From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Thu Aug 14 05:46:07 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Thu, 14 Aug 2014 11:46:07 +0200 Subject: [Rtk-users] SART with Elekta Dataset Message-ID: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Picture (Device Independent Bitmap) 1.jpg Type: image/jpeg Size: 10261 bytes Desc: Picture (Device Independent Bitmap) 1.jpg URL: From cyril.mory at creatis.insa-lyon.fr Thu Aug 14 08:57:48 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Thu, 14 Aug 2014 14:57:48 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: Message-ID: <53ECB24C.4030104@creatis.insa-lyon.fr> Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > Hello, > I?ve just begun to work with RTK. > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > My commands: > rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF > --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > One slice of the result: > With rtkfdk its working fine (like in the wiki). > Can you help me please? > Thanks. > Daniel > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cyril.mory at creatis.insa-lyon.fr Mon Aug 18 10:51:37 2014 From: cyril.mory at creatis.insa-lyon.fr (Cyril Mory) Date: Mon, 18 Aug 2014 16:51:37 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: References: <53ECB24C.4030104@creatis.insa-lyon.fr> Message-ID: <53F212F9.6060801@creatis.insa-lyon.fr> Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: > > Thanks! > > My goal is to use RTK for Reconstruction of Siemens Artis Zeego > projection data. You?ve worked with a Philips Scanner -- so as I > understood from your PhD thesis. If you have any advices regarding > writing code for the read-in of data this would be very helpful for me > (I?m a physicist and no real programmer...) > > Again thanks a lot and have a nice weekend! > > Daniel > > *Von:*Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] > *Gesendet:* Donnerstag, 14. August 2014 14:58 > *An:* Schoenke, Daniel; rtk-users at openrtk.org > *Betreff:* Re: [Rtk-users] SART with Elekta Dataset > > Hi Daniel, > > Welcome to RTK. It's always a pleasure to have new people around. > Your problem is a frequently encountered issue with SART: your > reconstruction volume is too small, and some of the object lies > outside of it. > > Another user recently had a similar issue, so I'll copy/paste what I > wrote for her : > An intuitive way to understand this is the following: the algorithm > tries to "explain" the attenuation it sees on the projections. If the > data has been partially truncated, some rays have "extra attenuation" > that has to go somewhere in the reconstructed volume. The center of > the reconstructed volume is constrained by all the other rays, so this > "extra attenuation" is pushed to the borders (or the corners). If, on > the other hand, the reconstructed volume is large enough, this "extra > attenuation" can go where it belongs, and does not cause hyper > attenuation on the borders. > > I've run the same command as you with spacing 2, and it gives a nice > reconstruction result (except for the border artifacts on the bottom > of the phantom, caused by the exact same phenomenon, which you can get > rid of by increasing the reconstructed volume's size again). So just > make sure your object is fully contained in the reconstructed volume, > and it should work. > > Best regards, > Cyril > > On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: > > Hello, > > I?ve just begun to work with RTK. > > Question: Why do I get strange results for the SART reco with the > Elekta dataset? > > My commands: > > rtkelektasynergygeometry --image_db > ../../../elekta/elekta/IMAGE.DBF --frame_db > ../../../elekta/elekta/FRAME.DBF --dicom_uid > 1.3.46.423632.135428.1351013645.166 -o elektaGeometry > > rtksart -g elektaGeometry -p > ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r > .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 > > One slice of the result: > > With rtkfdk its working fine (like in the wiki). > > Can you help me please? > > Thanks. > > Daniel > > > > > _______________________________________________ > > Rtk-users mailing list > > Rtk-users at public.kitware.com > > http://public.kitware.com/mailman/listinfo/rtk-users > > > > -- > -- > Cyril Mory, Post-doc > CREATIS > Leon Berard cancer treatment center > 28 rue La?nnec > 69373 Lyon cedex 08 FRANCE > > Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: From Daniel.Schoenke at medma.uni-heidelberg.de Tue Aug 19 09:19:14 2014 From: Daniel.Schoenke at medma.uni-heidelberg.de (Schoenke, Daniel) Date: Tue, 19 Aug 2014 15:19:14 +0200 Subject: [Rtk-users] SART with Elekta Dataset In-Reply-To: <53F212F9.6060801@creatis.insa-lyon.fr> References: <53ECB24C.4030104@creatis.insa-lyon.fr> <53F212F9.6060801@creatis.insa-lyon.fr> Message-ID: I get one image file (.bin or .nrrd format like data at http://www5.cs.fau.de/conrad/data/heart/) with all projections and a geometry matrix file (same formats available) from the Data Conversion Tool provided by Siemens. Today I get the recommendation to use the CONRAD toolkit for my purpose because it?s adapted for these data files. So I will at first try to get CONRAD. If this shall not work I?ll get in touch again. Thanks a lot! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Montag, 18. August 2014 16:52 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: AW: [Rtk-users] SART with Elekta Dataset Hi Daniel, Please keep the discussion about RTK on the mailing list : this conversation might come in handy to someone else in the future. Regarding the data acquisition, here is what you need : - Raw pixel data and meta-data of the projection images (number of pixels per line and column, number of lines and columns, number of projections, bits per pixel, size of the pixels in mm) - Geometry data (you can find some information on the RTK geometry on the wiki, http://wiki.openrtk.org/index.php/Main_Page#Geometry). It usually consists in a few (from 9 to 15) floats per projection I worked with a Philips C-Arm during my PhD indeed. I was employed by Philips, so I had the tools to export the raw data and geometry information readily available. Depending on how well-documented the data you have is, it may range from very easy to very difficult to extract this information (if you have to look for it manually in undocumented Dicom fields, for example, it may take time, while if all you need is a geometry converter, it's a matter of a few hours). Can you be precise on what kind of data you are able to extract from the Zeego ? Regards, Cyril On 08/15/2014 04:16 PM, Schoenke, Daniel wrote: Thanks! My goal is to use RTK for Reconstruction of Siemens Artis Zeego projection data. You?ve worked with a Philips Scanner - so as I understood from your PhD thesis. If you have any advices regarding writing code for the read-in of data this would be very helpful for me (I?m a physicist and no real programmer...) Again thanks a lot and have a nice weekend! Daniel Von: Cyril Mory [mailto:cyril.mory at creatis.insa-lyon.fr] Gesendet: Donnerstag, 14. August 2014 14:58 An: Schoenke, Daniel; rtk-users at openrtk.org Betreff: Re: [Rtk-users] SART with Elekta Dataset Hi Daniel, Welcome to RTK. It's always a pleasure to have new people around. Your problem is a frequently encountered issue with SART: your reconstruction volume is too small, and some of the object lies outside of it. Another user recently had a similar issue, so I'll copy/paste what I wrote for her : An intuitive way to understand this is the following: the algorithm tries to "explain" the attenuation it sees on the projections. If the data has been partially truncated, some rays have "extra attenuation" that has to go somewhere in the reconstructed volume. The center of the reconstructed volume is constrained by all the other rays, so this "extra attenuation" is pushed to the borders (or the corners). If, on the other hand, the reconstructed volume is large enough, this "extra attenuation" can go where it belongs, and does not cause hyper attenuation on the borders. I've run the same command as you with spacing 2, and it gives a nice reconstruction result (except for the border artifacts on the bottom of the phantom, caused by the exact same phenomenon, which you can get rid of by increasing the reconstructed volume's size again). So just make sure your object is fully contained in the reconstructed volume, and it should work. Best regards, Cyril On 08/14/2014 11:46 AM, Schoenke, Daniel wrote: Hello, I?ve just begun to work with RTK. Question: Why do I get strange results for the SART reco with the Elekta dataset? My commands: rtkelektasynergygeometry --image_db ../../../elekta/elekta/IMAGE.DBF --frame_db ../../../elekta/elekta/FRAME.DBF --dicom_uid 1.3.46.423632.135428.1351013645.166 -o elektaGeometry rtksart -g elektaGeometry -p ../../../elekta/elekta/img_1.3.46.423632.135428.1351013645.166/ -r .*.his -o sart.elekta.mha --verbose --spacing 1 --dimension 128 One slice of the result: [cid:7a88d6f7-52d2-4b5c-bd5f-534b77cd64c2] With rtkfdk its working fine (like in the wiki). Can you help me please? Thanks. Daniel _______________________________________________ Rtk-users mailing list Rtk-users at public.kitware.com http://public.kitware.com/mailman/listinfo/rtk-users -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -- -- Cyril Mory, Post-doc CREATIS Leon Berard cancer treatment center 28 rue La?nnec 69373 Lyon cedex 08 FRANCE Mobile: +33 6 69 46 73 79 -------------- next part -------------- An HTML attachment was scrubbed... URL: