From anais.capouillez at student.uliege.be Wed Apr 11 09:37:34 2018 From: anais.capouillez at student.uliege.be (anais.capouillez at student.uliege.be) Date: Wed, 11 Apr 2018 15:37:34 +0200 (CEST) Subject: [Rtk-users] More errors with 360 projections than 180 for SART Message-ID: <1998861004.8150225.1523453854483.JavaMail.zimbra@student.uliege.be> Hi, I'm reconstructing a homogeneous cylinder. My problem is that I have better results for 180 projections, than for 360 for SART (with default parameters), and the same geometry (except the number of projections). I don't understand why adding projections adds more errors, since it should be better (and for other algorithms this is indeed better with 360 projections). With 180 projections, the mean error on the cylinder is 1,6% , and the maximal error is 21,6%. While, with 360 projections, on the cylinder the mean error is 6,4%, and the maximal error is 169%. Even if I don't take into account voxels close to edges, there are more errors with 360 projections. Is there a reason why I obtain worse results with 360 projections? Since my files are too big, I added them here https://drive.google.com/drive/folders/1AYumjqNGUr8DZ0kxIbF7Ab1xet7srk-z?usp=sharing. I also added a file with the commands I used. Thank you very much. Ana?s From wuchao04 at gmail.com Wed Apr 11 15:39:07 2018 From: wuchao04 at gmail.com (Chao Wu) Date: Wed, 11 Apr 2018 21:39:07 +0200 Subject: [Rtk-users] More errors with 360 projections than 180 for SART In-Reply-To: <1998861004.8150225.1523453854483.JavaMail.zimbra@student.uliege.be> References: <1998861004.8150225.1523453854483.JavaMail.zimbra@student.uliege.be> Message-ID: Hi Anais, I am travelling so I cannot review your images, but please check if the following applies to you: Open issue: https://github.com/SimonRit/RTK/issues/151 Related thread: https://public.kitware.com/pipermail/rtk-users/2018-February/010575.html Regards, Chao 2018-04-11 15:37 GMT+02:00 : > Hi, > > I'm reconstructing a homogeneous cylinder. My problem is that I have > better results for 180 projections, than for 360 for SART (with default > parameters), and the same geometry (except the number of projections). > > I don't understand why adding projections adds more errors, since it > should be better (and for other algorithms this is indeed better with 360 > projections). > > With 180 projections, the mean error on the cylinder is 1,6% , and the > maximal error is 21,6%. While, with 360 projections, on the cylinder the > mean error is 6,4%, and the maximal error is 169%. > Even if I don't take into account voxels close to edges, there are more > errors with 360 projections. > > Is there a reason why I obtain worse results with 360 projections? > > Since my files are too big, I added them here https://drive.google.com/ > drive/folders/1AYumjqNGUr8DZ0kxIbF7Ab1xet7srk-z?usp=sharing. I also added > a file with the commands I used. > > > Thank you very much. > > Ana?s > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > https://public.kitware.com/mailman/listinfo/rtk-users > -------------- next part -------------- An HTML attachment was scrubbed... URL: From simon.rit at creatis.insa-lyon.fr Thu Apr 12 01:50:54 2018 From: simon.rit at creatis.insa-lyon.fr (Simon Rit) Date: Thu, 12 Apr 2018 07:50:54 +0200 Subject: [Rtk-users] More errors with 360 projections than 180 for SART In-Reply-To: References: <1998861004.8150225.1523453854483.JavaMail.zimbra@student.uliege.be> Message-ID: Hi Anais, The suggestion posted by Chao might be helpful to improve both results. My intuition is that this is caused by the fact that SART has no regularization and it is know that it will start to diverge from the solution after a few iterations. Setting the number of iteration is a way of avoiding divergence. Now, if you double the number of projections, I would say that one iteration goes two times faster to the solution. So I would compare a halved number of iterations when doubling the number of projections. I would suggest to use regularized conjugate gradient and to compare the results at convergence instead... I hope this helps, Simon On Wed, Apr 11, 2018 at 9:39 PM, Chao Wu wrote: > Hi Anais, > > I am travelling so I cannot review your images, but please check if the > following applies to you: > Open issue: https://github.com/SimonRit/RTK/issues/151 > Related thread: https://public.kitware.com/pipermail/rtk- > users/2018-February/010575.html > > Regards, > Chao > > 2018-04-11 15:37 GMT+02:00 : > >> Hi, >> >> I'm reconstructing a homogeneous cylinder. My problem is that I have >> better results for 180 projections, than for 360 for SART (with default >> parameters), and the same geometry (except the number of projections). >> >> I don't understand why adding projections adds more errors, since it >> should be better (and for other algorithms this is indeed better with 360 >> projections). >> >> With 180 projections, the mean error on the cylinder is 1,6% , and the >> maximal error is 21,6%. While, with 360 projections, on the cylinder the >> mean error is 6,4%, and the maximal error is 169%. >> Even if I don't take into account voxels close to edges, there are more >> errors with 360 projections. >> >> Is there a reason why I obtain worse results with 360 projections? >> >> Since my files are too big, I added them here >> https://drive.google.com/drive/folders/1AYumjqNGUr8DZ0kxIbF7 >> Ab1xet7srk-z?usp=sharing. I also added a file with the commands I used. >> >> >> Thank you very much. >> >> Ana?s >> _______________________________________________ >> Rtk-users mailing list >> Rtk-users at public.kitware.com >> https://public.kitware.com/mailman/listinfo/rtk-users >> > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > https://public.kitware.com/mailman/listinfo/rtk-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From anais.capouillez at student.uliege.be Thu Apr 12 16:50:07 2018 From: anais.capouillez at student.uliege.be (anais.capouillez at student.uliege.be) Date: Thu, 12 Apr 2018 22:50:07 +0200 (CEST) Subject: [Rtk-users] More errors with 360 projections than 180 for SART In-Reply-To: References: <1998861004.8150225.1523453854483.JavaMail.zimbra@student.uliege.be> Message-ID: <1777619911.8784950.1523566207667.JavaMail.zimbra@student.uliege.be> Hi, Thank you, this is probably the reason. Ana?s ----- Mail original ----- De: "Simon Rit" ?: "Chao Wu" Cc: "anais capouillez" , "rtk-users" Envoy?: Jeudi 12 Avril 2018 07:50:54 Objet: Re: [Rtk-users] More errors with 360 projections than 180 for SART Hi Anais, The suggestion posted by Chao might be helpful to improve both results. My intuition is that this is caused by the fact that SART has no regularization and it is know that it will start to diverge from the solution after a few iterations. Setting the number of iteration is a way of avoiding divergence. Now, if you double the number of projections, I would say that one iteration goes two times faster to the solution. So I would compare a halved number of iterations when doubling the number of projections. I would suggest to use regularized conjugate gradient and to compare the results at convergence instead... I hope this helps, Simon On Wed, Apr 11, 2018 at 9:39 PM, Chao Wu wrote: > Hi Anais, > > I am travelling so I cannot review your images, but please check if the > following applies to you: > Open issue: https://github.com/SimonRit/RTK/issues/151 > Related thread: https://public.kitware.com/pipermail/rtk- > users/2018-February/010575.html > > Regards, > Chao > > 2018-04-11 15:37 GMT+02:00 : > >> Hi, >> >> I'm reconstructing a homogeneous cylinder. My problem is that I have >> better results for 180 projections, than for 360 for SART (with default >> parameters), and the same geometry (except the number of projections). >> >> I don't understand why adding projections adds more errors, since it >> should be better (and for other algorithms this is indeed better with 360 >> projections). >> >> With 180 projections, the mean error on the cylinder is 1,6% , and the >> maximal error is 21,6%. While, with 360 projections, on the cylinder the >> mean error is 6,4%, and the maximal error is 169%. >> Even if I don't take into account voxels close to edges, there are more >> errors with 360 projections. >> >> Is there a reason why I obtain worse results with 360 projections? >> >> Since my files are too big, I added them here >> https://drive.google.com/drive/folders/1AYumjqNGUr8DZ0kxIbF7 >> Ab1xet7srk-z?usp=sharing. I also added a file with the commands I used. >> >> >> Thank you very much. >> >> Ana?s >> _______________________________________________ >> Rtk-users mailing list >> Rtk-users at public.kitware.com >> https://public.kitware.com/mailman/listinfo/rtk-users >> > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > https://public.kitware.com/mailman/listinfo/rtk-users > > From 1874415034 at qq.com Thu Apr 26 09:30:52 2018 From: 1874415034 at qq.com (=?gb18030?B?ZGFnZXR1bw==?=) Date: Thu, 26 Apr 2018 21:30:52 +0800 Subject: [Rtk-users] CBCT Reconstruction problem In-Reply-To: References: Message-ID: Dear RTK-users, I have been trying to run the rtkfdk code for CBCT reconstruction but the result was not quite right. I have attached a drawing of my setup and projections could be downloaded from: https://drive.google.com/drive/folders/1J0am5Hw67O8b5dmKNlEno4MOnbBZvsM0?usp=sharing Description of our CBCT setup: 1. The distance from Source to Isocenter is 602.00mm 2. The distance from Isocenter to Imaging Surface of the Detector is 403.84mm 3. Detector physical size: 434mm*434mm 4. Image Resolution: 2816*2816 5. angle step = 1 degree, angular covered is 205 degrees. The original size of projections was 2816*2816 and I resized them to (256,256) for faster calculation. This is how I performed my reconstruction: ( rtksimulatedgeometry -n 205 -o I:\RTK140_build_git_for_test\bin\Debug\geo.xml --sdd=1005.84 --sid=602.0 --proj_iso_y=-128.0 --proj_iso_x=-128.0 --arc=205.0 rtkfdk --verbose 1 --path=I:\RTK140_build_git_for_test\bin\Debug\timo1 --regexp=DCM$ -o I:\RTK140_build_git_for_test\bin\Debug\test3.mha --hardware="cuda" -g I:\RTK140_build_git_for_test\bin\Debug\geo.xml --hann 1.0 --dimension 256,256,256 ) Could you please tell me if I am setting the geometry correctly? Thank you in advance. Chole -------------- next part -------------- An HTML attachment was scrubbed... URL: From wuchao04 at gmail.com Thu Apr 26 10:57:04 2018 From: wuchao04 at gmail.com (Chao Wu) Date: Thu, 26 Apr 2018 16:57:04 +0200 Subject: [Rtk-users] CBCT Reconstruction problem In-Reply-To: References: Message-ID: Hi Chole, --proj_iso_x and --proj_iso_y use the same unit as in other geometrical parameters, so in your case those should be in mm not in pixels. Regards, Chao 2018-04-26 15:30 GMT+02:00 dagetuo <1874415034 at qq.com>: > Dear RTK-users, > > I have been trying to run the rtkfdk code for CBCT reconstruction but the > result was not quite right. > I have attached a drawing of my setup and projections could be downloaded > from: > https://drive.google.com/drive/folders/1J0am5Hw67O8b5dmKNlEno4MOnbBZv > sM0?usp=sharing > > Description of our CBCT setup: > 1. The distance from Source to Isocenter is 602.00mm > 2. The distance from Isocenter to Imaging Surface of the Detector is > 403.84mm > 3. Detector physical size: 434mm*434mm > 4. Image Resolution: 2816*2816 > 5. angle step = 1 degree, angular covered is 205 degrees. > > The original size of projections was 2816*2816 and I resized them to > (256,256) for faster calculation. > > This is how I performed my reconstruction: > ( rtksimulatedgeometry -n 205 -o I:\RTK140_build_git_for_test\bin\Debug\geo.xml > --sdd=1005.84 --sid=602.0 --proj_iso_y=-128.0 --proj_iso_x=-128.0 > --arc=205.0 > > rtkfdk --verbose 1 --path=I:\RTK140_build_git_for_test\bin\Debug\timo1 > --regexp=DCM$ -o I:\RTK140_build_git_for_test\bin\Debug\test3.mha > --hardware="cuda" -g I:\RTK140_build_git_for_test\bin\Debug\geo.xml > --hann 1.0 --dimension 256,256,256 > ) > > Could you please tell me if I am setting the geometry correctly? > > Thank you in advance. > > Chole > > > > > > _______________________________________________ > Rtk-users mailing list > Rtk-users at public.kitware.com > https://public.kitware.com/mailman/listinfo/rtk-users > > -------------- next part -------------- An HTML attachment was scrubbed... URL: