[IGSTK-Users] CTImageReader voxel origin

[Chris Heath]

Andinet,

While my IDL application took into account patient orientation (which  
is why my axes matched up), it did not take into account the absolute  
position of the patient as recorded in the DICOM header.  The field is  
called "Image Position Patient" (0032) and it contains the translation  
vector that I was missing.

Thanks,
Chris

On Jul 7, 2008, at 6:28 PM, Andinet Enquobahrie wrote:

> Hi Chris,
>
> Most likely your IDL application is not taking into account patient
> orientation (direction cosines ) information during index-physical
> coordinate system conversion.
>
> In IGSTK ( by way of the ITK image readers and image classes) ,
> we use spacing, origin and orientation information to convert  between
> the two coordinate systems.
>
> This should probably the first thing you might want to check. Check  
> your
> input dicom CT images  Orientation (0020,0037) . This tag specifies  
> the
> direction cosines of the first row and the first column with respect
> to the patient.
> Let us know what you find out...
>
> -Andinet
>
>
>> Hello,
>>
>> I am trying to integrate another application (in IDL) with IGSTK.   
>> The purpose of the IDL code is to locate the centers of fiducial  
>> markers from a CT image.  The code outputs the x, y, and z  
>> coordinates of the markers.  I would like to use these values with  
>> igstkLandmark3DRegistration to register tracker space to CT space.   
>> However, the coordinates I receive from my IDL code are offset by  
>> an obscure amount from their corresponding values as reported by  
>> IGSTK.  (I find the IGSTK-reported coordinates using the "picker"  
>> code from NeedleBiopsy). My CTImageSpatialObject's parent is  
>> m_WorldReference with the Identity transform relationship between  
>> them.
>>
>> In this setup, how can I determine where the origin of the CT image  
>> is?  Is there anyway to easily transform the CTImage so that all  
>> coordinates are positive and the origin (0,0,0) is located in the  
>> corner of the first slice?  I cannot simply find the translation  
>> between the two and apply it because the values I get from IGSTK  
>> are not as precise as the ones I get from IDL.  Also, the  
>> relationship changes for different CT Images.
>>
>> Thanks,
>> Chris
>>
>>
>> Christopher Heath
>> Vanderbilt University School of Engineering
>> christopher.c.heath at vanderbilt.edu
>>
>>
>>
>>
>
>
> -- 
> ==========================================================
> Andinet A. Enquobahrie, PhD
> R&D Engineer
> Kitware Inc.
>
> 28 Corporate Drive
> Clifton Park, NY 12065-8662
> Phone: 518-371-3971 x124
> www.kitware.com
>
>