error in NIFTItoNRRD handling of space directions #310
Comments
|
@vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? |
|
Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson [email protected] @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. |
|
Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" [email protected] Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson [email protected] @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — |
|
Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. |
|
Vince, I am working through several issues with properly representing Nrrd files and Nifti files from DWIConvert. Another popular tool for many dwi processing to convert dicom to NIFTI is the dcm2niix tool. I’ve previously had conversations with the author of that tool, and I thought we had resolved the interpretation issues. This is a discussion about some data that I am working with that is from BC Ho’s study. Could you please comment if you think the comments are correct, or if you think that perhaps there is a mis-interpretation of the data that I sent? Thanks, Hans From: "Vincent A. Magnotta" [email protected] Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — |
|
Hans, There are two slice orderings that occur with Mosaic images from Siemens. These are completely independent.
Chris may have never changed the image sorting order of the image and that is why he has not seen the change in slice ordering within a Mosaic image. Vince Associate Professor From: Hans Johnson [email protected] Vince, I am working through several issues with properly representing Nrrd files and Nifti files from DWIConvert. Another popular tool for many dwi processing to convert dicom to NIFTI is the dcm2niix tool. I’ve previously had conversations with the author of that tool, and I thought we had resolved the interpretation issues. This is a discussion about some data that I am working with that is from BC Ho’s study. Could you please comment if you think the comments are correct, or if you think that perhaps there is a mis-interpretation of the data that I sent? Thanks, Hans From: "Vincent A. Magnotta" [email protected] Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. |
|
Vince, That is helpful information. Do you happen to know where to find the DICOM tags that can be used to interpret those conditions? Thanks, From: "Vincent A. Magnotta" [email protected] Hans, There are two slice orderings that occur with Mosaic images from Siemens. These are completely independent.
Chris may have never changed the image sorting order of the image and that is why he has not seen the change in slice ordering within a Mosaic image. Vince Associate Professor From: Hans Johnson [email protected] Vince, I am working through several issues with properly representing Nrrd files and Nifti files from DWIConvert. Another popular tool for many dwi processing to convert dicom to NIFTI is the dcm2niix tool. I’ve previously had conversations with the author of that tool, and I thought we had resolved the interpretation issues. This is a discussion about some data that I am working with that is from BC Ho’s study. Could you please comment if you think the comments are correct, or if you think that perhaps there is a mis-interpretation of the data that I sent? Thanks, Hans From: "Vincent A. Magnotta" [email protected] Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — |
|
I know that this must be stored in the Siemens Private header fields of either 0029, 1010 or 0029, 1020. I will try to dig up more information of this. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> Vince, That is helpful information. Do you happen to know where to find the DICOM tags that can be used to interpret those conditions? Thanks, From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, There are two slice orderings that occur with Mosaic images from Siemens. These are completely independent.
Chris may have never changed the image sorting order of the image and that is why he has not seen the change in slice ordering within a Mosaic image. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> Vince, I am working through several issues with properly representing Nrrd files and Nifti files from DWIConvert. Another popular tool for many dwi processing to convert dicom to NIFTI is the dcm2niix tool. I’ve previously had conversations with the author of that tool, and I thought we had resolved the interpretation issues. This is a discussion about some data that I am working with that is from BC Ho’s study. Could you please comment if you think the comments are correct, or if you think that perhaps there is a mis-interpretation of the data that I sent? Thanks, Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. |
|
Thank you Vince. Hans From: "Vincent A. Magnotta" [email protected] I know that this must be stored in the Siemens Private header fields of either 0029, 1010 or 0029, 1020. I will try to dig up more information of this. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> Vince, That is helpful information. Do you happen to know where to find the DICOM tags that can be used to interpret those conditions? Thanks, From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans, There are two slice orderings that occur with Mosaic images from Siemens. These are completely independent.
Chris may have never changed the image sorting order of the image and that is why he has not seen the change in slice ordering within a Mosaic image. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]> Vince, I am working through several issues with properly representing Nrrd files and Nifti files from DWIConvert. Another popular tool for many dwi processing to convert dicom to NIFTI is the dcm2niix tool. I’ve previously had conversations with the author of that tool, and I thought we had resolved the interpretation issues. This is a discussion about some data that I am working with that is from BC Ho’s study. Could you please comment if you think the comments are correct, or if you think that perhaps there is a mis-interpretation of the data that I sent? Thanks, Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]> Hans That may be the case. It is odd that the sform is wrong while the qform is correct. Vince Associate Professor On Nov 19, 2016, at 8:07 PM, Hans Johnson <[email protected]mailto:[email protected]mailto:[email protected]> wrote: Thanks Vince. I’ll need to look into this again. I thought that if the qform is set, then the sform is NOT set, and is not used. Hans From: "Vincent A. Magnotta" <[email protected]mailto:[email protected]mailto:[email protected]> Hans, I looked at this some on Friday. Since both transforms originate from the image direction matrix, it would assume the bug could be in the helper function that generates the sform matrix from the direction cosines. Vince Associate Professor From: Hans Johnson <[email protected]mailto:[email protected]mailto:[email protected]> @vmagnottahttps://github.com/vmagnotta Vince, I'm looking at this code now. Do you happen to know where in the code the problem is occurring? — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — — Notice: This UI Health Care e-mail (including attachments) is covered by the Electronic Communications Privacy Act, 18 U.S.C. 2510-2521 and is intended only for the use of the individual or entity to which it is addressed, and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please notify the sender immediately and delete or destroy all copies of the original message and attachments thereto. Email sent to or from UI Health Care may be retained as required by law or regulation. Thank you. — |
|
@vmagnotta I think this is now fixed in #315 I have a test suite of 85 visualizations using both fslview and Slicer to ensure that corrections are done. I'm through 15 visualization validations (a problematic phillips, a GE, and a Siemens), and will complete the rest of the data visualizations tomorrow. dcm2niix -> view in fslview I also added a new feature described in #314 , and updated the version number to 4.8.0. |
There seems to be an issue with the proper handling of image orientation in DWIConvert. If a DICOM series is directly converted to NiFTI versus the conversion from DICOM->NRRD->NIFTI the NIFTI sform matrix is not consistent while the qform matrix is. Slicer shows the image in the same orientation likely because it is using the qform orientation.
The problem is related to the srow_z rotation portion of the transform which is getting multiplied by -1. The translation part of the matrix is correct.
The text was updated successfully, but these errors were encountered: