Hi all,
Small question related to the number of structural runs required for optimal results: I was wondering whether someone tested the assumption that using two scans rather than one improves the segmentation process and the cortical surface reconstruction / thickness estimation? I've not been able to find any publication on it, but maybe my search was not extensive enough.
Thanks a lot for your answer,
Marie
Hi Marie, I have not seen a publication on it. I've played around with it a bit and could not convince myself that there was a huge win in using 2, but it also depends a bit on your acquisition. If you have a lot of acceleration or really small voxels, then maybe it makes a bigger difference. Also, if you have a population that moves a lot, it might be a good idea to get to scans even if you only use one doug
On 07/15/2013 02:55 PM, Marie Schaer wrote:
Hi all,
Small question related to the number of structural runs required for optimal results: I was wondering whether someone tested the assumption that using two scans rather than one improves the segmentation process and the cortical surface reconstruction / thickness estimation? I've not been able to find any publication on it, but maybe my search was not extensive enough.
Thanks a lot for your answer,
Marie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Thanks a lot Doug for your super rapid response!
I was asking because I previously had the impression that using one run was giving better results, but lately with version 5.3 I've found that averaging 2 runs tends to reduce the amount of topological errors that have to be corrected. But as you said, it may largely be related to the acquisition's quality. Would you generally agree that if the contrast and resolution is good, then averaging two runs may add some "fuzziness" to the averaged scan, whereas it would increase the signal to noise in acquisitions that are not perfect in terms of contrast / resolution?
Marie
On Jul 15, 2013, at 12:04 PM, Douglas N Greve greve@nmr.mgh.harvard.edu wrote:
Hi Marie, I have not seen a publication on it. I've played around with it a bit and could not convince myself that there was a huge win in using 2, but it also depends a bit on your acquisition. If you have a lot of acceleration or really small voxels, then maybe it makes a bigger difference. Also, if you have a population that moves a lot, it might be a good idea to get to scans even if you only use one doug
On 07/15/2013 02:55 PM, Marie Schaer wrote:
Hi all,
Small question related to the number of structural runs required for optimal results: I was wondering whether someone tested the assumption that using two scans rather than one improves the segmentation process and the cortical surface reconstruction / thickness estimation? I've not been able to find any publication on it, but maybe my search was not extensive enough.
Thanks a lot for your answer,
Marie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
I only have anecdotal evidence on this. (using version 5.1) I found that if both scans had "a little" motion, then using both improved reconstruction. If one scan was pretty bad, and the other was pretty good, then using the pretty good one on its own was better than using both. If both were pretty bad, then using one or both didn't matter. (poor results either way) If the scan was good in the first place, I didn't repeat acquisition. In one subject, the first scan only had a little motion, and the 2nd scan was very good; in this case using both was worse than using just the very good scan.
Unfortunately I can't define "pretty good" or "pretty bad". My impression is that the only times reconstruction failed were in the presence of very large lesions, or in a subject who would have moved too much no matter what, and averaging even a dozen scans wouldn't help.
On 07/15/2013 03:19 PM, Marie Schaer wrote:
Thanks a lot Doug for your super rapid response!
I was asking because I previously had the impression that using one run was giving better results, but lately with version 5.3 I've found that averaging 2 runs tends to reduce the amount of topological errors that have to be corrected. But as you said, it may largely be related to the acquisition's quality. Would you generally agree that if the contrast and resolution is good, then averaging two runs may add some "fuzziness" to the averaged scan, whereas it would increase the signal to noise in acquisitions that are not perfect in terms of contrast / resolution?
Marie
On Jul 15, 2013, at 12:04 PM, Douglas N Greve greve@nmr.mgh.harvard.edu wrote:
Hi Marie, I have not seen a publication on it. I've played around with it a bit and could not convince myself that there was a huge win in using 2, but it also depends a bit on your acquisition. If you have a lot of acceleration or really small voxels, then maybe it makes a bigger difference. Also, if you have a population that moves a lot, it might be a good idea to get to scans even if you only use one doug
On 07/15/2013 02:55 PM, Marie Schaer wrote:
Hi all,
Small question related to the number of structural runs required for optimal results: I was wondering whether someone tested the assumption that using two scans rather than one improves the segmentation process and the cortical surface reconstruction / thickness estimation? I've not been able to find any publication on it, but maybe my search was not extensive enough.
Thanks a lot for your answer,
Marie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Yes, I think that tow runs can add some blurring. I think Martin may have looked at this too. Using 1 run instead of 2 will likely increase the number of topological defects, but most of these are fixed properly anyway.
doug
On 07/15/2013 03:19 PM, Marie Schaer wrote:
Thanks a lot Doug for your super rapid response!
I was asking because I previously had the impression that using one run was giving better results, but lately with version 5.3 I've found that averaging 2 runs tends to reduce the amount of topological errors that have to be corrected. But as you said, it may largely be related to the acquisition's quality. Would you generally agree that if the contrast and resolution is good, then averaging two runs may add some "fuzziness" to the averaged scan, whereas it would increase the signal to noise in acquisitions that are not perfect in terms of contrast / resolution?
Marie
On Jul 15, 2013, at 12:04 PM, Douglas N Greve greve@nmr.mgh.harvard.edu wrote:
Hi Marie, I have not seen a publication on it. I've played around with it a bit and could not convince myself that there was a huge win in using 2, but it also depends a bit on your acquisition. If you have a lot of acceleration or really small voxels, then maybe it makes a bigger difference. Also, if you have a population that moves a lot, it might be a good idea to get to scans even if you only use one doug
On 07/15/2013 02:55 PM, Marie Schaer wrote:
Hi all,
Small question related to the number of structural runs required for optimal results: I was wondering whether someone tested the assumption that using two scans rather than one improves the segmentation process and the cortical surface reconstruction / thickness estimation? I've not been able to find any publication on it, but maybe my search was not extensive enough.
Thanks a lot for your answer,
Marie
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: https://gate.nmr.mgh.harvard.edu/filedrop2 www.nmr.mgh.harvard.edu/facility/filedrop/index.html Outgoing: ftp://surfer.nmr.mgh.harvard.edu/transfer/outgoing/flat/greve/
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
The information in this e-mail is intended only for the person to whom it is addressed. If you believe this e-mail was sent to you in error and the e-mail contains patient information, please contact the Partners Compliance HelpLine at http://www.partners.org/complianceline . If the e-mail was sent to you in error but does not contain patient information, please contact the sender and properly dispose of the e-mail.
freesurfer@nmr.mgh.harvard.edu
-
Chris Watson -
Douglas N Greve -
Marie Schaer