Hi Bruce,

Yes - of course that helps :)

I did not even know that it was a hard question!

I had noticed that some results from freesurfer, and other segmentation techniques, subsequently fail manual QC.

Do you think that there would be value in some kind of parameter sensitivity analysis for freesurfer? Is the test suite publicly available?

Thank you very much. Mark

-----Original Message----- From: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Sent: November-15-13 8:14 AM To: Mark Alexiuk Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] are there any previous studies on the robustness of freesurfer default parameters as used in cortical thickness computation?

Hi Mark

that's a tough question to answer, as it depends on the parameter. I'm sure that the answer is "yes" for some, but in general we try very hard to adaptively estimate our parameters, which is why our tools work well on a wide variety of T1-weighted acquisitions. We also have a test suite with different acquisitions and pathologies that we run routinely to make sure that things continue to work at the same level of accuracy.

sorry, not sure that helps Bruce

On Thu, 14 Nov 2013, Mark Alexiuk wrote:

Hi,

My interest is in cortical thickness measurements for longitudinal Alzheimer data, namely the ADNI dataset.

My understanding is that recon-all can be used to generate cortical thicknesses.

I am trying to understand how the default parameter values used in recon-all were arrived at.

Are the default parameters optimized in some sense?

How robust are the default parameters with respect to the resulting cortical thickness?

Will a small change in a parameter value result in a significant difference in some cortical thicknesses?

I was able to find the following and would appreciate references to good papers on this topic.

Excerpt from abstract:

Standard manual tracing and FreeSurfer-based analyses were performed in 77 participants including 67 cognitively normal individuals and 10 individuals with early Alzheimer's disease?.

The manual and FreeSurfer approaches yielded nearly identical estimates of amyloid burden (intraclass correlation = 0.98) as assessed by the mean cortical binding potential.

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.007 3377

Thanks for any forthcoming comments.

Mark

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.