Christopher G. Schwarz

Changing the face of brain image processing research. Views expressed are my own.

[Christopher G. Schwarz @ChrisGSchwarz]

MCSA have now uploaded 3588 (de-faced) MRIs and longitudinal clinical and demographic data from over 22,000 participants for public sharing on https://t.co/iUaxyF0Ao8 ! This is only the beginning of our huge data sharing initiatives. Check it out! https://t.co/uQxE0ogKZp

[Christopher G. Schwarz @ChrisGSchwarz]

@_JakeVogel_ The paper very clearly still recommends defacing. It just tried to push back on the idea that defacing may not be enough. Still, I'm not sure why people want to argue toward less thorough defacing when we have better protection with equivalent or better downstream effects.

[Christopher G. Schwarz @ChrisGSchwarz]

New in @eic_nic: We compared correlations between AD imaging biomarkers and cognition, using unmodified images vs. images de-faced with mri_reface. All of the de-faced analyses were statistically identical both qualitatively and in statistical power.
https://t.co/Pz1F9gWb6b

[Christopher G. Schwarz @ChrisGSchwarz]

Our latest paper is now in NeuroImage: https://t.co/cpXOEYoZfZ We look at face recognition and de-facing for many common research brain MRI sequences. De-facing is not needed for current-gen fMRI, dMRI, and ASL, but mri_reface is recommended for all other tested sequences.

[Christopher G. Schwarz @ChrisGSchwarz]

We have released mri_reface version 0.3.2, with several minor improvements for DICOM, 2D FLAIR, and larger slices. By popular demand, starting with this release we also include a complete Docker image to standardize and simplify installing and running:
https://t.co/pfQm7IScbU

[Christopher G. Schwarz @ChrisGSchwarz]

@CorriveauNick I will certainly give it a try, when they are ready for submissions. I really hope that Elsevier/NI will allow editors to finish out current papers under review! NICL is conspicuously absent, though. The new journal wants to include its scope, but the NICL editors did not leave?

[Christopher G. Schwarz @ChrisGSchwarz]

@anna_marseglia_ @CorriveauNick I don't know an article dedicated to that topic specifically, but this one at least describes that we do it that way:
https://t.co/6f7l3ftj6A

[Christopher G. Schwarz @ChrisGSchwarz]

@anna_marseglia_ @CorriveauNick Sometimes when we make meta-ROIs of cortical thickness from FreeSurfer, we take a weighted median, weighting each region's thickness by its relative surface area, but it's just simple multiplication. This is analogous to volume-weighted medians for meta-ROIs of PET SUVR.

[Christopher G. Schwarz @ChrisGSchwarz]

Tip for people registering for AAIC: If trying to pay via credit card gives you an error "adding your card", select "Paypal" as your payment method, and then use your credit card inside the Paypal popup instead. Paypal->Card worked, but Card (Via Paypal) did not.

[Christopher G. Schwarz @ChrisGSchwarz]

In a new study by Robel Gebre and @prashvemuri, we compared 6 methods to harmonize FreeSurfer measurements across 2 different-vendor scanners using a direct cross-over dataset. All performed badly. It's disappointing, but a very important result to share!
https://t.co/KJImcHVuBG

[Christopher G. Schwarz @ChrisGSchwarz]

Future versions of mri_reface will require niftyreg >= 1.5.

[Christopher G. Schwarz @ChrisGSchwarz]

mri_reface users should ensure that you're using niftyreg (reg_aladin) version >= 1.5.x. mri_reface already warns about a "very old version", but users are reporting that registration failure rates can be much worse with 1.3.x. Compile from https://t.co/HqF6cU51KP

[Christopher G. Schwarz @ChrisGSchwarz]

My focus is mainly on brain images, but don't forget about re-identification risks from speech audio samples in research datasets as well! See this new thorough analysis by our @NaipMayo colleagues.

[Christopher G. Schwarz @ChrisGSchwarz]

@eHealthUOC @AdasLab @UOCuniversity @SpringerIME Hi, you have the wrong Schwarz. This paper isn't one of mine.

[Christopher G. Schwarz @ChrisGSchwarz]

@CorriveauNick after enough years of redundancy I gave up on my personal CV. Now I just have re aims and NIH.

[Prof. Melissa E. Murray 📸🔬 @DrNeuroChic]

The time has finally arrived, @HAIconference is back to gather the world's leaders and super fans of #AlzheimersDisease imaging & fluid biomarker research
✔️October 10 abstract opens
✔️ October 28 abstracts close
✔️ January 11-13 meeting dates in Miami
https://t.co/PAA0OmHXBB

[Christopher G. Schwarz @ChrisGSchwarz]

@relajoie @tessamharrison I migrated Mendeley to Zotero about a year ago also, once they added the internal PDF reader. The migration itself was pretty painful, and I still have a lot of papers with duplicate identical attachments that I slowly clean up, but overall I've had no regrets.

[Brian Gordon @BrianGordon81]

Careful methodological work doesn't always get the credit it deserves, by @shaneyflores at @MIRimaging just published a really nice piece of work looking at tau PET binding in the skull and whether it biases estimates of AD pathology
https://t.co/QvA38p2zI3

[Christopher G. Schwarz @ChrisGSchwarz]

@txusser I can do better than that: the paper is published and everything on the poster is in it: https://t.co/J0vRTodmxU

[Christopher G. Schwarz @ChrisGSchwarz]

At #AAIC2022? Do you manage data sharing for a brain imaging data set? Come by my poster 197 to learn about de-facing MRI and PET and CT scans to prevent potential re-identification of participants with face recognition