Schedule of Events | Search Abstracts | Invited Symposia | Symposia | Poster Sessions | Data Blitz

Structural Segmentation with Freesurfer and MUSE for use in fMRI Data in Healthy Young and Older Adults

Poster Session F - Tuesday, March 10, 2026, 8:00 – 10:00 am PDT, Fairview/Kitsilano Ballroom

Kana Kimura¹ (), Caitlin Bowman¹; ¹University of Wisconsin - Milwaukee

Cognitive[BC1.1] neuroscientists often compute functional analyses within predefined anatomical regions of interest (ROIs). Defining anatomical brain boundaries in different age groups is challenging due to age-related structural changes in older age.[BC2.1] An earlier version of one commonly used tool, Freesurfer (v6), has been shown to have differences across age groups in the segmentation of the hippocampus. Multi-atlas segmentation protocols have been developed that may help with this issue by taking the consensus label across multiple different segmentation protocols. To test differences in the suitability of different anatomy labeling approaches for use in fMRI studies of age differences, we compared two tools for automated labeling of brain anatomy (Freesurfer, v.8.1 0 and Nichart DLMUSE, v.0.1.7) in terms of the strength and variability of univariate activation elicited in several cortical and subcortical ROIs. Participants were cognitively healthy young (aged 18-30, n = 20) and older (aged 60-80, n = 16) adults, and their structural and functional images were collected while performing a task that involves viewing and remembering scene images. We did not find a significant overall effect of segmentation method or any significant interactions with segmentation method. The variability was also comparable, with slightly higher variability for multi-atlas segmentation in older adults. Thus, we found that single (freesurfer) vs. multi-atlas (Nichart DLMUSE) segmentation tools resulted in very similar estimates of functional activation in young and older adults, suggesting that researchers can use either tool when testing age differences in functional activation.

Topic Area: METHODS: Neuroimaging