Keynotes | K&Ts | GACs | Talks | Posters | Search

Poster B in Poster Session B: Tuesday, August 4, 2:00 – 3:45 pm, Kimmel Center, Shorin & Rosenthal Rooms

Neural Typicality in Naturalistic fMRI Defines an Interpretable Representation for Clinical Classification

Ido Har-Shalom¹, Roman Beliy¹, Dov Yellin¹, Aviv Jarbi¹, Michal Zamberg Elad¹, Hadar Kolb¹, Michal Ramot¹; ¹Weizmann Institute of Science

Presenter: Michal Ramot

Naturalistic paradigms evoke structured neural responses shared across individuals, providing a rich but high-dimensional signal for studying brain function. Translating this signal into clinically useful representations, however, remains challenging. Here, we use Neural Typicality (Ntyp), a voxelwise measure of alignment to a shared response during naturalistic movie viewing, to construct a compact representation of brain dynamics and apply it to diagnostic classification in people with multiple sclerosis (pwMS). Using this representation, we achieve reliable classification of pwMS versus healthy controls using simple feature extraction strategies, with data-driven decomposition yielding the strongest performance. Permutation testing confirms that classification reflects true disease-related structure. This was consistent across multiple movies acquired in separate scanning sessions. The dimensions driving classification correspond to coherent, bilaterally organized brain networks that can be projected back onto the cortex, demonstrating that predictive performance arises from structured, interpretable brain organization rather than a black-box model. The ability to inspect these networks directly links classification to interpretation. Together, these results show that naturalistic neural alignment defines a compact and meaningful representation of brain function that enables classification while revealing the neural structure underlying group differences.

Topic Area: Development, Individual Differences & Clinical Populations