سامانه همایش ها

صفحه اصلی / سی و دومین کنفرانس ملی و دهمین کنفرانس بین المللی مهندسی زیست پزشکی ایران

Personalized EEG Source Estimation in a Shape Drawing Task

نویسندگان :

Zakieh Hassanzadeh¹ Melisa Daryayi² Navid Entezari³ Fariba Bahrami⁴

1- دانشگاه تهران 2- دانشگاه تهران 3- دانشگاه تهران 4- دانشگاه تهران

کلمات کلیدی :

EEG source localization،, individual neural phenotypes،visuomotor learning،subject-specific modeling

چکیده :

Abstract— We investigated individual neural signatures during visuomotor learning using subject-specific EEG source localization. Five adults learned foot-drawing of geometric shapes across seven sessions, with EEG recorded during Sessions 1 and 7. Subject-specific MLP networks localized activity to 116 brain regions. Analysis of 145 trials revealed three distinct neural phenotypes: cognitive-control (cingulate-dominant), executive-planning (PFC-dominant), and visual-processing (V1-dominant). Unexpectedly, cingulate cortex showed highest activation probability (24.0%) over motor areas (19.5%). Learning patterns varied by shape: Star required 16.8% more total activation (summed across dominant regions), while B showed 4.1% reduction, indicating neural efficiency.. Critically, no brain regions were active across >2 subjects (maximum 40% overlap, Jaccard=0.23). Within-subject trial-to-trial consistency (0.71 ± 0.06) significantly exceeded between-subject consistency (0.59 ± 0.09), p<0.001, demonstrating that individuals maintain stable personal neural strategies rather than converging on universal patterns. These findings demonstrate motor learning relies on individual-specific neural solutions rather than universal mechanisms, mandating personalized approaches for BCIs and rehabilitation.