Reading movement straight from the brain, for the patients whose muscles cannot be read
Prosthetic hands are usually controlled by electromyography, which reads electrical activity in muscle. For conditions like Spinal Muscular Atrophy that is the one signal that is not there to read. Aditya built the alternative: a pipeline that classifies motor imagery straight from EEG, bypassing the muscle entirely. He benchmarked three deep learning architectures against three classical machine learning models on the PhysioNet Motor Movement/Imagery dataset, separating rest from imagined left-hand and right-hand movement, and ran every model twice, once on the five cleanest subjects and once across all 109, to separate raw accuracy from real generalization.
FIELDBrain Computer Interfaces & Machine Learning
RESULTAccepted at ISASF 2026, Sydney
VENUEISASF 2026, 2026
Aditya Arvind, YRI FellowBRAIN COMPUTER INTERFACE
BEFORE THE FELLOWSHIP
A high school student in New Jersey interested in machine learning and assistive technology, with no prior published research.
AFTER
First-author paper accepted for presentation at ISASF 2026 at Western Sydney University, an IEEE technical co-sponsored conference.
THE LEDGER
01First-author paper accepted at ISASF 2026, Western Sydney University, November 19, 2026
02ShallowConvNet with augmentation reached 75% accuracy, the strongest of all six architectures
03Benchmarked six models, EEGNet, ShallowConvNet and ResNet against SVM, LDA and Random Forest
04Trained across all 109 subjects in the PhysioNet dataset to test generalization, not just best-case accuracy
05Showed deep learning outperforms classical models at scale while classical models hold up better on limited data
