StereoEEG Dynamics During Visual-Motor Interaction with a Robotic Hand
We developed a method for decoding hand gestures from stereo-electroencephalographic (sEEG) signals using a transformer-based model. Seven patients with sEEG implants reproduced gestures performed by a prosthetic hand. Neural decoding was conducted offline on these data. Time-frequency features of the sEEG were processed using the transformer architecture with spatial and temporal attention. Participant-specific models were trained, achieving up to 0.83 accuracy for individual finger gestures. These results highlight the potential of combining sEEG with artificial intelligence (AI) for intuitive control of prosthetic limbs with brain activity.
Authors
- Mikhail Lebedev (ORCID: https://orcid.org/0000-0003-0355-8723)
- O. Andrei Dragoi
- Anna Makarova (ORCID: https://orcid.org/0000-0002-6902-1630)
- О Э Карпов (ORCID: https://orcid.org/0000-0002-5227-0657)
- Gurgen Soghoyan
- А. А. Зуев (ORCID: https://orcid.org/0000-0003-2974-1462)
- Timur Sajfutdinov
Institutions
- National Research University Higher School of Economics (RU)
- Skolkovo Institute of Science and Technology (RU)
- Institute of Evolutionary Physiology and Biochemistry (RU)
- Moscow International Higher Business School (RU)
- Institute for Information Transmission Problems (RU)
- National Medical and Surgical Center named after N.I. Pirogov (RU)
Publication Details
- Published
- 2025-08-25
- DOI
- https://doi.org/10.1109/cnn67635.2025.11177531
- Primary Topic
- Teleoperation and Haptic Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00