A Scalable Framework for Closed-Loop Neuromodulation with Deep Learning

This repository contains supporting code for the GX Framework.

A Preprint of this work is publicly available here. Please cite this work as:

Gebodh N, Miskovic V, Laszlo S, Datta A, Bikson M. A Scalable Framework for Closed-Loop Neuromodulation with Deep Learning. bioRxiv [Preprint]. 2023 Jan 20:2023.01.18.524615. doi: 10.1101/2023.01.18.524615. PMID: 36712027; PMCID: PMC9882307.

A preview of this work can be seen below:

All acompanying code will be released upon publication.

Data availability

All the data used for this work can be accessed below:

• Raw EEG, ECG, EOG data in .cnt formant
• Raw EEG, ECG, EOG data formated to comply with BIDS standard where data are in .set format (EEGlab)
• Raw downsampled EEG, ECG, EOG data (1k Hz) in .mat format for Experiment 1 and Experiment 2 (works with MATLAB and Python)
• Raw behavioral CTT data .csv format
• Questionnaire data in .xlsx format

Acknowledgments

Portions of this study were funded by X (formerly Google X), the Moonshot Factory. The funding source had no influence on study conduction or result evaluation. MB is supported by grants from Harold Shames and the National Institutes of Health: NIH-NIDA UG3DA048502, NIH-NIGMS T34 GM137858, NIH-NINDS R01 NS112996, and NIH-NINDS R01 NS101362. NG and MB are further supported by NIH-G-RISE T32GM136499.

Contact

Please feel free to reach out with any questions or suggestions.

• Email: ngebodh01@citymail.cuny.edu
• Twitter: @ngebodh