TY - GEN
T1 - Characterization of Data-driven ECoG Filters for Speech and Hand Motor Decoding in an ALS Patient
AU - Wallner, Josephine
AU - Angrick, Miguel
AU - Candrea, Daniel N.
AU - Luo, Shiyu
AU - Crone, Nathan E.
AU - Krusienski, Dean J.
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Brain-computer Interface (BCI) research has recently reached the stage where it has become feasible and prudent to perform long-term implantations and studies of efficacy in patients. One such clinical trial implanted a patient suffering from progressive amyotrophic lateral sclerosis with electrocorticographic arrays over the sensorimotor cortex for approximately 2.5 years, during which the patient performed a variety of speech and hand motor tasks in a controlled laboratory setting. The present analysis utilizes this longitudinal data to train interpretable convolutional neural networks using EEGNet to compare the decoding performance and resulting data-driven filters across hand motor, speech, and combined representations. These results aim to elucidate the common neural activations and substrates across speech and grasp tasks, and provide new insights into the spatial and spectral characteristics of the relevant neural features for the decoding tasks.
AB - Brain-computer Interface (BCI) research has recently reached the stage where it has become feasible and prudent to perform long-term implantations and studies of efficacy in patients. One such clinical trial implanted a patient suffering from progressive amyotrophic lateral sclerosis with electrocorticographic arrays over the sensorimotor cortex for approximately 2.5 years, during which the patient performed a variety of speech and hand motor tasks in a controlled laboratory setting. The present analysis utilizes this longitudinal data to train interpretable convolutional neural networks using EEGNet to compare the decoding performance and resulting data-driven filters across hand motor, speech, and combined representations. These results aim to elucidate the common neural activations and substrates across speech and grasp tasks, and provide new insights into the spatial and spectral characteristics of the relevant neural features for the decoding tasks.
KW - amyotrophic lateral sclerosis (ALS)
KW - brain-computer interface (BCI)
KW - convolutional neural networks (CNNs)
KW - electrocorticography (ECoG)
UR - https://www.scopus.com/pages/publications/105045265598
UR - https://www.scopus.com/pages/publications/105045265598#tab=citedBy
U2 - 10.1109/NER61569.2025.11589003
DO - 10.1109/NER61569.2025.11589003
M3 - Conference contribution
AN - SCOPUS:105045265598
T3 - International IEEE/EMBS Conference on Neural Engineering, NER
SP - 471
EP - 475
BT - 2025 12th International IEEE/EMBS Conference on Neural Engineering, NER 2025 - Proceedings
PB - IEEE Computer Society
T2 - 12th Annual International IEEE/EMBS Conference on Neural Engineering, NER 2025
Y2 - 11 November 2025 through 14 November 2025
ER -