What question did this study set out to answer?

This research aims to evaluate the effectiveness of geometric feature extraction for classifying motor imagery in EEG recordings.

synapse

⌘+K

synapse

⌘+K

March 26, 2026Open Access

Geometric Features for Brain-Computer Interface Classification: Benchmark Results on PhysioNet Motor Imagery

Key Points

This research aims to evaluate the effectiveness of geometric feature extraction for classifying motor imagery in EEG recordings.
Utilized EEG data from PhysioNet with 109 subjects and 64 channels.
Applied proprietary geometric feature extraction on spatial covariance matrices.
Compared results against pyRiemann baseline using Wilcoxon signed-rank test.
Achieved a +0.78pp improvement over pyRiemann baseline (p=0.0048).
Demonstrated strongest advantage with low channel counts (n=3-8).
Showed +8.5pp improvement in cross-subject transfer over standard MDM.

Abstract

OmniSciences Technical Report TR-2026-003 We present benchmark results for proprietary geometric feature extraction applied to EEG-based brain-computer interface (BCI) motor imagery classification. The method operates on spatial covariance matrices on the SPD manifold. Key results: PhysioNet (109 subjects, 64 channels): +0.78pp over pyRiemann baseline (p=0.0048, Wilcoxon signed-rank test) Strongest advantage at low channel counts (n=3-8), relevant for consumer BCI Cross-subject transfer: +8.5pp over standard MDM Patent-pending (U.S. Provisional Application No. 64/004,047).

Geometric Features for Brain-Computer Interface Classification: Benchmark Results on PhysioNet Motor Imagery

Key Points

Abstract

Cite This Study