Cortical-SSM: a deep state space model for motor imagery decoding from EEG signals

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: Not reported in abstract — stated only as 'outperformed baseline methods'; specific accuracy improvements or effect sizes are not disclosed

• Primary Outcome: Classification accuracy of motor imagery EEG signals in a subject-independent paradigm across two large-scale public benchmark datasets

• Nnt Or Sensitivity: Not applicable in conventional clinical sense; no sensitivity, specificity, AUC, or classification accuracy figures are provided in the abstract. For BCI benchmarking, relevant metrics would include mean classification accuracy (%), kappa coefficient, and information transfer rate (bits/min) — none of which are reported

• Confidence Interval: Not reported

Clinical Application

Computational feasibility depends on hardware requirements for the deep SSM architecture, which are not specified. Clinical deployment would require real-time decoding capability, robust artefact handling in uncontrolled environments, minimal user training, and regulatory approval — none of which have been demonstrated. EEG hardware costs and the need for trained operators remain barriers in routine clinical settings. Brain-computer interface technology for communication and rehabilitation is not currently listed on the Australian Register of Therapeutic Goods (ARTG) as a standard therapeutic device, and no PBS-listed BCI systems exist. The RACGP and relevant specialist colleges (AAN equivalent bodies, rehabilitation medicine) have not issued specific guidelines on BCI use. Australian NDIS funding may theoretically support assistive communication devices, but regulatory pathways via the TGA for software-as-a-medical-device (SaMD) classification would apply to any clinical deployment of Cortical-SSM. The study's findings are relevant to Australian researchers in neural engineering and rehabilitation medicine but have no immediate clinical practice implications. Australian clinical trials in this space would need HREC approval and TGA SaMD framework compliance. Theoretically applicable to individuals with motor impairments requiring brain-computer interface support, including patients with amyotrophic lateral sclerosis (ALS), high-level spinal cord injury, locked-in syndrome, or severe stroke with preserved cognitive function. However, clinical applicability is currently limited to research settings given the absence of prospective patient-level validation.

Abstract

Objective.Classification of electroencephalogram (EEG) signals obtained during motor imagery (MI) has substantial application potential, including for communication assistance and rehabilitation support for patients with motor impairments. These signals remain inherently susceptible to physiological artifacts (e.g. eye blinking, swallowing), which pose persistent challenges. Although Transformer-based approaches for classifying EEG signals have been widely adopted, they often struggle to capture fine-grained dependencies within them.Approach.To overcome these limitations, we propose Cortical-SSM, a novel architecture that extends deep state space models to capture integrated dependencies of EEG signals across temporal, spatial, and frequency domains. We validated our method across two large-scale public MI EEG datasets containing more than 50 subjects.Main results.Our method outperformed baseline methods on the two benchmarks. Furthermore, visual explanations derived from our model indicate that it effectively captures neurophysiologically relevant regions of EEG signals.Significance.These results indicate that Cortical-SSM provides a robust and interpretable alternative to attention-based architectures for MI EEG decoding. By enabling physiologically grounded feature learning, our method advances the reliability of subject-independent EEG classification and supports the development of practical and clinically deployable brain-computer interface systems.

References

1. 1.Suzuki, S., Nagashima, S., & Sugiura, K. (2026). Cortical-SSM: a deep state space model for motor imagery decoding from EEG signals. Journal of Neural Engineering. https://doi.org/10.1088/1741-2552/ae89e8