eess.SPNov 15, 2026

Cross-Subject Priors for Maximum a Posteriori Estimation of Latent Structure Influence Models in Motor-Imagery EEG With Few Trials

Authors: Sagnik De, Anil Kumar Gupta

Organizations: Institute of Radio Physics & Electronics, University of Calcutta, Kolkata 700009, India · HPC Infrastructure and EcoSystem Group, Centre for Development of Advanced Computing, Pune 411008, India

Abstract

Coupled hidden Markov models such as the latent structure influence model (LSIM) describe how the sensorimotor rhythms of the two hemispheres evolve and interact during motor imagery, but subject-specific maximum likelihood (ML) estimates are unstable when few trials are available. This paper derives a maximum a posteriori (MAP) estimator for the LSIM whose conjugate prior equals the expected statistics of λλ pseudo-trials from a population model fitted to other subjects. The M-step remains in closed form, expectation maximization keeps its monotone ascent, and ML (λ=0λ=0) and the population model (λ→∞λ\to\infty) are limiting cases. The prior strength was selected on 35 subjects of the PhysioNet EEG Motor Movement/Imagery Dataset and applied unchanged to 70 held-out subjects and to BCI Competition IV-2a. With λ=1λ=1 the held-out log-likelihood exceeded that of ML in 61 of 70 subjects (Holm-adjusted p=9.1×10−8p=9.1\times10^{-8}) and in all 70 subjects with five training trials per class. A prior learned only from the PhysioNet data improved the held-out likelihood of all nine BCI IV-2a subjects at five trials per class and reduced the error of the joint transition matrix by 61%. Decoding accuracy did not improve (p=0.17p=0.17). Cross-subject priors therefore stabilize few-trial dynamic models of motor-imagery EEG, but they do not improve decoding, which in these data relies mainly on structure shared across subjects.

Figures & tables

Explore similar work

Jul 25, 2026

Bayesian Complete-Pooling in Cross-Subject Classification for Motor Imagery Electroencephalogram

Brain-computer interfaces (BCIs) have long sought calibration-free operation, yet classifiers are typically benchmarked on discrimination alone. Discrimination is blind to calibration, a meaningful gap given that electroencephalogram (EEG) signals are nonstationary and point-estimate classifiers can become overconfident under distribution shift. We conducted a large-scale study contrasting Bayesian complete-pooling models against frequentist baselines for cross-subject, left-hand versus right-hand motor imagery EEG classification across 20 datasets. Each of six frequentist pipelines was paired with an analogous Bayesian pipeline sharing identical, deterministic feature extraction, so that only the classifier was Bayesian. Bayesian classifiers were fit via Markov chain Monte Carlo. Our primary metric was the Brier score, decomposed into reliability and resolution, alongside the area under the ROC curve for discrimination and Shannon entropy for sharpness. For each metric we fit a random-effects meta-analysis with Knapp-Hartung adjustment, verified by leave-one-out influence analysis. Bayesian complete-pooling produced statistically significant improvements in reliability and increases in predictive uncertainty (lower sharpness), with 95% confidence intervals of [-0.0029, -0.0001] for reliability and [0.0022, 0.0220] bits for entropy. Still, neither effect held statistical significance when datasets with flagged posterior sampling were excluded. Brier score, resolution, and discrimination showed no significant differences. Between-study heterogeneity was low across all metrics. Bayesian pipelines consumed roughly thirteen times more energy than their frequentist counterparts, a cost that remains modest relative to common household appliances. This modest cost makes partial-pooling across subjects and sessions, the structure complete-pooling omits, a feasible next step.
Jul 24, 2026

Subject-Level Heterogeneity in EEG Motor Imagery Decoding: A Large-Scale Benchmark and Portfolio-Based Reduction of the Search Space

Robust EEG motor imagery decoding remains limited by strong inter-individual variability, making it difficult to identify pipelines that generalize across users. We present a large-scale, standardized within-session benchmark of decoding pipelines across three public datasets: Cho2017 (52 subjects), PhysionetMI (109 subjects), and Zhou2016 (4 subjects). Using a common MOABB LeftRightImagery setting, two frequency bands (8-15 Hz and 8-30 Hz), and a broad combination of feature extraction, preprocessing, and classification steps, we analyzed 216,714 raw evaluation rows, which after structured aggregation yielded 44,928, 109,000, and 4,192 subject-level observations respectively. Covariance tangent-space projection (cov-tgsp) and Common Spatial Patterns (CSP) consistently defined the strongest methodological families, though their relative ordering was dataset-dependent. On Cho2017, the best family-level mean accuracy came from cov-tgsp in 8-30 Hz (0.712 +/- 0.140), whereas Zhou2016 favored CSP (0.832 +/- 0.121 in 8-15 Hz). These aggregate rankings concealed substantial subject-level heterogeneity: 42 distinct winning pipelines across 52 Cho2017 subjects, and 93 across 109 PhysionetMI subjects. We then used the benchmark as an empirical performance landscape for building compact portfolios of pipelines of size K. Several construction procedures were compared, including a ranking-based Top-K Mean heuristic and search-based strategies. Results were broadly consistent, with Top-K Mean giving the best trade-off. A single best global pipeline already retained 94.2% of the oracle in Cho2017 and 81.8% in PhysionetMI; at K = 12, oracle retention rose to 96.5% and 90.0%. The landscape is therefore subject-dependent, and this heterogeneity can be exploited through compact portfolios that make personalization more feasible.
Jul 3, 2026

Stacked LoRA for Subject-Adaptive EEG Foundation Models in Motor Imagery Decoding

Electroencephalography (EEG) decoding for brain-computer interfaces (BCIs) faces a major challenge: substantial inter-subject variability limits effective cross-subject generalization. Consequently, practical systems still rely largely on subject-specific models trained from scratch and requiring individual recalibration. EEG foundation models have recently emerged as a promising alternative; however, even large pretrained models cannot simply be used as fixed feature extractors and still require additional adaptation before they can be reliably applied to downstream tasks. In this work, we address this challenge through targeted adaptation strategies. Building on recent EEG foundation models such as REVE, LaBraM, and LUNA, we examine the impact of different low-rank adaptation strategies on motor imagery classification. We propose a framework that structurally decouples subject-invariant knowledge from subject-specific neural signatures: the low-rank update at each adapted layer is split into a Global adapter, trained jointly across all subjects, and Subject-Specific adapters, each absorbing individual variability. To assess the contribution of each path, we compare three adaptation strategies: (i) subject-specific LoRA (ii) global LoRA and (iii) stacked LoRA, combining both Global and Subject Specific adapters. Experiments on BCI Competition IV-2a, PhysioNet Motor Imagery, and the clinical Zuo2025 benchmark show that Stacked LoRA effectively mitigates inter-subject variability, achieving the best accuracy in the large majority of backbone and dataset combinations. Our analysis further reveals that the optimal balance between the global and subject-specific paths depends on the target population: a shared adapter is sufficient for large, diverse cohorts, whereas subject-specific adaptation is decisive under the high inter-session variability of clinical recordings.