Stochastic Resonance as a Diagnostic Tool for Neural Pathologies

Schilling et al. propose stochastic resonance (SR) as a compensatory mechanism in tinnitus, where neural noise amplifies weak signals. This research explores SR as a diagnostic biomarker for other conditions (e.g., Parkinson’s tremors or epilepsy). By modeling neural pathways as stochastic processes with controllable noise parameters, we could identify pathological states where SR fails or overcompensates. For example, in EEG data, deviations from expected SR patterns might signal early neurodegeneration. Unlike Schilling et al.'s focus on auditory perception, we extend SR to multisensory integration using geometric reversibility (O'Byrne & Cates, 2025) to quantify entropy production. This diverges from traditional ML-based diagnostics (Khan et al., 2023) by grounding analysis in stochastic thermodynamics.

References:

1. Integrating Machine Learning and Stochastic Pattern Analysis for the Forecasting of Time-Series Data. A. B. F. Khan, K. Kamalakannan, nisar. ahmed (2023). SN Computer Science.
2. Geometric theory of (extended) time-reversal symmetries in stochastic processes: II. Field theory. J. O'Byrne, M. E. Cates (2025). Journal of Statistical Mechanics: Theory and Experiment.
3. Predictive coding and stochastic resonance as fundamental principles of auditory phantom perception. A. Schilling, W. Sedley, Richard C. Gerum, C. Metzner, Konstantin Tziridis, Andreas K. Maier, H. Schulze, F. Zeng, K. Friston, P. Krauss (2023). Brain : a journal of neurology.