Aron Jeremiah: Unravelling Propagation: Exploring Spinal Cord Injury Dynamics through Machine Learning

<p dir="ltr">Spinal cord injury (SCI) disrupts signal transmission within the nervous system, motivating the development of implantable interfaces for both stimulation and sensing. This work investigates the feasibility of classifying injury status using electrophysiological signals acquired from freely moving rats with custom intraspinal neural implants. Time-series features were systematically extracted using tsfresh and evaluated for statistical relevance to injury state. Significant features were used to train a supervised LightGBM classifier, with performance assessed via leave-one-subject-out cross-validation. The model achieved an average Area Under the Curve (AUC) of 0.85, with a peak of 0.96, demonstrating reliable discrimination between healthy and injured conditions. Statistical analysis identified approximately 310 features (p < 0.001) associated with SCI, indicating measurable alterations in neural signal characteristics. These results demonstrate a data-driven framework for electrophysiological state classification and lay the groundwork for advanced diagnostic and prognostic tools in spinal injury assessment.</p>