Intraoperative neurophysiological monitoring (IONM) uses electrical stimulation to track neural integrity during high-risk surgeries. A key challenge is that stimulation to assess integrity of motor pathways induces patient movement, disrupting the surgical field. This causes irregular application of stimulation and therefore surveillance gaps. This emphasizes the need for automated, adaptive strategies. We retrospectively analyzed stimulation intervals in 130 procedures across two centers, comparing transcranial (TES) and direct cortical stimulation (DCS). Stimulation patterns varied markedly within and across surgeries and centers. We propose a closed-loop workflow combining accelerometer-based movement tracking, and machine learning (ML) based signal forecasting. Such a system could adapt stimulation to surgical context and use ML to reduce variability, supporting consistent motor surveillance in minimally invasive neurosurgery.
