Gaze-based Technology as a Tool for Surgical Skills Assessment and Training in Urology

Abstract

To assess the sensitivity of gaze-based metrics in detecting cognitive demands imposed by surgical procedures. We analyzed urologists' gaze entropy and velocity while performing 2 standardized high-fidelity simulated stone procedures with different levels of complexity. Using a wearable eye tracker device (mounted onto an eyeglass frame), we measured gaze entropy and velocity in 15 urologists, members of the Andalusian health-care system, while they performed an extraction of a stone in the bladder (low complexity) and an extraction of a stone in the lumbar ureter (high complexity). We also collected performance and subjective data. Gaze entropy and velocity were significantly higher when surgeons performed the most complex surgical procedure: the visual exploration pattern became less stereotyped (ie, more random) and faster. Surgeons' performance and perceived task complexity differed accordingly, confirming the gaze-based results. Gaze-based metrics might have great potential as objective and nonintrusive indices to assess surgeons' cognitive (over)load, potentially being a complementary assessment tool to quantify the learning curve for surgical procedures.