Smartwatch-based wrist tremor assessment in neurosurgical simulator training

While surgical skill is acquired during training, objectively evaluating training's impact is challenging. However, new possibilities in measuring wrist tremor during surgical procedures have emerged due to novel wearable sensors. In this study, the effect of simulator-based neurosurgical resident training on wrist tremor was assessed using a smartwatch. A study involving nine neurosurgical residents performing simulator-based training on unruptured intracranial aneurysm clipping procedures was conducted. Residents performed four training sessions on two days (180 clippings in total). Three cohorts were formed receiving in between the two days either: no training, video-based training, or video-based training with augmented reality. Wrist acceleration was measured using a smartwatch. Wrist tremor was calculated using the acceleration data in time and frequency domains (peak power, total power, and peak frequency). Video-based training with augmented reality showed a statistically significant reduction in total power as well as in peak power, with reductions of 56% (p = 0.0186) and 69% (p = 0.0021), between day 0 and day 14, respectively. Wrist tremor assessment using a smartwatch confirmed the positive impact of video-based and augmented reality training on neurosurgical residents' confidence in surgical training procedures by reducing wrist tremor, making it a valuable tool for skill assessment.