Objective motion-based performance assessment in STA-MCA microanastomosis: a postprocedural debriefing platform

The aim of this study was to develop and demonstrate a postprocedural motion analysis platform capable of objectively quantifying intraoperative surgeon micromovements to facilitate evidence-based self-debriefing in suturing superficial temporal artery-middle cerebral artery microanastomosis.

The authors analyzed microscope-recorded microanastomoses performed by a cerebrovascular neurosurgeon. Seven suture-placement clips were extracted from operative videos. A computer vision tracking pipeline was implemented in Python using OpenCV to track the needle holder position across the video clips. Four motion metrics were computed for each video clip: absolute error, velocity, acceleration, and jerk. A unified reference distribution was constructed from all clips, and 95th percentile thresholds were defined for each metric. The metrics were displayed as dynamic synchronized plots alongside the corresponding video. Spectral decomposition of the position signal was performed to generate time-frequency representations, and average power in the 6- to 8-Hz band was computed as a quantitative tremor measure. To demonstrate reliable longitudinal performance tracking, the authors applied their platform to 8 recorded procedures between 2023 and 2026 and plotted the mean jerk per suture over time.

Reliable tracking was achieved across 92.09% of analyzed footage. Two-dimensional trajectories were reconstructed for a total of 13,183 analyzed frames encompassing 219.82 seconds of suture-placement video (7 clips). Jerk analysis revealed multiple segments exceeding the 95th percentile threshold, flagging moments for surgical review. Time-frequency analysis revealed transient broadband power bursts and 2 persistent spectral bands (6-8 Hz and < 4 Hz). Tremor intensity analysis in the 6- to 8-Hz range revealed temporal tremor fluctuations, a finding undetectable by other metrics. Longitudinal performance tracking was feasible and demonstrated a stable jerk profile, consistent with expert-level performance.

A motion-based postprocedural debriefing platform for the performance of microanastomosis is technically feasible and capable of generating objective, quantitative performance metrics from standard operative video. Unlike prior work focused on expert-novice differentiation, this platform was designed to support individualized self-assessment and reflective practice for surgeons at any experience level. Future work should validate the platform prospectively and correlate motion-based metrics with clinical outcomes.