Dynamic evaluation of instrument trajectories that delivers real-time guidance and quantitative metrics on surgical motion quality, helping surgeons achieve consistent, high-quality suturing outcomes.
Three integrated stages combine to convert raw endoscopic video into actionable suturing guidance.
High-definition endoscopic video is streamed into the pipeline. Deep-learning models segment surgical instruments, tissue, and suture material on every frame.
Six-degree-of-freedom instrument poses are reconstructed and compared against reference trajectories derived from expert surgical recordings to compute deviation scores.
Scores and guidance overlays are rendered on the OR display with sub-50 ms latency, giving the surgeon immediate, non-distracting cues during the procedure.