OBJECTIVE: Methods of evaluating surgical performance are mainly subjective. This study introduces a method of evaluating surgical performance using a quantitative analysis of tool tip kinematics. METHODS: One experienced surgeon performed eight rat microvascular anastomoses over a 2-day interval. An optoelectronic motion analysis system acquired tool tip trajectories at frequencies of 30 Hz. On the basis of a hierarchical decomposition, the procedure was segmented into specific surgical subtasks (free space movement, needle placement and knot throws) from which characteristic measures of performance (tool tip trajectory, excursion and velocity) were evaluated. Comparisons of performance measures across each procedure were indexed (D scale) using the Kolmogorov-Smirnov statistic. RESULTS: Despite the marker occlusions, tool tip data were obtained 92 +/- 7% (mean +/- SD) of the time during manipulation tasks. Missing data segments were interpolated across gaps of less than 10 sample points with errors less than 0.4 mm. The anastomoses were completed in 27 +/- 4 min (range 20.5-31.4 min) with 100% patency. Tool tip trajectories and excursions were comparable for each hand, while right and left hand differences were found for velocity. Performance measures comparisons across each procedure established the benchmark for an experienced surgeon. The D-scale range was between 0 and 0.5. CONCLUSION: The study establishes a reproducible method of quantitating surgical performance. This may enhance assessment of surgical trainees at various levels of training.
OBJECTIVE: Methods of evaluating surgical performance are mainly subjective. This study introduces a method of evaluating surgical performance using a quantitative analysis of tool tip kinematics. METHODS: One experienced surgeon performed eight rat microvascular anastomoses over a 2-day interval. An optoelectronic motion analysis system acquired tool tip trajectories at frequencies of 30 Hz. On the basis of a hierarchical decomposition, the procedure was segmented into specific surgical subtasks (free space movement, needle placement and knot throws) from which characteristic measures of performance (tool tip trajectory, excursion and velocity) were evaluated. Comparisons of performance measures across each procedure were indexed (D scale) using the Kolmogorov-Smirnov statistic. RESULTS: Despite the marker occlusions, tool tip data were obtained 92 +/- 7% (mean +/- SD) of the time during manipulation tasks. Missing data segments were interpolated across gaps of less than 10 sample points with errors less than 0.4 mm. The anastomoses were completed in 27 +/- 4 min (range 20.5-31.4 min) with 100% patency. Tool tip trajectories and excursions were comparable for each hand, while right and left hand differences were found for velocity. Performance measures comparisons across each procedure established the benchmark for an experienced surgeon. The D-scale range was between 0 and 0.5. CONCLUSION: The study establishes a reproducible method of quantitating surgical performance. This may enhance assessment of surgical trainees at various levels of training.
Authors: Joshua D Knopf; Rahul Kumar; Michael Barats; Paul Klimo; Frederick A Boop; L Madison Michael; Jonathan E Martin; Markus Bookland; David S Hersh Journal: World Neurosurg Date: 2020-09-02 Impact factor: 2.104