| Summary: | The current methods of assessment of surgical performance for robot-assisted surgery are subjective. In this paper, we propose a cognitive-based method for objective evaluation of performance. Changes in brain functional networks were extracted and their relationship with performance level was investigated. We used electroencephalogram data recorded from a mentor surgeon's brain while supervising and performing surgical tasks of varying complexity [urethrovesical anastomosis (UVA) and lymph-node dissection (LND)]. Multilayer community detection techniques were used to extract functional network communities at frequency bands of θ, α, lower β, upper β, and γ. Results showed different detected communities while supervising and performing LND (more complex). However, for UVA (less complex), the majority of functional communities were similar. This is likely because, in less complicated tasks, the trainee's performance more closely matched the mentor's expectation. Entropy and power distribution through frequency bands showed minimum thermodynamic stability during α and the maximum during y. The relaxation time for channels with high entropy level was also extracted as a brain functional metric at thermodynamic stability state. These metrics may be used to quantify changes of brain functional network as performance improves.
|
|---|
