Cognitive Activity and Surgical Capability: Impact of Cognitive Load on Simulated Cholecystectomy Performance
Nicholas E. Anton, MS, Wendy S. Li, MD, Nathan G. Behrens, MD, Brielle M. Warnock, MD, Noosha D. Deravi, MD, Dimitrios Stefanidis, MD, PhD
Indiana University School of Medicine, Indianapolis Indiana
Background
•Cognitive load (CL) refers to the information demands on working memory.1
•When excessive, CL can degrade surgical performance.2
•Self-report CL measures have important limitations in objectively capturing impact of CL on surgical performance.
•Physiological sensors like electroencephalogram (EEG) or heart rate monitors could objectively capture indices of surgeons’ CL.3
•Study Objective: Determine whether objective CL sensing modalities are associated with surgical performance.
Methods
•Surgical novices and experienced surgeons participated.
•Performed cholecystectomy procedures on explanted porcine livers laparoscopically and robotically.
•A heart rate monitor, mobile eye tracker, and 14-lead saline-based EEG measured CL during performance.
•Spare attentional capacity (i.e., available working memory) was measured with a secondary visuospatial task superimposed on the viewing monitor. Participants used a foot pedal when they detected a particular pattern.
•Performance videos were evaluated by two blinded raters using the Global Operative Assessment of Laparoscopic Skills (GOALS and the Global Evaluate Assessment of Robotic Skills (GEARS), respectively.
•Performance differences between groups and surgical modalities and correlations were analyzed non-parametrically.
Results
•Thirteen participants (62% Female, 46% experienced, 100% right-handed).
•Weighted Kappa revealed high inter-rater reliability (ranging from 0.78 to 0.96 across GOALS/GEARS domains).
•Experienced surgeons had significantly better depth perception and efficiency than novices laparoscopically and had significantly better performance robotically in all GEARS domains (p < 0.05).
•Experienced surgeons had significantly better secondary task performance than novices on both platforms (p < 0.05).
•There were no significant differences in performance or CL between laparoscopic and robotic cholecystectomies.
•Heart rate and eye tracking metrics were not correlated with performance.
•EEG-measured cortical activation (reflective of CL) in the prefrontal, premotor, motor, occipital, and parietal cortices negatively correlated with surgical performance (p < 0.05).
•The number of correct identifications in the secondary task positively correlated with performance (R = 0.58, p = 0.003) while the number of missed identifications negatively correlated (R = -0.5, p = 0.01).
Conclusions
•Across simulated laparoscopic and robotic cholecystectomies, greater theta band cortical activity, which indicates heightened cognitive load, was associated with poorer surgical performance.
•More spare attentional capacity, indicating less CL, was associated with better surgical performance.
•EEG and secondary task measures show promise in objectively measuring CL during surgical performance.