Visual working memory influences the performance in virtual image–guided surgical intervention

This study addresses for the first time the relationship between working memory and performance measures in image-guided instrument navigation with Minimally Invasive Surgical Trainer-Virtual Reality (MIST-VR) and GI Mentor II (a simulator for gastroendoscopy). In light of recent research on simulator training, it is now prime time to ask why in a search for mechanisms rather than show repeatedly that conventional curriculum for simulation training has effect.

The participants in this study were 28 Swedish medical students taking their course in basic surgery. Visual and verbal working memory span scores were assessed by a validated computer program (RoboMemo) and correlated with visual–spatial ability (MRT-A test), total flow experience (flow scale), mental strain (Borg scale), and performance scores in manipulation and diathermy (MD) using Procedicus MIST-VR and GI Mentor 11 (exercises 1 and 3).

Significant Pearson’s r correlations were obtained between visual working memory span scores for visual data link (a RoboMemo exercise) and movement economy (r = −0.417; p < 0.05), total time (r = −0.495; p < 0.01), and total score (r = −0.390; p < 0.05) using MIST-MD, as well as total time (r = −0.493; p < 0.05) and efficiency of screening (r = 0.469; p < 0.05) using GI Mentor 11 (exercise 1). Correlations also were found between visual working memory span scores in rotating data link (another RoboMemo exercise) and both total time (r = −0.467; p < 0.05) and efficiency of screening (r = −0.436; p < 0.05) using GI Mentor 11 (exercise 3). Significant Pearson’s r correlations also were found between visual–spatial ability scores and several performance scores for the MIST and GI Mentor II exercises.

Findings for the first time demonstrate that visual working memory for surgical novices may be important for performance in virtual simulator training with two well-known and validated simulators.