Wireless body sensor for objective assessment of surgical performance on a standardised FLS task

Advances in Body Sensor Networks have prompted increasing numbers of low cost, miniaturised sensors being used in many different applications with one being the capture of hand movement data for surgical skills assessment. Despite these advances, existing assessment techniques are still predominantly subjective and resource demanding. Combining surgical training with a reliable objective assessment technique would ensure that trainees are correctly evaluated and credentialed as they progress through their training hence, ensuring competence and reducing critical medical errors. This paper proposes the use of wearable, wireless inertial sensors for capturing motion data and enabling objective assessment of trainee surgeons' performance in carrying out one of the FLS (Fundamentals of Laparoscopic surgery) tasks; the peg transfer. A novel approach has been developed for the segmenting of specific peg movements enabling performance to be measured entirely objectively. The features derived from the whole task as well as features for each of the segmented movements were analysed through unsupervised machine learning algorithms to look for useful measures of performance as well as patterns to identify differences between expert and trainee performance. Encouraging results in the peg transfer task, where a successful classification of expertise was obtained for all participants against gold standard assessment, prompt further investigation into the development of advanced performance metrics for a wider range of surgical training tasks.