JMIR Publications

ABSTRACT

Background:

Robotic-assisted surgery (RAS) has grown rapidly in recent decades and several RAS procedures have become the standard. However, the physical and mental demands of minimally invasive surgical (MIS) techniques can lead to ergonomic shortcomings for surgeons. Advances in wearable technology and artificial intelligence (AI) favor the development of innovative solutions to analyze and improve ergonomic conditions during surgical practice.

Objective:

The main objective is the development and validation of a predictive model of localized muscle fatigue from EMG data during conventional laparoscopic surgery (LAP) and RAS.

Methods:

Four different tasks were performed on LAP and RAS: dissection, labyrinth, peg transfer and suturing. A wireless EMG sensor system was used to record muscle activity. JASA graphs were used to evaluate the localized muscle fatigue. A dataset was generated for each task as a function of surgeons’ expertise level and surgical type. Each dataset was scaled as preprocessing and divided into two datasets: 80% for training and 20% for test. Multiple Linear Regression (MLR) and Multilayer Perceptron (MLP) were applied as predictive techniques and validated on all test datasets. R2 coefficient and root mean square error (RMSE) were used to measure the accuracy of the models.

Results:

RAS showed less muscle fatigue for novice surgeons compared to conventional laparoscopic practice, although it was higher for expert surgeons. The predictive model achieved satisfactory R2 and RMSE coefficients for all parameters extracted from the EMG signal, predicting with high accuracy localized muscle fatigue values. The MLR predictive model demonstrated superior performance relative to the MLP model.

Conclusions:

Wearable technology and artificial intelligence techniques have been successfully applied for the development and validation of a novel predictive model based on MLR and MLP to predict localized muscle fatigue in MIS.