This study aims to predict the values of ergonomic (body position, angular rate), kinematic (steps count), and physiological (ECG signal, beats per minute, blood pressure, and body temperature) parameters by using a wearable electrocardiogram (ECG) sensor worn by surgeons during the performance of minimally invasive surgical (MIS) procedures (conventional laparoscopy and robotic-assisted surgery -RAS-). For this purpose, data related to the surgeon’s ECG sensor parameters were collected during forty-four MIS sessions conducted by eighteen surgeons with different levels of experience. Once the dataset was generated, two preprocessing techniques were applied: scaling and normalization. These two datasets were subsequently divided into two subsets: one containing 80% of the data for training and cross-validation, and the other containing 20% of data for testing. Several Artificial Intelligence (AI) techniques were applied to the training dataset to develop the predictive models. Finally, these models were validated on cross-validation and test datasets. PCA results showed that the physiological parameters (ECG signal, beats per minute, blood pressure and body temperature) were the most representative parameters of the ECG sensor. The predictive analysis results showed that XGBoost achieved the best results for the training dataset, while Multiple Linear Regression demonstrated the best performance for the cross-validation and test datasets combined with the scaled preprocessing technique, achieving the highest R \(^2\) coefficient and the lowest error for each parameter analyzed. The linear models were successfully validated on both cross-validation and test datasets, underscoring the potential for prediction of factors contributing to surgeon’s health improvement during MIS procedures.