Figure 4.

Block diagram of the proposed method including pseudocode: at first data are separated into ten interpatient folds, then 10-fold CV is being set, and the training process is repeated for all groups of folds. In each loop the model is trained and validated for nine folds, then tested on the 10th fold. Next, the micro-average values of the sensitivity, specificity, score, and accuracy are calculated. As soon as training is finished for all folds, average sensitivity, specificity, score, and accuracy are extracted.