Figure 2.

Training and validation scheme for machine learning methods. The database is split, and 70% of the data are used for training and validation of the method and 30% for testing. The model is trained with a training set and scored on the test set (metrics), and then the process is repeated k-times. After this training, pattern discrimination is then tested in a different subset of patients (test set, 30% of the database). The whole process is then repeated until the learning stabilizes and stops improving. The results presented in this study are obtained from the evaluation of this subset.