*3.4. Coincident Bands Evaluation Methodology*

When the optimization algorithms employed in *PF2* and *PF3* selected the bands, different bands were identified for each test image. Thus, the six HS test images were evaluated employing the same selected bands, i.e., using the coincident and non-coincident selected bands in each test image obtained in the leave-one-patient-out cross-validation for the SVM training and classification. The procedure was as follows:


This process was repeated until reaching the possible six coincidences. However, in our case, a maximum of three levels of coincidence were obtained.
