Introduction

We show here an explanation of the evaluation procedures done in the Thesis. Firstly, we explain how we evaluated the classifications tasks performed in the work. Note that we did two different works related with classification: classification of the breasts according to their internal density in the four classes used in the BIRADS standard, and the classification of the RoIs when evaluating the false positive reduction method. In the former, we evaluated the classification using confusion matrices, while in the latter we used ROC analysis. Both techniques are analyzed in Section .

On the other hand, we have also results obtained from the evaluation of the detection algorithm. In contrast with the above results, these are obtained from automatically segmented images, and the results are obtained by comparing these images with the corresponding manually annotated image. In this task, ROC and FROC analysis are performed, which are described in Section .