Evaluación y mejora de la calidad interna y calidad de uso de un computer-aided diagnostic y detection system para el reconocimiento automático de patologías de la mama en imágenes mamográficas

Abstract

The following work aims to guarantee the quality in each of the stages of the development cycle of a Computer-Aided Diagnostic and Detection System for the Automatic Recognition of Pathologies of the breast in Mammographic Images, as one of the products of the Research Project REDU: PREDU-2016-013. For this purpose, the ISO 25000 standard, the usability heuristic rules and the ISO 19505-2 standard were combined. The construction of the system was framed in the 4 phases proposed in the methodology of the agile unified process (AUP) and in each phase an incremental prototype was built. To achieve this, the activities specified in said methodology were carried out: 1) Survey of software and stakeholder requirements; 2) Diagramming phase; 3) Interface design; 4) Development phase and 5) Product evaluation. Then, a user satisfaction questionnaire was designed focusing on two types of users: normal and experts. The results of the evaluation in this first version show that the product evaluated has a satisfactory quality level obtaining 9.86/10 in the evaluation of quality in use and 8.45/10 in the internal quality evaluation.