DDC Tool: Aplicación Móvil para la Detección Automatizada de Displasia de Cadera en Niños

Abstract

Abstract: His work presents the development of a mobile application based on artificial intelligence for the automated detection of developmental dysplasia of the hip in children through the analysis of pelvic radiographs. The system integrates deep learning models for the segmentation of relevant anatomical structures, as well as an automated morphogeometric analysis for the estimation of acetabular angles, and a classification module that supports the assessment of disease severity. The system performance was evaluated by comparing the automated measurements with those performed by a medical specialist, using statistical metrics and concordance analysis. The obtained results show low average errors and high accuracy within clinically acceptable thresholds, as well as good agreement between both methods, particularly after the exclusion of outliers. These findings demonstrate that the proposed application can serve as a diagnostic support tool, contributing to the reduction of subjectivity in radiographic interpretation and improving the early detection of developmental dysplasia of the hip, especially in resource-limited settings.