Deep Learning for Healthcare: A Web-Microservices System Ready for Chest Pathology Detection

The automation of medical diagnosis has accelerated thanks to the integration of artificial intelligence (AI), particularly in interpreting pathologies in chest X-rays. This study presents a web microservices system that uses a deep learning model to classify thoracic pathologies. The system improves clinical decision-making by providing visual aids, including heat maps for model explainability and a comprehensive set of medical image manipulation tools. The back-end, developed using a microservices architecture, ensures robust data management, secure user authentication, and efficient AI model integration. The results highlight the system’s accuracy in detecting pathologies with an average AUC of 0.89, an easy-to-use interface, and the transformative impact of AI explainability in clinical settings.