A Service Oriented Architecture for Clinical Decision Support Systems Based on Artificial Intelligence

The proliferation of digital health applications has created an abundance of different data that doctors must synthesize promptly to make the best patient care decisions. This mass of data contributes to physician burnout, a critical problem in healthcare. At the same time, there is growing enthusiasm for the integration of artificial intelligence (AI) technologies, such as Machine Learning and Deep Learning, to improve physicians' decision-making processes. Establishing a mechanism to integrate human feedback becomes critical to instilling confidence in AI models, employing human-in-the-loop implementation models and participatory design approaches. This study aims to emphasize the importance of integrating data from various sources and extracting explicit knowledge to address the challenges of clinical decision flows. The overall objective is to provide a modern service-oriented architecture that facilitates the collection of clinical data from heterogeneous sources by facilitating the deployment of AI innovation in healthcare. The dynamic modeling of the system is proposed from a real-life case study, in which an existing medical system has been enhanced through the proposed architecture.