The growing availability of medical data has sparked fresh interests in Computerized Clinical Decision Support Systems (CDSS), thanks to recent breakthroughs in machine and deep learning. CDSS has showed a lot of promise in terms of improving healthcare, enhancing the safety of patients and minimizing treatment costs. The application of CDSS, nonetheless, is unsafe since an insufficient or defective CDSS may possibly degrade healthcare quality and place patients at potential threat. Furthermore, the deployment of a CDSS may fail when the CDSS's output is ignored by its intended users owing to a lack of confidence, relevance, or actionability. We offer literature-based advice for the various elements of CDSS adoption, with a particular emphasis on Artificial Intelligence (AI) and Machine Learning (ML) systems: quality assurance, deployment, commissioning, acceptability tests, and selection, in this research. A critical selection process will assist in the process of identifying CDSS, which effectively suits the localized sites’ needs and preferences. Acceptance testing ensures that the chosen CDSS meets the specified standards and meets the safety criteria. The CDSS will be ready for safe clinical usage at the local site once the commissioning procedure is completed. An efficient system implementation must result in a smooth rollout of the CDSS to well-trained end-users with reasonable expectations. Furthermore, quality assurance will ensure that the CDSS's levels are maintained and that any problems are discovered and resolved quickly. We conclude this research by discussing the methodical adoption process for CDSS to assist in avoiding issues, enhance the safety of patients and increasing quality of service.
Transparent Artificial Intelligence – A Conceptual Framework for Evaluating AI-based Clinical Decision Support Systems
