Performance and Reliability Landscape of AI Models in Healthcare Diagnostics

Healthcare diagnostic has been significantly transformed by Artificial Intelligence (AI) that leads to personalized treatment, precise prediction of disease, and early detection. There is also greater deal of work done on privacy preservation using AI as well as some emerging studies towards trustworthiness too in clinical studies. Existing review work are noted to be quite fragmented making it quite challenging to extract a holistic inference towards influence of AI on healthcare diagnostic.  Therefore, the proposed study introduces a comprehensive, compact, and highly systematic review of AI approaches exercised towards healthcare diagnostic using Machine Learning (ML), Deep Learning (DL). Adopting PRISMA methodology, the proposed study presents varied scale of methodologies used in clinical study. The study contributes to highlighting emerging trends, unsolved challenges, frequently used approach, etc.