Perspectives on Artificial Intelligence in Clinical Healthcare Applications

Abstract

The concept of artificial intelligence (AI) has a lengthy history. It has become evident that achieving human-level intelligence is more complex than initially expected. Presently, there is a resurgence of interest in AI, driven by a substantial augmentation in computer capacity and an even greater proliferation of data, with advancements in AI technologies such as deep learning. Healthcare is seen as the subsequent sector poised for transformation using artificial intelligence. Although AI methodologies are highly effective for developing some algorithms, biological applications present distinct problems. We suggest six recommendations the 6Rs to enhance AI initiatives in the biomedical domain, particularly in clinical healthcare, and to promote dialogue between AI researchers and medical practitioners: (1) Formulate a pertinent and clearly articulated clinical inquiry; (2) Acquire appropriate data that is representative and of high quality; (3) Ensure the ratio of patients to their variables aligns with the AI methodology; (4) Establish a direct and causal relationship between the data and the ground truth; (5) Ensure regulatory compliance to facilitate validation; and (6) Employ the appropriate AI methodology.