Artificial intelligence versus clinicians: systematic review of design, reporting standards, and claims of deep learning studies

Short summary

A systematic review of 81 non-randomized deep learning studies in medical imaging found that only 6% were tested in a real-world clinical setting, 58% had high risk of bias, and data/code availability was lacking in 95% and 93% of studies, respectively.

AI-generated from the title and abstract; the full text is not read.