Table 2. Highly cited AI algorithms graded by their interpretability, dynamicity, precision, autonomy, fairness, and reproducibility.

Primary author	Algorithm application	Explainable	Dynamic	Precise	Autonomous	Fair	Reproducible
Gulshan	Detecting diabetic retinopathy	No	N/A	Yes	Yes	Yes	No
Iorio	Predicting tumor sensitivity to pharmacotherapies	Yes	N/A	Yes	No	Yes	No
Kamnitsas	Brain lesion segmentation	Yes	N/A	Yes	Yes	Yes	Yes
Ott	Predicting human lymphocyte antigen binding	No	N/A	Yes	Yes	Yes	Yes
Savova	Extracting information from clinical free text in EHRs	No	N/A	Yes	Yes	No	Yes
Tajbakhsh	Medical image classification, detection, and segmentation	No	N/A	Yes	Yes	Yes	No
Wolfe	Identifying and assessing severity of fibromyalgia	Yes	N/A	Yes	No	No	No
Xiong	Predicting splicing regulation for mRNA sequences	Yes	N/A	Yes	Yes	No	No

AI, artificial intelligence; EHR, electronic health record; N/A, not applicable (i.e., temporal changes and continuous monitoring were not applicable to these algorithms and their intended use).