Artificial intelligence algorithms are being created both investigationally and commercially. Evaluation of their performance is important for developers, investigators, clinical physicians, and regulatory agencies. No clear consensus exists on what metrics are best for algorithmic evaluation for AI and ML applications in radiology. We review the basics of the confusion matrix, continue to single number summary values such as accuracy, F1 score, and ɸ coefficient, and then discuss Receiver Operator Curves and their derivatives, Precision Recall Curves, and Cost Curves. Recommendations are made for potential future directions and what currently may be best practices in algorithmic evaluation metrics.
Comments: 17 Pages. v3: spelling corrections, added Cohen's Kappa, image and formula corrections
Unique-IP document downloads: 50 times
Vixra.org is a pre-print repository rather than a journal. Articles hosted may not yet have been verified by peer-review and should be treated as preliminary. In particular, anything that appears to include financial or legal advice or proposed medical treatments should be treated with due caution. Vixra.org will not be responsible for any consequences of actions that result from any form of use of any documents on this website.
Add your own feedback and questions here:
You are equally welcome to be positive or negative about any paper but please be polite. If you are being critical you must mention at least one specific error, otherwise your comment will be deleted as unhelpful.