AI-based image analysis in clinical testing: lessons from cervical cancer screening
Cet article présente une approche pour améliorer la performance des algorithmes d'intelligence artificielle destinés à la reconnaissance d'images dans le cadre du dépistage des cancers du col de l'utérus
Novel screening and diagnostic tests based on artificial intelligence (AI) image recognition algorithms are proliferating. Some initial reports claim outstanding accuracy followed by disappointing lack of confirmation, including our own early work on cervical screening. This is a presentation of “lessons learned”, organized as a conceptual step-by-step approach to bridge the gap between the creation of an AI algorithm and clinical efficacy. The approach includes the following fundamental principles: 1) Specify rigorously what the algorithm is designed to identify and what the test is intended to measure, eg, screening, diagnostic, or prognostic. 2) Design the AI algorithm to minimize the most clinically important errors. For example, many equivocal cervical images cannot yet be labeled because the borderline between cases and controls is blurred. To avoid a misclassified case-control dichotomy, we have isolated the equivocal cases and formally included an intermediate, indeterminate class (case > indeterminate > control). 3) Evaluate AI algorithms like any other test, using clinical epidemiologic criteria. Repeatability of the algorithm at the borderline, for indeterminate images, has proven extremely informative. Distinguishing between internal and external validation is also essential. 4) Link the AI algorithm results to clinical risk estimation. Absolute risk (not relative) is the critical metric for translating a test result into clinical use. 5) Generate risk-based guidelines for clinical use that match local resources and priorities. We are particularly interested in applications to lower-resource settings to address health disparities. We note that similar principles apply to other domains of AI-based image analysis for medical diagnostic testing.
Journal of the National Cancer Institute , article en libre accès, 2022