Nature Medicine, Published online: 12 August 2026; doi:10.1038/s41591-026-04568-3
Automated visual evaluation (AVE) could be transformative for cervical cancer screening in low-resource settings. However, AVE research has yet to yield robust clinical tools. Frugal approaches are needed that emphasize efficient, multimodal models, explainable outputs, local calibration and evaluation in real-world screening workflows.
