Assessments of tumor-infiltrating lymphocytes in breast cancer using machine learning are reviewed and mapped for translational readiness and methodological gaps.
Key Takeaways
- Review mapped 51 histopathology-image ML studies across detection, prognostication, and treatment-response prediction
- External validation was reported in 31.6% of detection
- Calibration and decision-curve analysis were nearly absent, with calibration reported in only one prognostication study