RaLMPH: Reliability-aware Learning for Multi-Pathologist Harmonization in Whole-Slide Image Classification
Researchers have developed RaLMPH, a novel framework for Whole-Slide Image (WSI) analysis that addresses the challenge of inter-pathologist variability in diagnostic labeling. Unlike existing methods that assume a single correct label or global annotator reliability, RaLMPH models local neighborhood structure and expert uncertainty to identify trustworthy regions within WSIs. This allows for sample-wise local annotator ranking and adaptive fusion of labels based on reliability, leading to improved performance in computational pathology. AI