CAMELYON17
PulseAugur coverage of CAMELYON17 — every cluster mentioning CAMELYON17 across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
PRISM benchmark evaluates pathology foundation model reliability
Researchers have introduced PRISM, a new benchmark designed to evaluate the reliability of pathology foundation models. PRISM assesses models across four key areas: model choice, label availability, calibration, and out…
-
New MFE-MIL framework enhances whole slide image analysis in pathology
Researchers have developed a new framework called Masked Feature Encoding for Multiple Instance Learning (MFE-MIL) to improve the analysis of whole slide images in computational pathology. This method uses a feature-spa…
-
New method reveals spatial shortcut patterns in vision models
Researchers have developed a new method to identify and characterize shortcut learning in vision models by grouping per-image contribution maps into recurring spatial patterns. This approach, utilizing K-means and non-n…
-
Test-time adaptation in pathology models can destabilize explanations
A new benchmark study on test-time adaptation (TTA) in computational pathology reveals that while TTA methods improve model accuracy, they can significantly alter the model's explanations. Researchers found that frozen-…
-
New ProsMAE framework enhances histopathology representation learning
Researchers have developed ProsMAE, a novel multi-source Masked Autoencoder framework designed for histopathology representation learning. This approach utilizes tiles from diverse datasets like PANDA, CAMELYON17, and B…
-
New AI method aligns cellular sheaves with attention for pathology localization
Researchers have developed a new method for interpreting weakly-supervised pathology localization in whole-slide images by combining cellular sheaves with classifier attention. This approach aims to improve the trustwor…
-
Federated learning framework FedHD aligns WSI features for collaborative pathology
Researchers have introduced FedHD, a new federated learning framework designed for collaborative digital pathology. This framework addresses challenges posed by diverse architectures and feature extractors across instit…