Search papers, labs, and topics across Lattice.
This study introduces ProsMAE, a multi-source Masked Autoencoder framework designed to enhance histopathology representation learning by leveraging diverse datasets, including PANDA, CAMELYON17, and BRACS. By pretraining on these varied sources, the model effectively addresses challenges such as stain variation and tissue artifacts, leading to improved performance in ISUP grade classification. The results show that ProsMAE outperforms the standard MAE baseline, achieving a higher mean validation quadratic weighted kappa (QWK) on the disjoint PANDA split, highlighting its potential for robust diagnostic applications in computational pathology.
ProsMAE outperforms traditional methods by leveraging diverse histopathology datasets to significantly improve ISUP grade classification accuracy.
Whole slide images (WSIs) provide rich diagnostic information for computational pathology, but their gigapixel scale, stain variation, scanner differences, tissue artifacts, and limited expert annotation make robust model training challenging. This paper presents a multi-source Masked Autoencoder (MAE) framework, named ProsMAE, for histopathology representation learning. Tiles from Prostate cANcer graDe Assessment (PANDA), CAncer MEtastases in LYmph nOdes challeNge 2017 (CAMELYON17), and BReAst Carcinoma Subtyping (BRACS) are used for ProsMAE pretraining to expose the encoder to diverse tissue morphology and acquisition conditions. The learned encoder is transferred for International Society of Urological Pathology (ISUP) grade classification through ProsCLS, using a frozen encoder and a linear classification head. ProsMAE achieved a higher mean validation quadratic weighted kappa (QWK) than the vanilla MAE frozen linear-probe baseline under the evaluated disjoint PANDA split. Repeated-split evaluation remains necessary to further establish robustness across split compositions.
