Joint multi-task learning improves weakly-supervised biomarker prediction in computational pathology

arXiv:2403.03891v1 Announce Type: cross
Abstract: Deep Learning (DL) can predict biomarkers directly from digitized cancer histology in a weakly-supervised setting. Recently, the prediction of continuous biomarkers through regression-based DL has seen an increasing interest. Nonetheless, clinical decision making often requires a categorical outcome. Consequently, we developed a weakly-supervised joint multi-task Transformer architecture which has been trained and evaluated on four public patient cohorts for the prediction of two key predictive biomarkers, microsatellite instability (MSI) and homologous recombination deficiency (HRD), trained …

abstract arxiv cancer categorical clinical computational continuous cs.cv cs.lg decision decision making deep learning eess.iv making multi-task learning pathology prediction regression through type weakly-supervised