Focused Active Learning for Histopathological Image Classification

arXiv:2404.04663v1 Announce Type: new
Abstract: Active Learning (AL) has the potential to solve a major problem of digital pathology: the efficient acquisition of labeled data for machine learning algorithms. However, existing AL methods often struggle in realistic settings with artifacts, ambiguities, and class imbalances, as commonly seen in the medical field. The lack of precise uncertainty estimations leads to the acquisition of images with a low informative value. To address these challenges, we propose Focused Active Learning (FocAL), which combines …

abstract acquisition active learning algorithms arxiv class classification cs.ai cs.cv data digital digital pathology however image machine machine learning machine learning algorithms major medical medical field pathology solve struggle type