Deep weakly-supervised learning methods for classification and localization in histology images: A survey
Deep weakly-supervised learning methods for classification and localization in histology images: A survey
Rony, Jérôme and Belharbi, Soufiane and Dolz, Jose and Ayed, Ismail Ben and McCaffrey, Luke and Granger, Eric
arXiv 2019
Abstract : Using state-of-the-art deep learning models for the computer-assisted diagnosis of diseases like cancer presents several challenges related to the nature and availability of labeled histology images. In particular, cancer grading and localization in these images normally relies on both image- and pixel-level labels, the latter requiring a costly annotation process. In this survey, deep weakly-supervised learning (WSL) models are investigated to identify and locate diseases in histology images, without the need for pixel-level annotations. Given training data with image-level labels, these models allow to simultaneously classify histology images and yield pixel-wise localization scores, thereby identifying the corresponding regions of interest. These models are organized into two main approaches that differ in their mechanism for building attention maps to localize salient regions – (1) bottom-up approaches based on forward-pass information through a network, either by spatial pooling of representations/scores, or by detecting class regions; and (2) top-down approaches based on backward-pass information within a network, inspired by human visual attention. Since relevant WSL models have mainly been investigated within the computer vision community, and validated on natural scene images, we assess the extent to which they apply to histology images which have challenging properties, e.g. very large size, non-salient and highly unstructured regions, stain heterogeneity, and coarse/ambiguous labels. The most relevant deep WSL models (e.g. , CAM, WILDCAT and Deep MIL) are compared experimentally in terms of accuracy (classification and pixel-wise localization) on several public benchmark histology datasets for breast and colon cancer (BACH ICIAR 2018, BreakHis, CAMELYON16, and GlaS). Furthermore, to benchmark large-scale evaluations of WSL methods for histology, we propose a protocol to build WSL datasets from Whole Slide Imaging, with publicly available deterministic code and coordinates of the sampled patches. Results indicate that several deep learning models, and in particular WILDCAT and deep MIL, can provide a high level of classification accuracy, although pixel-wise localization of cancer regions remains an issue for such images.
