Multi-Layer Pseudo-Supervision for Histopathology Tissue Semantic Segmentation using Patch-level Classification Labels

Abstract

Tissue-level semantic segmentation is a vital step in computational pathology. Fully-supervised models have already achieved outstanding performance with dense pixel-level annotations. However, drawing such labels on the giga-pixel whole slide images is extremely expensive and time-consuming. In this paper, we use only patch-level classification labels to achieve tissue semantic segmentation on histopathology images, finally reducing the annotation efforts. We proposed a two-step model including a classification and a segmentation phases. In the classification phase, we proposed a CAM-based model to generate pseudo masks by patch-level labels. In the segmentation phase, we achieved tissue semantic segmentation by our proposed Multi-Layer Pseudo-Supervision. Several technical novelties have been proposed to reduce the information gap between pixellevel and patch-level annotations. As a part of this paper, we introduced a new weakly-supervised semantic segmentation (WSSS) dataset for lung adenocarcinoma (LUAD-HistoSeg). We conducted several experiments to evaluate our proposed model on two datasets. Our proposed model outperforms two state-ofthe-art WSSS approaches. Note that we can achieve comparable quantitative and qualitative results with the fully-supervised model, with only around a 2% gap for MIoU and FwIoU. By comparing with manual labeling, our model can greatly save

CAM-TMIL: A Weakly-Supervised Segmentation Framework for Histopathology based on CAMs and MIL

“…Thus, they proposed a Class-specific Erasing method to erase the target and perform segmentation. Han C et al presented a Progress Dropout Attention, a progressive drop coefficient to restrain the CAMs generated by the classification network from only focusing on the most discriminative part and making the model learn the spatial information as more as possible [6]. Sun K et al designed a network called Erased CAM Supervision Net (ECS-Net), selecting features by the threshold of confidence to erase and preventing the classification network from learning the notable object features in its training procedure [7].…”

WSSS4LUAD: Grand Challenge on Weakly-supervised Tissue Semantic Segmentation for Lung Adenocarcinoma

“…CAM-based models [67] have been widely used in weakly-supervised semantic segmentation (WSSS), including histopathology images [68]. The organizers of WSSS4LUAD challenge also proposed a CAM-based model [69] to achieve WSSS for tissue segmentation. They introduced a progressive dropout attention mechanism to gradually deactivate the most distinguishable regions.…”

“…We introduce our previous proposed approach MLPS [69], a weakly-supervised semantic segmentation model tailored for the histopathology images. We use the patch-level labels collected in the previous phase to train MLPS.…”