Weakly-Supervised Classification of HER2 Expression in Breast Cancer Haematoxylin and Eosin Stained Slides

Abstract

<jats:p>Human epidermal growth factor receptor 2 (HER2) evaluation commonly requires immunohistochemistry (IHC) tests on breast cancer tissue, in addition to the standard haematoxylin and eosin (H&amp;E) staining tests. Additional costs and time spent on further testing might be avoided if HER2 overexpression could be effectively inferred from H&amp;E stained slides, as a preliminary indication of the IHC result. In this paper, we propose the first method that aims to achieve this goal. The proposed method is based on multiple instance learning (MIL), using a convolutional neural network (CNN) that separately processes H&amp;E stained slide tiles and outputs an IHC label. This CNN is pretrained on IHC stained slide tiles but does not use these data during inference/testing. H&amp;E tiles are extracted from invasive tumour areas segmented with the HASHI algorithm. The individual tile labels are then combined to obtain a single label for the whole slide. The network was trained on slides from the HER2 Scoring Contest dataset (HER2SC) and tested on two disjoint subsets of slides from the HER2SC database and the TCGA-TCIA-BRCA (BRCA) collection. The proposed method attained 83.3 % classification accuracy on the HER2SC test set and 53.8 % on the BRCA test set. Although further efforts should be devoted to achieving improved performance, the obtained results are promising, suggesting that it is possible to perform HER2 overexpression classification on H&amp;E stained tissue slides.</jats:p>

CAD systems for colorectal cancer from WSI are still not ready for clinical acceptance

“…Moreover, labelling the tiles represents a significant effort with regards to the workload of the pathologists. Therefore, techniques such as multiple instance learning (MIL), have been adapted to computational pathology problems [68][69][70] . MIL only requires slide level labels and the original supervised problem is converted to a weakly-supervised problem.…”

Strategies for Enhancing the Multi-Stage Classification Performances of HER2 Breast Cancer from Hematoxylin and Eosin Images

“…Oliveira et al [ 25 ] developed a CNN model based on multiple instance learning (MIL) approaches identifying HER2 status from H&E images. Initially, the CNN model was pre-trained from IHC images on the HER2SC dataset.…”

Automated Grading of Breast Cancer Histopathology Images Using Multilayered Autoencoder

“…BC image analysis can be applied to handle numerous pathology jobs, such as mitosis detection [18,23], tissue segmentation [24], histological classification or cancer grading [15]. Automatic analysis is commonly performed by using hematoxylin and eosin (H&E)-stained slides and deep learning techniques to enhance model performance [25]. Breast cancer diagnosis from histopathological images always remains the benchmark in clinical pathology [20][21][22][23][24].…”