Detection of HER2 from Haematoxylin-Eosin Slides Through a Cascade of Deep Learning Classifiers via Multi-Instance Learning

David La Barbera; António Polónia; Kevin Roitero; Eduardo Conde-Sousa; Vincenzo Della Mea

doi:10.3390/jimaging6090082

Detection of HER2 from Haematoxylin-Eosin Slides Through a Cascade of Deep Learning Classifiers via Multi-Instance Learning

J Imaging. 2020 Aug 23;6(9):82. doi: 10.3390/jimaging6090082.

Authors

David La Barbera¹, António Polónia^{2

3}, Kevin Roitero¹, Eduardo Conde-Sousa^{3

4}, Vincenzo Della Mea¹

Affiliations

¹ Department of Mathematics, Computer Science and Physics, University of Udine, 33100 Udine, Italy.
² Department of Pathology, Ipatimup Diagnostics, Institute of Molecular Pathology and Immunology, University of Porto, 4169-007 Porto, Portugal.
³ i3S-Instituto de Investigação e Inovação em Saúde, Universidade do Porto, 4169-007 Porto, Portugal.
⁴ INEB-Instituto de Engenharia Biomédica, Universidade do Porto, 4169-007 Porto, Portugal.

Abstract

Breast cancer is the most frequently diagnosed cancer in woman. The correct identification of the HER2 receptor is a matter of major importance when dealing with breast cancer: an over-expression of HER2 is associated with aggressive clinical behaviour; moreover, HER2 targeted therapy results in a significant improvement in the overall survival rate. In this work, we employ a pipeline based on a cascade of deep neural network classifiers and multi-instance learning to detect the presence of HER2 from Haematoxylin-Eosin slides, which partly mimics the pathologist's behaviour by first recognizing cancer and then evaluating HER2. Our results show that the proposed system presents a good overall effectiveness. Furthermore, the system design is prone to further improvements that can be easily deployed in order to increase the effectiveness score.

Keywords: HER2; convolutional neural networks; deep learning classification; digital pathology; multiple instance learning; whole slide image processing.