Learn like a Pathologist: Curriculum Learning by Annotator Agreement for Histopathology Image Classification

• Published in: 2021 IEEE Winter Conference on Applications of Computer Vision (WACV) pp. 2472 - 2482
• Main Authors: Wei, Jerry, Suriawinata, Arief, Ren, Bing, Liu, Xiaoying, Lisovsky, Mikhail, Vaickus, Louis, Brown, Charles, Baker, Michael, Nasir-Moin, Mustafa, Tomita, Naofumi, Torresani, Lorenzo, Wei, Jason, Hassanpour, Saeed
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.01.2021
• Subjects: Computer vision; Conferences; Histopathology; Image classification; Learning systems; Task analysis; Training
• ISSN: 2642-9381
• DOI: 10.1109/WACV48630.2021.00252

Applying curriculum learning requires both a range of difficulty in data and a method for determining the difficulty of examples. In many tasks, however, satisfying these requirements can be a formidable challenge.In this paper, we contend that histopathology image classification is a compelling use case for curriculum learning. Based on the nature of histopathology images, a range of difficulty inherently exists among examples, and, since medical datasets are often labeled by multiple annotators, annotator agreement can be used as a natural proxy for the difficulty of a given example. Hence, we propose a simple curriculum learning method that trains on progressively-harder images as determined by annotator agreement.We evaluate our hypothesis on the challenging and clinically-important task of colorectal polyp classification. Whereas vanilla training achieves an AUC of 83.7% for this task, a model trained with our proposed curriculum learning approach achieves an AUC of 88.2%, an improvement of 4.5%. Our work aims to inspire researchers to think more creatively and rigorously when choosing contexts for applying curriculum learning.