Regression Constraint for an Explainable Cervical Cancer Classifier

TL;DR

This paper develops a deep learning-based classifier for cervical cancer screening that achieves high accuracy and uses attribution methods to interpret learned features, enhancing explainability.

Contribution

It introduces a novel regression constraint in the classifier to improve interpretability and demonstrates its effectiveness on the Herlev dataset.

Findings

74.5% accuracy on severity classification

94% accuracy on normal/abnormal classification

Use of attribution methods to identify discriminative features

Abstract

This article adresses the problem of automatic squamous cells classification for cervical cancer screening using Deep Learning methods. We study different architectures on a public dataset called Herlev dataset, which consists in classifying cells, obtained by cervical pap smear, regarding the severity of the abnormalities they represent. Furthermore, we use an attribution method to understand which cytomorphological features are actually learned as discriminative to classify severity of the abnormalities. Through this paper, we show how we trained a performant classifier: 74.5\% accuracy on severity classification and 94\% accuracy on normal/abnormal classification.