Denoising Adversarial Autoencoders: Classifying Skin Lesions Using Limited Labelled Training Data

TL;DR

This paper introduces a semi-supervised denoising adversarial autoencoder that effectively classifies skin lesions by leveraging large unlabelled datasets and limited labelled data, outperforming traditional methods.

Contribution

The novel model combines adversarial and denoising autoencoders for semi-supervised learning in medical image classification, specifically for skin lesion diagnosis.

Findings

Superior classification accuracy with limited labelled data

Effective utilization of unlabelled data for feature learning

Both adversarial and denoising components enhance performance

Abstract

We propose a novel deep learning model for classifying medical images in the setting where there is a large amount of unlabelled medical data available, but labelled data is in limited supply. We consider the specific case of classifying skin lesions as either malignant or benign. In this setting, the proposed approach -- the semi-supervised, denoising adversarial autoencoder -- is able to utilise vast amounts of unlabelled data to learn a representation for skin lesions, and small amounts of labelled data to assign class labels based on the learned representation. We analyse the contributions of both the adversarial and denoising components of the model and find that the combination yields superior classification performance in the setting of limited labelled training data.