Denoising adversarial autoencoders: classifying skin lesions using limited labelled training data

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.

Machine Learning and Deep Learning Methods for Skin Lesion Classification and Diagnosis: A Systematic Review

“…Finally, ad hoc clustering was performed using a pre-trained deep Levenberg-Marquardt neural network. Creswell et al [161] proposed denoising adversarial autoencoders to classify limited and imbalanced skin lesion images. Harangi [162] ensembled four different CNNs to investigate the impact on performance.…”

A Review of Generative Adversarial Networks in Cancer Imaging: New Applications, New Solutions

“…The scans from healthy subjects were used to train the autoencoder model to learn the distribution of healthy images and detect pathological images as outliers. Creswell et al [231] proposed a semi-supervised Denoising Adversarial Autoencoder (ssDAAE) to learn a representation based on unlabelled skin lesion images. The semi-supervised part of their CNN-based architecture corresponds to malignancy classification of labelled skin lesions based on the encoded representations of the pretrained DAAE.…”

Visual saliency based global–local feature representation for skin cancer classification

“…For example, Harangi [5] create the ensemble of DCNNs to achieve high classification accuracy, which combine the extracted depth features from different networks. Creswell et al [35] propose a novel semi‐supervised auto‐encoder which is able to use vast amounts of unlabelled data to learn a representation for skin lesions. Zhang et al [36] propose a synergistic deep learning model in which two DCNNs simultaneously learn the features of two images and determine whether the pair of input images belongs to the same class.…”