Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery

TL;DR

This paper introduces AnoGAN, an unsupervised deep learning method using GANs to detect anomalies in medical images, aiding marker discovery without requiring annotated training data.

Contribution

The paper presents AnoGAN, a novel unsupervised GAN-based approach for anomaly detection in medical imaging, reducing reliance on annotated datasets and enabling discovery of new disease markers.

Findings

Successfully identified retinal anomalies such as fluid and hyperreflective foci.

Demonstrated effective anomaly detection on OCT retinal images.

Provided a new unsupervised framework for medical image marker discovery.

Abstract

Obtaining models that capture imaging markers relevant for disease progression and treatment monitoring is challenging. Models are typically based on large amounts of data with annotated examples of known markers aiming at automating detection. High annotation effort and the limitation to a vocabulary of known markers limit the power of such approaches. Here, we perform unsupervised learning to identify anomalies in imaging data as candidates for markers. We propose AnoGAN, a deep convolutional generative adversarial network to learn a manifold of normal anatomical variability, accompanying a novel anomaly scoring scheme based on the mapping from image space to a latent space. Applied to new data, the model labels anomalies, and scores image patches indicating their fit into the learned distribution. Results on optical coherence tomography images of the retina demonstrate that the approach correctly identifies anomalous images, such as images containing retinal fluid or hyperreflective foci.