Identifying and Categorizing Anomalies in Retinal Imaging Data

TL;DR

This paper presents a novel method for detecting and categorizing anomalies in retinal images using a deep autoencoder trained on healthy data, enabling unsupervised identification of pathological regions without expert annotations.

Contribution

It introduces an unsupervised approach combining deep autoencoders and one-class SVMs for anomaly detection and categorization in medical imaging, bypassing the need for labeled data.

Findings

Successfully identified pathologic regions without annotations

Outperformed standard embeddings in classification tasks

Categorized anomalies into clinically meaningful classes

Abstract

The identification and quantification of markers in medical images is critical for diagnosis, prognosis and management of patients in clinical practice. Supervised- or weakly supervised training enables the detection of findings that are known a priori. It does not scale well, and a priori definition limits the vocabulary of markers to known entities reducing the accuracy of diagnosis and prognosis. Here, we propose the identification of anomalies in large-scale medical imaging data using healthy examples as a reference. We detect and categorize candidates for anomaly findings untypical for the observed data. A deep convolutional autoencoder is trained on healthy retinal images. The learned model generates a new feature representation, and the distribution of healthy retinal patches is estimated by a one-class support vector machine. Results demonstrate that we can identify pathologic regions in images without using expert annotations. A subsequent clustering categorizes findings into clinically meaningful classes. In addition the learned features outperform standard embedding approaches in a classification task.