Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discovery
arXiv:1703.05921
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.
To be published in the proceedings of the international conference on Information Processing in Medical Imaging (IPMI), 2017
References in corpus (4)
Cited by in corpus (33)
- GAN-based Synthetic Medical Image Augmentation for increased CNN Performance in Liver Lesion Classification
- Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders
- Deep Anomaly Detection Using Geometric Transformations
- Searching for New Physics with Deep Autoencoders
- A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
- Anomaly Detection with Generative Adversarial Networks for Multivariate Time Series
- DFR: Deep Feature Reconstruction for Unsupervised Anomaly Segmentation
- Exploiting Epistemic Uncertainty of Anatomy Segmentation for Anomaly Detection in Retinal OCT
- Towards Explaining Anomalies: A Deep Taylor Decomposition of One-Class Models
- When will the mist clear? On the Interpretability of Machine Learning for Medical Applications: a survey
- A Comparison Study of Credit Card Fraud Detection: Supervised versus Unsupervised
- Anomaly Detection for Skin Disease Images Using Variational Autoencoder
- Pseudo-healthy synthesis with pathology disentanglement and adversarial learning
- VELC: A New Variational AutoEncoder Based Model for Time Series Anomaly Detection
- Are generative deep models for novelty detection truly better?
- Airline Passenger Name Record Generation using Generative Adversarial Networks
- Unsupervised Learning of Anomaly Detection from Contaminated Image Data using Simultaneous Encoder Training
- Improving Missing Data Imputation with Deep Generative Models
- Visual Anomaly Detection for Images: A Survey
- Unsupervised RGBD Video Object Segmentation Using GANs
- Estimation of Dimensions Contributing to Detected Anomalies with Variational Autoencoders
- RSM-GAN: A Convolutional Recurrent GAN for Anomaly Detection in Contaminated Seasonal Multivariate Time Series
- Detecting Anomalous Faces with 'No Peeking' Autoencoders
- GONet: A Semi-Supervised Deep Learning Approach For Traversability Estimation
- Coupled IGMM-GANs for deep multimodal anomaly detection in human mobility data
- Squeezed Convolutional Variational AutoEncoder for Unsupervised Anomaly Detection in Edge Device Industrial Internet of Things
- Chest X-ray Inpainting with Deep Generative Models
- Group Anomaly Detection using Deep Generative Models
- Anomaly Detection in Astronomical Images with Generative Adversarial Networks
- Conformal Anomaly Detection on Spatio-Temporal Observations with Missing Data
- Generic Semi-Supervised Adversarial Subject Translation for Sensor-Based Human Activity Recognition
- Perceiving Music Quality with GANs
- P-WAE: Generalized Patch-Wasserstein Autoencoder for Anomaly Screening