Through training on unlabeled data, anomaly detection has the potential to impact computer-aided diagnosis by outlining suspicious regions. Previous work on deep-learning-based anomaly detection has primarily focused on the reconstruction error. We argue instead, that pixel-wise anomaly ratings derived from a Variational Autoencoder based score approximation yield a theoretically better grounded and more faithful estimate. In our experiments, Variational Autoencoder gradient-based rating outperforms other approaches on unsupervised pixel-wise tumor detection on the BraTS-2017 dataset with a ROC-AUC of 0.94.
@article{arxiv.1912.00003,
title = {A Case for the Score: Identifying Image Anomalies using Variational Autoencoder Gradients},
author = {David Zimmerer and Jens Petersen and Simon A. A. Kohl and Klaus H. Maier-Hein},
journal= {arXiv preprint arXiv:1912.00003},
year = {2019}
}