Exploiting Epistemic Uncertainty of Anatomy Segmentation for Anomaly Detection in Retinal OCT.
Philipp SeeböckJosé Ignacio OrlandoThomas SchleglSebastian M. WaldsteinHrvoje BogunovicSophie KlimschaGeorg LangsUrsula Schmidt-ErfurthPublished in: IEEE Trans. Medical Imaging (2020)
Keyphrases
- anomaly detection
- optical coherence tomography
- oct images
- imaging modalities
- intrusion detection
- medical images
- anomalous behavior
- detecting anomalies
- fundus images
- network traffic
- network anomaly detection
- retinal images
- network intrusion detection
- computer security
- behavior analysis
- unsupervised anomaly detection
- one class support vector machines
- detecting anomalous
- image analysis
- intrusion detection system
- image segmentation
- medical imaging
- low contrast
- network security
- computer vision
- detect anomalies
- negative selection algorithm
- data mining
- malware detection
- network intrusion
- unsupervised learning
- knowledge discovery