Unsupervised anomaly localization in high-resolution breast scans using deep pluralistic image completion. (arXiv:2305.03098v1 [eess.IV])

Automated tumor detection in Digital Breast Tomosynthesis (DBT) is a
difficult task due to natural tumor rarity, breast tissue variability, and high
resolution. Given the scarcity of abnormal images and the abundance of normal
images for this problem, an anomaly detection/localization approach could be
well-suited. However, most anomaly localization research in machine learning
focuses on non-medical datasets, and we find that these methods fall short when
adapted to medical imaging datasets. The problem is alleviated when we solve
the task …

anomaly anomaly detection arxiv automated dbt detection digital image images localization natural normal unsupervised