Reliable Mislabel Detection for Video Capsule Endoscopy Data
This addresses the challenge of obtaining accurate annotations in medical imaging, which is crucial for reliable deep learning classification, though it is incremental as it builds on existing mislabel detection methods.
The paper tackles the problem of mislabeled data in medical imaging by introducing a framework for mislabel detection, validated on two large Video Capsule Endoscopy datasets, resulting in improved anomaly detection performance after dataset cleaning.
The classification performance of deep neural networks relies strongly on access to large, accurately annotated datasets. In medical imaging, however, obtaining such datasets is particularly challenging since annotations must be provided by specialized physicians, which severely limits the pool of annotators. Furthermore, class boundaries can often be ambiguous or difficult to define which further complicates machine learning-based classification. In this paper, we want to address this problem and introduce a framework for mislabel detection in medical datasets. This is validated on the two largest, publicly available datasets for Video Capsule Endoscopy, an important imaging procedure for examining the gastrointestinal tract based on a video stream of lowresolution images. In addition, potentially mislabeled samples identified by our pipeline were reviewed and re-annotated by three experienced gastroenterologists. Our results show that the proposed framework successfully detects incorrectly labeled data and results in an improved anomaly detection performance after cleaning the datasets compared to current baselines.