Quantifying and leveraging predictive uncertainty for medical image assessment |
| |
Institution: | 1. Siemens Healthineers, Digital Technology and Innovation, Princeton, NJ, USA;2. Siemens Healthineers, Digital Technology and Innovation, Bangalore, India;3. University of Michigan, Department of Radiation Oncology, Ann Arbor, MI, USA;4. Department of Radiology, Massachusetts General Hospital, Boston, MA, USA;5. Harvard Medical School, Boston, MA, USA;1. Centre for Medical Image Computing, University College London, London, United Kingdom;2. Université Côte dAzur, Inria, Epione Team, 06902 Sophia Antipolis, France;3. School of Biomedical Engineering and Imaging Sciences, King’s College London, London, United Kingdom;4. NIHR Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and the Institute of Ophthalmology, University College London, London, United Kingdom;1. College of Computer Science and Software Engineering, Shenzhen University, Shenzhen, China, 518060;2. National-Regional Key Technology Engineering Laboratory for Medical Ultrasound, Guangdong Key Laboratory for Biomedical Measurements and Ultrasound Imaging, School of Biomedical Engineering, Health Science Center, Shenzhen University, Shenzhen, China, 518060;1. Department of Electrical and Computer Engineering, University of California, Los Angeles, United States;2. Department of Radiology and Diagnostic Imaging, University of Alberta, Edmonton, Canada;3. Servier Virtual Cardiac Centre, Mazankowski Alberta Heart Institute, Edmonton, Canada;4. Department of Computing Science, University of Alberta, Edmonton, Canada;1. Centre for Computational Imaging and Simulation Technologies in Biomedicine (CISTIB), School of Computing, University of Leeds, Leeds, UK;2. Leeds Institute for Cardiovascular and Metabolic Medicine (LICAMM), School of Medicine, University of Leeds, Leeds, UK;3. Queen Square Institute of Neurology, University College London, London, UK;4. Centre for Medical Image Computing, Department of Computer Science, University College London, London, UK;5. Oxford Center for Clinical Magnetic Resonance Research (OCMR), Division of Cardiovascular Medicine, John Radcliffe Hospital, University of Oxford, Oxford, UK;6. William Harvey Research Institute, Barts Heart Centre, Barts Health NHS Trust, Queen Mary University of London, London, UK;7. Medical Imaging Research Center (MIRC), University Hospital Gasthuisberg, and Cardiovascular Science and Electronic Engineering Departments, KU Leuven, Leuven, Belgium;1. Processing Speech and Images, Department of Electrical Engineering, KU Leuven, Kasteelpark Arenberg 10/2440, Leuven 3001, Belgium;2. Medical Imaging Research Center, UZ Leuven, Herestraat 49/7003, Leuven 3000, Belgium;3. icometrix, Kolonel Begaultlaan 1b/12, Leuven 3000, Belgium;4. Department of Anatomy, University of Pretoria, Private Bag X20, Pretoria 0028, Republic of South-Africa |
| |
Abstract: | The interpretation of medical images is a challenging task, often complicated by the presence of artifacts, occlusions, limited contrast and more. Most notable is the case of chest radiography, where there is a high inter-rater variability in the detection and classification of abnormalities. This is largely due to inconclusive evidence in the data or subjective definitions of disease appearance. An additional example is the classification of anatomical views based on 2D Ultrasound images. Often, the anatomical context captured in a frame is not sufficient to recognize the underlying anatomy. Current machine learning solutions for these problems are typically limited to providing probabilistic predictions, relying on the capacity of underlying models to adapt to limited information and the high degree of label noise. In practice, however, this leads to overconfident systems with poor generalization on unseen data. To account for this, we propose a system that learns not only the probabilistic estimate for classification, but also an explicit uncertainty measure which captures the confidence of the system in the predicted output. We argue that this approach is essential to account for the inherent ambiguity characteristic of medical images from different radiologic exams including computed radiography, ultrasonography and magnetic resonance imaging. In our experiments we demonstrate that sample rejection based on the predicted uncertainty can significantly improve the ROC-AUC for various tasks, e.g., by 8% to 0.91 with an expected rejection rate of under 25% for the classification of different abnormalities in chest radiographs. In addition, we show that using uncertainty-driven bootstrapping to filter the training data, one can achieve a significant increase in robustness and accuracy. Finally, we present a multi-reader study showing that the predictive uncertainty is indicative of reader errors. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|