Utilizing Knowledge Distillation in Deep Learning for Classification of Chest X-Ray Abnormalities
Open Access
- 1 September 2020
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 160749-160761
- https://doi.org/10.1109/access.2020.3020802
Abstract
Automatic screening and diagnosis of lung abnormalities from chest X-ray images has been recently drawing attention from the computer vision and medical imaging communities. Previous studies of deep neural networks have predominantly demonstrated the effectiveness of lung disease binary classification procedures. However, large numbers of medical images—which can be labeled with a variety of existing or suspected pathologies—are required to be interpreted and reported upon daily by an individual radiologist; this poses a challenge in maintaining a consistently high diagnosis accuracy. In this paper, we present a competitive study of knowledge distillation (KD) in deep learning for classification of abnormalities in chest X-ray images. This method aims to either distill knowledge from cumbersome teacher models into lightweight student models or to self-train these student models, to generate weakly supervised multi-label lung disease classifications. Our approach was based on multi-task deep learning architectures that, in addition to multi-class classification, supported the visualizations utilized in saliency maps of the pathological regions where an abnormality was located. A self-training KD framework, in which the model learned from itself, was shown to outperform both the well-established baseline training procedure and the normal KD, achieving the AUC improvements of up to 6.39% and 3.89%, respectively. Through application to the publicly available ChestX-ray14 dataset, we demonstrated that our approach efficiently overcame the interdependency of 14 weakly annotated thorax diseases and facilitated the state-of-the-art classification compared with the current deep learning baselines.Keywords
Funding Information
- Basic Science Research Program through the National Research Foundation of Korea
- Ministry of Education (NRF-2020R1I1A3074141)
- Brain Research Program through the NRF funded by the Ministry of Science, ICT and Future Planning (NRF-2019M3C7A1020406)
- Korea National University of Transportation in 2020
This publication has 37 references indexed in Scilit:
- Learning Deconvolution Network for Semantic SegmentationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2015
- Preparing a collection of radiology examinations for distribution and retrievalJournal of the American Medical Informatics Association, 2015
- Going deeper with convolutionsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2015
- Fully convolutional networks for semantic segmentationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2015
- Region-Based Convolutional Networks for Accurate Object Detection and SegmentationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2015
- Two public chest X-ray datasets for computer-aided screening of pulmonary diseases2014
- Activities of the Korean Institute of TuberculosisOsong Public Health and Research Perspectives, 2014
- Previous Lung Diseases and Lung Cancer Risk: A Systematic Review and Meta-AnalysisPLOS ONE, 2011
- The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial of the National Cancer Institute: History, organization, and statusControlled Clinical Trials, 2000
- Development of a Digital Image Database for Chest Radiographs With and Without a Lung NoduleAmerican Journal of Roentgenology, 2000