COVID-19 Interpretable Diagnosis Algorithm Based on a Small Number of Chest X-Ray Samples

doi:10.1007/s12204-021-2393-2

Abstract

Abstract: The COVID-19 medical diagnosis method based on individual’s chest X-ray (CXR) is achieved difficultly in the initial research, owing to difficulties in identifying CXR data of COVID-19 individuals. At the beginning of the study, infected individuals’ CXRs were scarce. The combination of artificial intelligence (AI) and medical diagnosis has been advanced and popular. To solve the difficulties, the interpretability analysis of AI model was used to explore the pathological characteristics of CXR samples infected with COVID-19 and assist in medical diagnosis. The dataset was expanded by data augmentation to avoid overfitting. Transfer learning was used to test different pre-trained models and the unique output layers were designed to complete the model training with few samples. In this study, the output results of four pre-trained models in three different output layers were compared, and the results after data augmentation were compared with the results of the original dataset. The control variable method was used to conduct independent tests of 24 groups. Finally, 99.23% accuracy and 98% recall rate were obtained, and the visual results of CXR interpretability analysis were displayed. The network of COVID-19 interpretable diagnosis algorithm has the characteristics of high generalization and lightweight. It can be quickly applied to other urgent tasks with insufficient experimental data. At the same time, interpretability analysis brings new possibilities for medical diagnosis.

Key words: COVID-19, chest X-ray (CXR), interpretability, data augmentation, transfer learning, convolutional neural network

CLC Number:

TP 183
R 445

BU Ran (卜冉), XIANG Wei∗ (向伟), CAO Shitong (曹世同). COVID-19 Interpretable Diagnosis Algorithm Based on a Small Number of Chest X-Ray Samples[J]. J Shanghai Jiaotong Univ Sci, 2022, 27(1): 81-89.

References 16

[1]	CHEN N S, ZHOU M, DONG X, et al. Epidemiologicaland clinical characteristics of 99 cases of 2019 novelcoronavirus pneumonia in Wuhan, China: A descriptivestudy [J]. The Lancet, 2020, 395(10223): 507-513.
[2]	LU Y Q, SHEN L, HE B. Application of artificial intelligencein assisted diagnosis and treatment of cardiovasculardisease [J]. Journal of Shanghai JiaotongUniversity (Medical Science), 2020, 40(2): 259-262 (inChinese).
[3]	ZHANG L, CHEN Q, JIANG B B, et al. Preliminarystudy on motion artifacts removal of coronaryCT angiography using generative adversarial network[J]. Journal of Shanghai Jiao Tong University (MedicalScience), 2020, 40(9): 1229-1235 (in Chinese).
[4]	LIU Y H, ZHANG F D, ZHANG Q Y, et al. Cross-viewcorrespondence reasoning based on bipartite graphconvolutional network for mammogram mass detection[C]//2020 IEEE/CVF Conference on ComputerVision and Pattern Recognition (CVPR). Seattle, WA:IEEE, 2020: 3811-3821.
[5]	RIBEIRO M T, SINGH S, GUESTRIN C. “Whyshould I trust you?” Explaining the predictions of anyclassifier [C]//Proceedings of the 22nd ACM SIGKDDInternational Conference on Knowledge Discovery andData Mining. San Francisco, CA: ACM, 2016: 1135-1144.
[6]	PALATNIK DE SOUSA I, VELLASCO M M B R,COSTA DA SILVA E. Local interpretable modelagnosticexplanations for classification of lymph nodemetastases [J]. Sensors, 2019, 19(13): 2969.
[7]	SIMONYAN K, ZISSERMAN A. Verydeep convolutional networks for large-scaleimage recognition [EB/OL]. [2021-04-12].https://arxiv.org/abs/1409.1556.
[8]	SZEGEDY C, VANHOUCKE V, IOFFE S, et al.Rethinking the inception architecture for computervision [EB/OL]. [2021-04-12]. https://arxiv.org/abs/1512.00567
[9]	HU X C, MU H Y, ZHANG X Y, et al. Meta-SR:A magnification-arbitrary network for super-resolution[C]//2019 IEEE/CVF Conference on Computer Visionand Pattern Recognition (CVPR). Long Beach,CA: IEEE, 2019: 1575-1584.
[10]	ZHANG X F, WU G. Data augmentation methodbased on generative adversarial network [J]. ComputerSystems Applications, 2019, 28(10): 201-206 (in Chinese).
[11]	BRUNESE L, MERCALDO F, REGINELLI A, etal. Explainable deep learning for pulmonary diseaseand coronavirus COVID-19 detection from X-rays[J]. Computer Methods and Programs in Biomedicine,2020, 196: 105608.
[12]	COZZI D, ALBANESI M, CAVIGLI E, et al. Chest Xrayin new Coronavirus Disease 2019 (COVID-19) infection:Findings and correlation with clinical outcome[J]. La Radiologia Medica, 2020, 125(8): 730-737.
[13]	SCHIAFFINO S, TRITELLA S, COZZI A, et al. Diagnosticperformance of chest X-ray for COVID-19pneumonia during the SARS-CoV-2 pandemic in Lombardy,Italy [J]. Journal of Thoracic Imaging, 2020,35(4): W105-W106.
[14]	CHENG K Y, SHIWX, ZHAN Y Z. Research Progresson explicability of deep learning [J]. Journal of ComputerResearch and Development, 2020, 57(6): 1208-1217 (in Chinese).
[15]	SELVARAJU R R, COGSWELL M, DAS A, et al.Grad-CAM: Visual explanations from deep networksvia gradient-based localization [J]. International Journalof Computer Vision, 2020, 128(2): 336-359.
[16]	ANTHIMOPOULOS M, CHRISTODOULIDIS S,EBNER L, et al. Lung pattern classification for interstitiallung diseases using a deep convolutional neuralnetwork [J]. IEEE Transactions on Medical Imaging,2016, 35(5): 1207-1216.