Evaluation of efficiency of artificial intelligence for chest radiograph interpretation for pulmonary tuberculosis screening in mobile x-ray vehicle
Article Sidebar
Main Article Content
Abstract
Background: Tuberculosis (TB) is one of the major global health threats. The chest radiograph (CXR) is one of the primary tools for detecting TB, especially pulmonary TB. Artificial Intelligence (AI) is increasingly used with radiological technology by developing AI software for health screening by CXR.
Objectives: To compare the pulmonary TB screening results between the radiologist and AI software from the mobile x-ray screening vehicle of the Faculty of Allied Health Sciences, Thammasat University.
Materials and methods: 449 patients (408 normal, 41 abnormal) were exposed for chest radiograph at the mobile x-ray screening vehicle of Faculty of Allied Health Sciences, Thammasat University. The retrospective data was randomly collected between 2016 and 2018. The methods were divided into three steps: quality control for the x-ray machine, transferring the radiograph from digital radiography to PACS and AI, and displaying the results on the monitor with the StatPlus program.
Results: The mobile x-ray machine has passed the quality control test. In addition, the TB interpretation by AI showed Area Under Curve of 0.859 and the study demonstrated high specificity of 0.995 but low sensitivity of 0.722. The positive predictive value (PPV) was 0.951, which was less than the Negative Predictive Value of 0.963.
Conclusion: Artificial intelligence is becoming a healthcare supporter to help radiologists analyze and interpret chest radiographs and provide a fast diagnosis.
Article Details
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Personal views expressed by the contributors in their articles are not necessarily those of the Journal of Associated Medical Sciences, Faculty of Associated Medical Sciences, Chiang Mai University.
References
World Health Organization. Global tuberculosis report 2020. Geneva: World Health Organization; 2021.
Department of disease control. National tuberculosis control programme guidelines, Thailand, 2018. [in Thai]. Bangkok: 2018; 50-57.
Yu K, Beam A, Kohane I. Artificial intelligence in healthcare. Nat Biomed Eng 2018 October; 2: 719-31. doi: 10.1038/s41551-018-0305-z.
Staff news brief. FUJIFILM showcases AI for digital radiology [Internet]. 2019 Dec [cited 2020 Nov 11]. Available from: https://www.appliedradiology.com/articles/fujifilm-showcases-ai-for-digital radiography
Michelfeit F. Siemens healthineers uses artificial intelligence to take x-ray diagnostics to a new level [Internet]. 2020 Jun [cited 2020 Nov 15]. Available from: https://www.siemens-healthineers.com/press-room/press-releases/ysioxpree-ai-chest.html
Samsung brings together medical imaging and AI for radiologists at RSNA 2018 [Internet]. 2018 Nov [cited 2020 Nov 15]. Available from: https://news.samsung.com/global/samsung-brings-together-medical-imaging-and-ai-for-radiologists-at-rsna-2018
Department of medical sciences, Ministry of public health. Quality standard of diagnostic x-ray systems. 1st ed. Bangkok: The agricultural cooperative federation of Thailand limited; 2015.
Maduskar P, Muyoyeta M, Ayles H, Hogeweg L, Peters L, Ginneken B. Detection of tuberculosis using digital chest radiography: automated reading vs. interpretation by clinical officers. Int J Tuberc Lung Dis 2013; 17: 1613-20.
Hwang S, Kim H, Jeong J, Kim H. A novel approach for tuberculosis screening based on deep convolutional neural networks. Medical imaging 2016: computer-aided diagnosis 2016 Mar; 9785. doi: 10.1117/12.2216198.
Muyoyeta M, Maduskar P, Moyo M, Kasese N, Milimo D, Spooner R, et al. The sensitivity and specificity of using a computer aided diagnosis program for automatically scoring chest X-rays of presumptive TB patients compared with Xpert MTB/RIF in Lusaka Zambia. PLOS ONE 2014 Apr; 9(4): e93757. doi: 10.1371/journal.pone.0093757.
El-Solh A, Goodnough S, Serghani J, Brydon B. Predicting active pulmonary tuberculosis using an artificial neural network. Clinical investigations; chest. 1999 May; 116: 968-73.
Lakhani P, Sundaram B. Deep learning at chest radiography: automated classification of pulmonary tuberculosis by using convolutional neural networks. Radiology. 2017 Aug; 284(2): 574–82. doi: 10.1148/radiol.2017162326.
Rajpurkar P, Irvin J, Ball R, Zhu K, Yang B, Mehta H, et al. Deep learning for chest radiograph diagnosis: A retrospective comparison of the CheXNeXt algorithm to practicing radiologists. PLoS Med. 2018 Nov; 15(11):e1002686. doi: 10.1371/journal.pmed.1002686.
Steiner A, Mangu C, Hombergh J, Deutekom H, Ginneken B, Clowes P, et al. Screening for pulmonary tuberculosis in a Tanzanian prison and computer-aided interpretation of chest X-rays. Public Health Action 2015 Dec; 5(4): 249-54. doi: 10.5588/pha.15.0037.