Development of diagnostic algorithm using machine learning for distinguishing between active tuberculosis and latent tuberculosis infection

BMC Infectious Diseases

Research

Yimin Cai¹  Ying Xue²  Ying Luo³  Wei Liu³  Huijuan Song³  Guoxing Tang³  Ziyong Sun³  Yi Huang³  Feng Wang³  Qi Wang⁴ 

[1] Department of Epidemiology and Biostatistics, Key Laboratory of Environmental Health of Ministry of Education, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Hangkong Road 13, Wuhan, China;Department of Immunology, School of Basic Medicine, Tongji Medical College, Huazhong University of Science and Technology, Hangkong Road 13, Wuhan, China;Department of Laboratory Medicine, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Jiefang Road 1095, 430030, Wuhan, China;Télécom Physique Strasbourg, Illkirch-Graffenstaden, France;

关键词: Diagnostic algorithm;    Machine learning;    Active tuberculosis;    Latent tuberculosis infection;    Discrimination;   

DOI  :  10.1186/s12879-022-07954-7

 received in 2022-09-19, accepted in 2022-12-19,  发布年份 2022

来源: Springer

【 摘 要 】

BackgroundThe discrimination between active tuberculosis (ATB) and latent tuberculosis infection (LTBI) remains challenging. The present study aims to investigate the value of diagnostic models established by machine learning based on multiple laboratory data for distinguishing Mycobacterium tuberculosis (Mtb) infection status.MethodsT-SPOT, lymphocyte characteristic detection, and routine laboratory tests were performed on participants. Diagnostic models were built according to various algorithms.ResultsA total of 892 participants (468 ATB and 424 LTBI) and another 263 participants (125 ATB and 138 LTBI), were respectively enrolled at Tongji Hospital (discovery cohort) and Sino-French New City Hospital (validation cohort). Receiver operating characteristic (ROC) curve analysis showed that the value of individual indicator for differentiating ATB from LTBI was limited (area under the ROC curve (AUC) < 0.8). A total of 28 models were successfully established using machine learning. Among them, the AUCs of 25 models were more than 0.9 in test set. It was found that conditional random forests (cforest) model, based on the implementation of the random forest and bagging ensemble algorithms utilizing conditional inference trees as base learners, presented best discriminative power in segregating ATB from LTBI. Specially, cforest model presented an AUC of 0.978, with the sensitivity of 93.39% and the specificity of 91.18%. Mtb-specific response represented by early secreted antigenic target 6 (ESAT-6) and culture filtrate protein 10 (CFP-10) spot-forming cell (SFC) in T-SPOT assay, as well as global adaptive immunity assessed by CD4 cell IFN-γ secretion, CD8 cell IFN-γ secretion, and CD4 cell number, were found to contribute greatly to the cforest model. Superior performance obtained in the discovery cohort was further confirmed in the validation cohort. The sensitivity and specificity of cforest model in validation set were 92.80% and 89.86%, respectively.ConclusionsCforest model developed upon machine learning could serve as a valuable and prospective tool for identifying Mtb infection status. The present study provided a novel and viable idea for realizing the clinical diagnostic application of the combination of machine learning and laboratory findings.

【 授权许可】

CC BY   
© The Author(s) 2022

【 图 表 】

Fig. 2

Fig. 2

Fig. 1

40517_2022_243_Article_IEq12.gif