Search

Diagnosing tuberculous pleurisy (TP) remains a clinical challenge and the best method to diagnose it is controversial. Although several studies have investigated the performance of pleural fluid (PF) T-SPOT for pleural tuberculosis (plTB) diagnosis, the heterogeneity of its accuracy exists. Therefore, we performed an updated meta-analysis of the existing evidence on the utility of PF T-SPOT to diagnose TP. PubMed and EmBase were searched for relevant English articles up to July 29, 2019. Statistical analysis was performed using Stata, Revman, and Meta-Disc. Pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), and diagnostic odds ratio (DOR) were determined. Summary receiver operating characteristic (SROC) curves and the area under the curve (AUC) were used to summarize the overall diagnostic performance. A total of 13 studies (997 patients with TP and 656 patients without TP) were identified and enrolled to meta-analysis, giving the following pooled values for diagnostic accuracy of PF T-SPOT: sensitivity, 0.91 (95% CI, 0.89–0.92, I 2 = 80.9%); specificity, 0.88 (95% CI, 0.86–0.91, I 2 = 87.3%); PLR, 6.28 (95% CI, 2.88–13.69, I 2 = 93.3%); NLR, 0.12 (95% CI, 0.07–0.21, I 2 = 84.9%); DOR, 59.74 (95% CI, 24.13–147.93, I 2 = 78.3%); and the area under the SROC curve, 0.95 (95% CI, 0.93–0.97). Our meta-analysis suggests that PF T-SPOT has important diagnostic value for plTB. However, the standardization of the operating procedure needs to be further promoted, which would make the results more credible. [ABSTRACT FROM AUTHOR]

• The present study firstly applied CNN based on T-SPOT assay to differentiate MTB infection status. • CNN based on T-SPOT assay could successfully discriminate ATB from LTBI. • The diagnostic performance of model based on CNN is better than spot number or TBAg/PHA ratio. The study aims to investigate the potential of convolutional neural network (CNN) based on spot image of T-SPOT assay for distinguishing active tuberculosis (ATB) from latent tuberculosis infection (LTBI). CNN was applied to recognize and classify T-SPOT spot image. Logistic regression was used to establish prediction model based on CNN. Areas under the receiver operating characteristic curve (AUCs) of early secreted antigenic target 6 (ESAT-6) CNN, culture filtrate protein 10 (CFP-10) CNN, and phytohemagglutinin (PHA) CNN were more than 0.7 in differentiating ATB from LTBI, while the performance of these indicators was significantly better than that of spot number. Furthermore, prediction model based on the combination of CNNs yielded an AUC of 0.898. The model presented a sensitivity of 85.76% and a specificity of 90.23%. The current study identified CNN based on T-SPOT spot image with the potential to serve as a tool for TB diagnostics. [ABSTRACT FROM AUTHOR]