Tuberculosis is one of the major infectious diseases that seriously endanger human health. Since 2014, it has surpassed human immunodeficiency virus/acquired immunodeficiency syndrome as the first infectious disease in patients with single pathogens. China is the third-largest country in the world in terms of high burden of tuberculosis. In 2016, there were about 900 000 new cases of tuberculosis in China. China is facing a severe tuberculosis epidemic, especially for the early diagnosis of tuberculosis and misdiagnosis of tuberculosis, which leads to delay in treatment and the spread of tuberculosis. With the application of artificial intelligence in the medical field, machine learning and deep learning methods have shown important value in the diagnosis of tuberculosis. This article will explain the application status and future development of machine learning and deep learning in the diagnosis of tuberculosis.
Delay in diagnosis of tuberculosis and the presence of drug-resistant tuberculosis are huge threats to global tuberculosis disease control. Early detection of active tuberculosis, especially the detection of drug-resistant Mycobacterium tuberculosis strains, is necessary. This paper emphasizes on the application of the molecular diagnostic techniques in the field of rapid detection of Mycobacterium tuberculosis and drug-resistant Mycobacterium tuberculosis, and discusses the performance of current molecular diagnostic techniques in solving clinical detection difficulties. The paper aims to provide the theoretical thinking for the diagnosis of tuberculosis in the future.
ObjectiveTo explore the single locus mutation that related to hepatitis B virus (HBV) co-infection by means of genome-wide association study (GWAS) in Chinese Han patients with pulmonary tuberculosis (TB).MethodsA total of 946 patients with pulmonary TB enrolled between March 2013 and March 2018 were genotyped by Illumina Human Omni Express gene chip. After quality control, 389 972 single nucleotide polymorphisms (SNPs) of 703 patients with single TB infection and 53 patients with TB-HBV co-infection were included in the follow-up association analysis.ResultsThe SNP with the strongest statistical correlation signal was rs118122819 (P=2.923×10−12, odds ratio=7.933) located on chromosome 8p23.1. Other potential susceptibility genes included CDH4 (rs73309833), MARCH1 (rs3797020), and DNER (rs13393112), etc. In addition, a strong linkage imbalance between rs118122819 and rs4840365 (D’=0.88, r2=0.76) was found, while rs4840365 was located in the MFHAS1 gene region.ConclusionsThis study provides evidence for the presence of susceptibility gene locus for HBV co-infection in pulmonary TB patients, and provides important clues for the mechanism research, disease prevention, and treatment of co-infection. But these associations must be replicated and validated in larger studies.
ObjectiveTo explore the relationship between hexokinase domain-containing protein 1 (HKDC-1) single nucleotide polymorphism (SNP) and first-line anti-tuberculosis drug-induced liver injury (ATDILI) in tuberculosis patients in western China.MethodsFrom November 2016 to April 2018, 746 tuberculosis patients treated in West China Hospital of Sichuan University were collected and divided into ATDILI group and non-ATDILI group according to the liver function indicators. DNA was extracted by QIAamp® DNA Blood Mini Kit (Qiagen, Germany). Seven SNPs of the HKDC-1 gene were genotyped by high-throughput genotyping technique and the differences between the two groups were compared.ResultsThere were 118 ATDILI and 628 non-ATDILI cases enrolled in this study. In clinical symptoms, the differences in incidences of fever and weight loss between the two groups were statistically significant (P=0.004, 0.024). The C allele at rs906219 was associated with low susceptibility to ATDILI [odds ratio (OR)=0.737, 95% confidence interval (CI) (0.556, 0.957), P=0.033], and the additive model and dominant model showed that CC/CA genotype had a lower risk of ATDILI than AA genotype [CC vs. AA: OR=0.563, 95%CI (0.325, 0.976), P=0.039; CC+CA vs. AA: OR=0.533, 95%CI (0.348, 0.817), P=0.004].ConclusionThe SNP of rs906219 in HKDC-1 is correlated with ATDILI occurrence in tuberculosis patients in western China, which provides clues for personalized anti-tuberculosis treatment.