Advances in Detecting Latent Tuberculosis Infection in High-Risk Groups: from Immunodiagnostics to Machine Learning (Systematic Review)

The detection of latent tuberculosis infection (LTBI) is essential for improving the epidemiological situation in the region. As tuberculosis prevalence declines, identifying LTBI in risk groups becomes increasingly significant. At present, no method exists that can reliably confirm or exclude the presence of latent tuberculosis infection. In this study, we analyzed 71 publications on methods for LTBI detection, including modern immunological tests based on Mycobacterium tuberculosis antigens ESAT-6 and CFP-10, as well as interferon-γ release assays (IGRAs). Developing diagnostic strategies for LTBI is particularly relevant for individuals with diabetes mellitus, congenital or acquired immunodeficiency, those receiving immunosuppressive therapy, patients undergoing hemodialysis, as well as for children and the elderly. The search for novel LTBI biomarkers using artificial intelligence, aimed at improving detection and predicting the progression to active tuberculosis seems to be promising advancement in phthisiology.