AI Technologies Used for Diagnostic Modalities in Drug-Resistant Tuberculosis Diagnosis

Drug-resistant tuberculosis (DR-TB) is described as tuberculosis that is not responding to one or more anti-TB medications. A person is diagnosed with DR-TB if they do not respond to regular TB treatment; then a more challenging treatment regimen and a longer treatment duration are required for DR-TB. Furthermore, DR-TB has the same capacity to spread and infect people as regular TB, even though early detection may shorten the duration and reduce the cost of TB treatment. MDR-TB and XDR-TB are two severe variants of DR-TB. MDR-TB remains a public health concern and a threat to health safety. Nowadays, artificial intelligence (AI)–driven dialog-based object query systems (DBOQS) are developing into more popular across a range of industries and offer significant practical application possibilities. The DBOQS will eventually incorporate the prototype-produced ensemble deep learning (EDL) to assist medical professionals in using DRCS as a classification tool and in diagnosing DR-TB. EDL generates 31.25% more accuracy than normal deep learning and enhances DR-TB classification by 1.17–43.43% compared with existing approaches. At the conclusion of the testing period, 99.70% of users said that they would like to continue using DRCS as a useful diagnostic tool, and the system was able to boost user confidence to 95.1% and dependability to 95.8%.