An accurate artificial intelligence system for the detection of pulmonary and extra pulmonary Tuberculosis

Sharma, Anshu; Sharma, Anurag; Malhotra, Rahul; Singh, Parulpreet; Chakrabortty, Ripon K.; Mahajan, Shubham; Pandit, Amit Kant

doi:10.1016/j.tube.2021.102143

Cited by 8 publications

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, the result images can be transferred via mobile phone technology from rural settings to server site and data can be analyzed with AI [76]. Artificial intelligence can be used to diagnose pulmonary and extrapulmonary TB [77] and to predict drug-resistant and drug-susceptible Mtb strains [78].…”

Section: Artificial Intelligencementioning

confidence: 99%

Tuberculosis Diagnosis: Updates and Challenges

Shivakumar¹,

Shettigar²

2023

Bacterial Infectious Diseases Annual Volume 2023

View full text Add to dashboard Cite

Tuberculosis (TB) is caused by a single infectious agent, Mycobacterium tuberculosis, and a public health concern due to increased cases of drug-resistance and high mortality rates. Rapid identification of tuberculosis is necessary for its early treatment and to prevent the emergence of drug-resistant strains. For effective management of patients, rapid, cost-effective, and point-of-care (POC) diagnostic methods are required. The commonly used screening and identification methods are clinical examination, radiography, sputum smear microscopy, culture method, serological method, and tuberculin skin test. In addition, several molecular methods such as NAAT based GeneXpert, loop-mediated isothermal amplification (LAMP), line probe assay (LPA), whole genome sequencing (WGS) and other non-invasive methods of lateral flow urine lipoarabinomannan assay (LF-LAM) and eNose assays are developed. Sputum smear microscopy, Xpert MTB/RIF, and LED-Fluorescence microscopy (LED-FM) are the preferred methods to use in peripheral laboratories. The non-invasive methods of tuberculosis diagnosis are more beneficial in patients from whom collecting sputum sample is difficult particularly in children and HIV co-infected patients. Molecular methods can simultaneously identify the pathogen, M. tuberculosis, and mutations in drug-resistance genes. Even though, many advanced methods are currently available, accurate and affordable diagnostic method for tuberculosis is still challenging. Here, we review and highlight the uses and challenges of currently available conventional and advanced diagnostic methods of tuberculosis screening and diagnosis.

show abstract

Section: Artificial Intelligencementioning

confidence: 99%