Artificial intelligence-based radiographic extent analysis to predict tuberculosis treatment outcomes: a multicenter...

Area of Science:

• Pulmonary Medicine
• Radiology
• Infectious Diseases

Background:

• Predicting treatment outcomes for pulmonary tuberculosis remains challenging.
• Effective treatments exist, but patient management requires better prognostic tools.

Purpose of the Study:

• To identify factors predicting treatment success and culture conversion in pulmonary tuberculosis.
• To evaluate the role of artificial intelligence (AI)-based chest X-ray analysis and Xpert MTB/RIF assay cycle threshold (Ct) values.

Main Methods:

• Retrospective study of 230 adults with rifampicin-susceptible pulmonary tuberculosis across six South Korean centers (2019).
• Analysis included patient characteristics, AI-based tuberculosis extent scores from chest X-rays, and Xpert Ct values.

Main Results:

• 89.6% of patients achieved treatment success.
• AI-based radiographic tuberculosis extent scores significantly correlated with treatment success (OR 0.938) and culture conversion.
• Xpert Ct values did not significantly correlate with major treatment outcomes.

Conclusions:

• AI-based radiographic scoring at diagnosis is a significant predictor of treatment success and culture conversion.
• AI analysis of chest X-rays shows potential for personalized patient management in pulmonary tuberculosis.