CAD Detects Pulmonary Tuberculosis, Reduces Radiologist Workload
A retrospective paired diagnostic-yield study in Yichang, China (2022–2024) evaluated JF CXR-1 CAD versus radiologists on 93,319 chest X‑rays including 273 bacteriologically confirmed pulmonary tuberculosis (PTB) cases. CAD identified 83.9% of confirmed cases versus radiologists' 25.6% but had lower PPV (1.70% vs 10.31%), enabling an 85.5% reduction in radiologist workload through selective review. Authors recommend tiered AI prescreening with radiologist review for high-risk scores (>0.75), noting stronger CAD performance in township facilities.
Key Points
- 1Demonstrates CAD achieved 83.9% diagnostic yield among 273 bacteriologically confirmed PTB cases
- 2Highlights substantially higher sensitivity versus radiologists (25.6%), indicating improved detection in resource-limited settings
- 3Enables selective human review, cutting radiologist workload by 85.5% while preserving identified PTB cases
Scoring Rationale
Strong real-world evidence and operational gains, limited by single-region study and lower positive predictive value.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
