AI Tool Surpasses Radiologists in Diagnosing Pulmonary TB: Study
The tool, ULTR-AI, surpassed World Health Organization (WHO) targets for non-sputum-based TB triage, showing 93% sensitivity and 81% specificity.
An AI-based lung ultrasound tool has outperformed human experts in diagnosing pulmonary tuberculosis (TB), according to a study presented at ESCMID Global 2025.
The tool, ULTR-AI, surpassed World Health Organization (WHO) targets for non-sputum-based TB triage, showing 93% sensitivity and 81% specificity.
The study was conducted at a tertiary urban center in Benin, West Africa. Of the 504 patients included after exclusions, 192 (38%) were confirmed to have pulmonary TB. Among them, 15% were HIV-positive, and 13% had a previous history of TB. A standardized 14-point lung ultrasound scan protocol was used, and human experts interpreted the results.
The reference standard was the MTB Xpert Ultra molecular sputum test.
Addresses Diagnostic Gaps in High-Burden Regions
The AI suite analyses ultrasound images from smartphone-connected portable devices, offering a rapid, sputum-free diagnostic method. Its use could be especially relevant in high-burden, resource-limited settings where access to radiologists and chest X-ray equipment is limited.
“Early screening and rapid diagnosis are critical components of the WHO's ‘End TB Strategy,’ yet many high-burden countries experience substantial patient dropouts at the diagnostic stage due to the high cost of chest x-ray equipment and a shortage of trained radiologists,” the study noted.
Adding insights, Dr. Véronique Suttels, lead study author, said, “The ULTR-AI suite leverages deep learning algorithms to interpret lung ultrasound in real-time, making the tool more accessible for TB triage, especially for minimally trained healthcare workers in rural areas. This technology can help diagnose patients faster and more efficiently by reducing operator dependency and standardizing the test.”
AI Integration Enables Point-of-Care Diagnosis
The ULTR-AI system includes three models: one that directly predicts TB from images, one that detects human-recognized patterns, and another that combines both to maximize diagnostic accuracy.
“Our model detects human-recognizable lung ultrasound findings—like large consolidations and interstitial changes—but an end-to-end deep learning approach captures even subtler features beyond the human eye,” said Dr. Suttels. “We hope this will help identify early pathological signs such as small sub-centimetre pleural lesions common in TB.”
She added, “A key advantage of our AI models is the immediate turnaround time once they are integrated into an app. This allows lung ultrasound to function as a true point-of-care test with good diagnostic performance at triage, providing instant results while the patient is still with the healthcare worker. Faster diagnosis could also improve linkage to care, reducing the risk of patients being lost to follow-up.”
