Opinion: How AI can rejuvenate imaging equipment to ramp up TB screening

In the fight against tuberculosis, early detection is crucial. Yet, millions of cases go undiagnosed every year, particularly in low- and middle-income countries.

While health care solutions powered by artificial intelligence have accelerated TB detection, scaling up screening remains a challenge in many LMICs due to infrastructure limitations — including shortage of radiologists and clinicians and lack of health infrastructure in remote regions. This is why instead of costly replacements and procurement, integrating AI into existing X-ray systems could offer a practical, cost-effective approach — enhancing efficiency, optimizing resources, improving detection rates, and overall broadening health care access.

The missed millions

Despite WHO's End TB Strategy for 2030, millions remain undiagnosed. Approximately 50% of TB cases are asymptomatic and slip through traditional screening efforts. AI-integrated screening can bridge this gap, strengthening active case detection and overall national surveillance by enhancing diagnostic accuracy, improving efficiency in high-burden settings, and supporting health care workers in low-resource environments.

For example, in India, where district hospitals play a crucial role in TB detection, integrating AI into routine chest X-ray screenings could significantly enhance case identification. With approximately 760 district hospitals across the country, even a conservative estimate of 50 chest X-rays per hospital per day translates to over 13 million chest X-rays annually. The National TB Prevalence Survey 2019-21 indicated that 42.6% of TB cases were identified because of upfront X-ray screening, independent of the presence of classical TB symptoms, underscoring AI’s potential in detecting additional cases through incidental TB screening.

This is also the case in other high TB-burden countries, such as Ethiopia in Africa, and the Philippines and Indonesia in Southeast Asia. If you take Indonesia’s case, an estimated 1.1 million people are contracting TB, out of which approximately 300,000 cases went undiagnosed. To close this detection gap, the government aims to screen 7 to 8 million people annually. The potential of an AI-powered screening pathway implemented across the existing health infrastructure could potentially transform TB detection at scale and help pick up exponentially more cases.

Overcoming the radiologist shortage with AI in Lesotho

Various countries are already implementing AI-driven TB screening and are witnessing the benefits firsthand. For example, Lesotho faces one of the world’s highest TB burdens, compounded by a severe shortage of radiologists. Recognizing the need for timely and accurate screening, the country’s Ministry of Health has integrated AI-powered chest X-ray solutions into its health care system.

According to Llang Bridget Maama-Maime, national TB program manager at the Ministry of Health in Lesotho, AI-powered solutions have allowed health care facilities to enhance diagnostic efficiency while continuing to use existing X-ray machines. With limited access to radiologists, integrating AI into existing imaging systems has enabled rapid, high-quality X-ray interpretations, reducing the dependency on human readers and improving TB detection. This approach optimizes available resources and ensures that diagnostic services remain accessible, particularly in resource-constrained settings.

Since implementation, over 6,000 scans have been processed in seven locations, identifying more than 1,100 potential TB cases, reinforcing how AI-powered imaging can help bridge diagnostic gaps.

Scaling AI with health worker training in Nigeria

When Nigeria first piloted AI-powered chest X-ray screening, it became evident that alongside deploying the technology, building the capacity of health care workers was critical for sustainability and strengthening the health system. Dr. Chukwuma Anyaike, director of public health at the Federal Ministry of Health in Nigeria, emphasized the importance of training, noting that equipping health care workers with the necessary skills helps minimize false positives and false negatives. He highlighted that rather than reinventing existing systems, the focus should be on optimizing current resources and ensuring continuous improvement in TB detection efforts.

Why AI-driven rejuvenation matters 

AI-powered imaging enables faster, more accurate TB screening in key and vulnerable communities, improving early detection and treatment initiation rates. AI solutions also enhance existing machines, increasing reach without additional hardware investments.

Findings from India’s health technology assessment study further validated the cost-effectiveness of AI-assisted TB screening.

The study found that one AI solution for chest X-ray screening has a favorable cost-effectiveness profile, with an incremental cost-effectiveness ratio, or ICER, well below India’s per capita GDP threshold, confirming its viability for national TB programs. ICER measures the additional cost required to interpret suspected TB cases compared to routine setups. Furthermore, AI-based screening was found to reduce costs per case while enhancing diagnostic efficiency.

Empowering health care providers 

AI integration also enables faster, more confident clinical decisions, even in facilities without radiologists, and has proven to provide scalability across multiple disease areas for comprehensive lung health management.

Moreover, studies have shown AI can differentiate between pneumonia and TB with up to 97.7% accuracy, surpassing the 79.9% accuracy achieved by physicians.

The way forward 

Despite AI’s proven benefits, many LMICs struggle to access and adopt global AI innovations. With WHO’s 2035 TB elimination targets fast approaching, countries cannot rely solely on traditional screening methods. Reviving existing X-ray infrastructure with AI is also proving an effective model for health system strengthening.

To fully leverage AI’s potential, however, continued investment, innovation, and collaboration are essential. By integrating AI-powered screening, LMICs can significantly expand TB detection capacity so that millions more access timely diagnosis and treatment. The question is no longer whether AI can transform TB care, but rather how quickly we can implement and scale these solutions to save more lives.

Visit Qure.ai for more information on AI in TB Care and Global Health