Revolutionizing TB Diagnosis: AI’s Breakthrough in Chest X-ray Detection

Revolutionizing TB Diagnosis: AI's Breakthrough in Chest X-ray Detection.

AI software has demonstrated the ability to accurately detect tuberculosis (TB) from chest X-rays with promising results. This is what a study that will be presented at the European Congress of Clinical Microbiology and Infectious Diseases (ECCMID) in Copenhagen, Denmark, from April 15th to the 18th shows.

Tuberculosis (TB) is a big reason why people die and get sick all over the world. It causes 1.6 million deaths a year, making it is the 13th leading cause of death globally and the second biggest infectious killer, after COVID-19.

In low-resource settings, chest X-rays play an important role in the diagnosis of patients unable to produce good-quality sputum samples for microbiological analysis. Using software to look for problems on X-rays could help with diagnosis in places where there aren’t enough doctors.

However, there is a lack of good quality studies assessing its diagnostic accuracy, as highlighted recently by the World Health Organization (WHO).

In AI-assisted chest X-rays, India has a powerful technology to screen for presumptive TB. The AI algorithm qXR, developed by Mumbai based, can help detect people with presumptive TB early and in less than a minute.

With an expected 3 million undiagnosed patients in 2021, it is important to come up with new strategies and tools to improve the way TB is found in places with few resources and a high number of cases.
We’ve shown that AI software is at least as good at finding TB as a trained doctor, and that all it takes for analysis is a simple shot from a cell phone.
In places with few resources and a lot of tuberculosis but not enough doctors, chest X-rays could be taken with a cell phone and sent to an AI system for analysis.
This would make it easier to read more chest X-rays and, most importantly, find more cases of TB.

— Dr. Frauke Rudolf, Department of Infectious Diseases, Aarhus University Hospital, Aarhus, Denmark

Systematically looking for TB to find it early is an important part of the “End TB” plan. A few months ago, the Indian agency in charge of drugs approved qXR. qXR also meets the WHO requirement with >90% sensitivity and >70% specificity in people older than 15 years.

WHO’s ambitious goal of “eliminating” TB by 2025 will remain possible only if early diagnosis and initiation of care for millions of people with TB becomes a reality. Large scale use of AI assisted chest X-rays for screening is the first step to achieve this goal.

In Vietnam, a communitywide screening of people older than 15 years using a molecular test in 2014-2017 resulted in lower prevalence of pulmonary TB in 2018 than standard passive case detection alone.

Unlike in Vietnam, the use of qXR to read digital X-rays before molecular testing as part of community screening will reduce TB prevalence and minimize the number of molecular tests required to detect TB.

To improve applicability in low-resource settings, the AI was given mobile phone photographs of analog (non-digital) CXRs.

Chest X-rays from 498 patients were analyzed retrospectively. Fifty-seven (11%) of these patients had been diagnosed with TB, 41 clinically and 16 through PCR tests (Xpert MTB/Rif).

The AI software was as good as or better than a trained radiologist at identifying the PCR-confirmed cases. It correctly identified 75% of all PCR-confirmed cases (sensitivity of 75%) and 85.7% of non-TB cases (specificity of 85.7%).

The less experienced radiologist’s assessments had a sensitivity of 62.5% (they correctly picked up 62.5% of the PCR-confirmed cases) and a specificity of 91.7% (they correctly identified 91.7% of those who didn’t have TB).

The experienced radiologist’s assessments were 75% sensitive and 82.0% specific.

The radiologists didn’t agree too much on the data, and neither did the program or all of the radiologists together.

Several AI algorithms and machine learning models have been developed and trained on large datasets of chest X-ray images to identify patterns and abnormalities associated with TB. These AI systems can assist healthcare providers in the screening and diagnosis of TB in a more efficient and accurate manner.

Here are some key points of Chest X-rays to consider:

High Accuracy: AI-powered systems have achieved high levels of accuracy in detecting TB from chest X-rays. Some studies have reported sensitivity and specificity rates that are comparable to or even surpass those of human radiologists.

Speed: AI algorithms can analyze X-ray images quickly, potentially reducing the time it takes to provide a diagnosis. This speed can be especially valuable in regions with a high TB burden, where prompt detection and treatment are crucial.

Consistency: AI systems offer consistent performance regardless of the time of day or the workload of healthcare professionals, reducing the risk of human error in diagnosis.

Enhanced Screening: These AI tools can be used for mass screening programs in areas with a high prevalence of TB, helping to identify cases early and prevent further transmission of the disease.

Resource Optimization: By automating the initial screening process, AI can help allocate healthcare resources more efficiently, ensuring that patients who require further evaluation and treatment are prioritized.

Accessibility: AI-powered TB detection can be deployed in remote or underserved areas where access to trained radiologists is limited, improving healthcare access and outcomes.

While AI holds great promise in TB detection, it’s important to note that these systems should be used as supportive tools for healthcare professionals rather than replacements. AI systems should undergo rigorous validation, and their integration into clinical practice should adhere to regulatory and ethical guidelines to ensure patient safety and the reliability of results. Additionally, ongoing human oversight and expert interpretation of AI-generated findings are essential to confirm diagnoses and make treatment decisions.

Read More

The Impact of Missed Childhood Tuberculosis. Cases on Global TB Control Goals