Lung Cancer Breath Test Uses AI to Detect Volatile Biomarkers

A portable breath-based cancer screening tool developed at the University of Texas at Dallas is showing strong potential for early detection of thoracic cancers. By analyzing volatile organic compounds (VOCs) in exhaled breath and applying machine learning, the device can identify cancer-linked chemical signatures with 90 percent accuracy in initial clinical tests. This approach could make noninvasive cancer screening more accessible, especially in primary care settings.

The sensor is designed to detect eight specific VOCs that are known to correlate with cancer metabolism. These compounds are present in trace amounts and require highly sensitive detection methods. The biosensor uses screen-printed electrodes and nanomaterials to capture and measure VOC concentrations in real time, offering a noninvasive alternative to traditional diagnostic procedures.

To interpret the complex data generated by the sensor, the team developed machine learning models that classify breath samples based on their chemical profiles. The AI component enables the system to distinguish between healthy and cancerous breath signatures with minimal false positives, making it suitable for early screening applications.

The study involved 67 participants, including 30 patients with biopsy-confirmed thoracic cancer. Breath samples were collected and analyzed using the sensor, and the results were compared to clinical diagnoses. The system correctly identified cancer-related VOC patterns in 90 percent of confirmed cases. Researchers emphasized that the test could be administered during routine checkups, offering a low-cost method for early referral and follow-up care.

The COVID-19 pandemic served as a catalyst for this innovation, highlighting the value of rapid, noninvasive diagnostics. Breath analysis offers a window into metabolic changes and disease states, and the UT Dallas team is contributing to the emerging field of breathomics. Their sensor is designed to be low-cost, portable, and scalable, with potential applications in routine checkups, population-wide screening programs, and even remote healthcare settings.