Dr. Shalini Prasad, a leading researcher and professor in the bioengineering department at UT Dallas, emphasized the breakthrough potential of this technology, stating that it may enable clinicians to detect lung cancer during its initial, more treatable stages. The research aims to establish a quick, affordable, and noninvasive screening tool that utilizes breath analysis, which could significantly improve patient outcomes and aid in the timely management of thoracic cancers.

Notably, the electrochemical biosensor developed by the research team is capable of detecting eight specific VOCs associated with thoracic cancers. After testing this device on breath samples from 67 patients—including 30 with biopsy-confirmed thoracic cancer—the researchers achieved an impressive success rate of accurately identifying the VOCs in 90% of confirmed cancer cases. This high level of accuracy demonstrates the potential efficacy of using breath analysis as a diagnostic tool in cancer screening.

The origins of this project closely align with global health challenges raised during the COVID-19 pandemic. At that time, there was an urgent need to explore noninvasive technologies that could assist in the rapid screening and isolation of virus transmission. Dr. Prasad noted that leveraging breath analysis was compelling due to the connection between respiratory metabolites and potential indicators of disease, showcasing the clinically relevant insights derived from human breath.

The proposed technology falls within the emerging field of breathomics—a discipline focusing on the analysis of compounds present in exhaled breath to diagnose diseases and monitor various health conditions. The significant variation in metabolites in breath can signal early disease onset, positioning this research, particularly when augmented by AI, as a vital complementary approach to traditional diagnostic methodologies.

Artificial intelligence plays an integral role within the framework of this research, as Dr. Prasad highlighted the complex data produced by breath analysis. The challenge lies in discerning which data points are clinically significant and which are not. Machine learning algorithms contribute to this filtering process, emphasizing the importance of interdisciplinary collaboration with computer science experts to develop effective analytical models that enhance diagnostic capabilities.

Collaboration was a cornerstone of this research endeavor, as Dr. Prasad worked alongside Dr. Ovidiu Daescu, a computer science expert who assisted in refining the machine learning models and validating the technological approach. The interdisciplinary teamwork harnesses the strengths of bioengineering and computational methodologies, ensuring that the developed breath profiling device is robust and ready for clinical application.

The implications of such a device are promising, with the potential to transform cancer detection practices in the medical field. Early detection of lung cancer remains a critical concern, as it stands as the leading cause of cancer-related mortality both in the U.S. and globally. By utilizing minimally invasive technologies such as breath-analysis, the research team aims to institute methods for early detection of thoracic malignancies while minimizing the patient burden associated with traditional diagnostic procedures.

Looking ahead, Dr. Prasad expressed the team’s commitment to further advancing the technology, specifically seeking more extensive clinical validation. She envisions a future where routine breath tests could be integrated into standard primary care visits, alongside traditional blood tests, allowing healthcare providers to offer proactive recommendations based on patients’ breath biomarker profiles.

This push towards making breath analysis a mainstream diagnostic tool encapsulates an ethos of leveraging cutting-edge research to enhance patient care—transforming how diseases are detected and monitored in everyday healthcare settings. By moving beyond traditional methodologies, this research signifies a critical step toward integrating innovative technologies within clinical practices.

Key contributions to this research project were also made by doctoral student Nikini Subawickrama, first author Dr. Anirban Paul, and several other scholars from both UT Dallas and the UT Southwestern Medical Center. Their collective efforts affirm the significant collaboration required to pioneer new biomedical technologies that can reshape the landscape of disease diagnosis and patient management.

As research in this field continues to evolve, the potential for electrochemical breath profiling—especially when coupled with artificial intelligence—offers a forward-thinking approach to cancer detection that bridges technological innovation with pressing healthcare needs. Continued exploration and validation of these methods could lead to more effective screening options, ultimately saving lives through timely diagnosis and intervention.

This groundbreaking development not only holds promise for lung cancer detection but could also extend to other health conditions, emphasizing the versatility and potential impact of breath analysis research. As scientists continue to unlock the secrets of breathomics, we stand at the threshold of a new era in disease detection and management, driven by the confluence of engineering, computer science, and medicine.

Subject of Research: Biosensor technology for cancer detection
Article Title: Electrochemical breath profiling for early thoracic malignancy screening
News Publication Date: 1-Aug-2025
Web References: DOI
References: Sensing and Bio-Sensing Research
Image Credits: University of Texas at Dallas

Keywords

Bioengineering, Health and medicine, Cancer, Lung cancer, Artificial intelligence, Machine learning, Breath analysis, Biosensors.

Colorectal cancer remains a formidable health challenge, ranking as the fourth most prevalent malignancy in the United Kingdom. While survival rates drastically improve with early diagnosis, late-stage discovery yields a dismal five-year survival of merely 10%. The insidious nature of CRC symptoms, often vague or nonspecific, complicates timely detection and complicates referral decisions for colonoscopies, which, despite being the gold standard, carry logistical and patient compliance issues.

Breath analysis offers a compelling solution that hinges on detecting CRC-specific VOCs emitted through the respiratory system. These organic compounds, byproducts of tumor metabolism or host-tumor interactions, create distinct chemical fingerprints that gas chromatography–mass spectrometry (GC-MS) can discern with high sensitivity. The COBRA2 protocol establishes a rigorous framework to collect, analyze, and interpret these VOC profiles to enhance early CRC detection accuracy.

The study design incorporates a total enrollment of 720 participants, meticulously divided into two cohorts: 470 control subjects scheduled for colonoscopy with no CRC diagnosis, and 250 patients confirmed to have colorectal adenocarcinoma through histological examination. This case-control setup enables the researchers to contrast VOC patterns robustly and develop predictive algorithms that differentiate cancerous cases from non-cancer controls.

To ensure the integrity of breath samples and minimize confounding variables, participants adhere to a clear fluid diet for a minimum of 4–6 hours before sample collection. Sampling occurs at outpatient clinics, intentionally avoiding bowel preparation that could alter VOC signatures. This methodological attention to detail heightens the reliability of VOC data and solidifies the foundation for the ensuing machine learning analyses.

The analytical phase employs advanced gas chromatography–mass spectrometry, a technique that systematically separates and identifies the myriad VOCs within each breath specimen. By quantifying these compounds, researchers aim to pinpoint specific VOC profiles or molecular signatures that correlate strongly with CRC presence, distinguishing them from benign conditions or healthy states.

A pivotal facet of the study is the integration and comparative assessment of the faecal immunochemical test (FIT), a widely used non-invasive screening tool that detects occult blood in stool samples. Researchers intend to evaluate whether combining FIT results with breath VOC data enhances the diagnostic power beyond each modality alone, potentially refining screening accuracy and reducing false negatives.

After initial model development, the COBRA2 framework entails an independent validation phase with up to 250 participants split evenly between controls and CRC cases. This step tests the model’s generalizability and predictive reliability in a fresh cohort, an essential process to affirm the clinical value and reproducibility of the breath test in varied settings.

Exploratory statistical and machine learning techniques play crucial roles in model building. These methods sift through complex, multidimensional VOC data to identify patterns and relationships that human analysis might overlook. Machine learning algorithms offer adaptive, data-driven prediction tools that can evolve with expanding datasets and clinical insights, paving the way for precise, personalized cancer screening strategies.

The ultimate goal is to craft decision rules that support frontline healthcare providers in triaging patients efficiently. A breath test that accurately flags high-risk individuals could streamline referrals for colonoscopy, reduce patient burden, and optimize resource allocation within healthcare systems. By detecting CRC earlier, this approach holds promise not just for survival improvement but also for enhancing the quality of life through less invasive diagnostics.

The COBRA2 initiative’s relevance extends beyond its immediate clinical implications. Breath analysis technology harnesses cutting-edge biomarker science, metabolomics, and analytical chemistry, symbolizing a broader shift toward non-invasive diagnostics in oncology. This represents a paradigm change where molecular signatures replace or augment tissue biopsies and imaging, ushering in an era of precision medicine driven by accessible technology.

ClinicalTrials.gov registration (NCT05844514) formalizes this study in the international research landscape, ensuring transparency, adherence to rigorous protocols, and facilitating prospective participant engagement. This registration also enables real-time monitoring of milestones and dissemination of forthcoming results that could influence global screening guidelines.

The breath test’s patient-centered advantages cannot be overstated. Avoiding bowel preparation and invasive endoscopic procedures reduces physical discomfort and psychological stress, thereby may improve patient compliance and screening uptake. In public health contexts where CRC burden is significant, such innovations could substantially impact screening participation rates and downstream outcomes.

If successful, COBRA2’s predictive model will invite further validation in more heterogeneous, unselected symptomatic populations. Real-world application demands testing beyond controlled case-control cohorts to understand performance amidst clinical variability, comorbidities, and population diversity, shaping practical integration into routine healthcare.

Moreover, the prospect of combining breath VOC analysis with established screening tools like FIT illustrates a forward-thinking, multimodal diagnostic landscape. By layering orthogonal biomarkers, clinicians gain a richer, more nuanced decision-making framework, balancing sensitivity and specificity that might otherwise be unattainable with single tests alone.

In closing, the COBRA2 breath testing study epitomizes translational research at its best — transforming a scientific discovery in molecular signatures into a feasible diagnostic tool with the potential to change cancer outcomes. The integration of biochemical innovation, computational analytics, and clinical validation exemplifies a multidisciplinary endeavor poised to reshape colorectal cancer detection and perhaps inspire similar strategies across oncology disciplines.

Subject of Research: Non-invasive breath testing for early detection of colorectal cancer using volatile organic compound analysis

Article Title: Non-invasive breath testing to detect colorectal cancer: protocol for a multicentre, case–control development and validation study (COBRA2 study)

Article References:
Fadel, M.G., Murray, J., Woodfield, G. et al. Non-invasive breath testing to detect colorectal cancer: protocol for a multicentre, case–control development and validation study (COBRA2 study). BMC Cancer 25, 1230 (2025). https://doi.org/10.1186/s12885-025-14520-2

Image Credits: Scienmag.com

DOI: https://doi.org/10.1186/s12885-025-14520-2