A new research paper has revealed how new medical devices, combined with advanced machine learning techniques, could offer a promising alternative to traditional spirometry for diagnosing Chronic Obstructive Pulmonary Disease (COPD).
The paper, developed by TidalSense, incorporated data from five clinical studies conducted in collaboration with Portsmouth Hospitals University NHS Trust, the NIHR Community Healthcare MedTech, IVD Cooperative at the University of Oxford, Modality GP Partnership and uMed.
Interpretable machine learning was used to diagnose COPD and assess severity using 75-second carbon dioxide (CO2) breath records captured in these studies with TidalSense’s N-Tidal capnometer.
According to the research, the approach achieved impressive results in both COPD diagnosis and severity assessment, highlighting its potential to transform diagnostic capabilities within healthcare practices, particularly in resource-strained primary care environments.
Dr Ameera Patel, CEO of TidalSense, said: “Healthcare systems around the world are in urgent need of alternative diagnostic tools to address significant operational challenges in diagnosing COPD with spirometry which has resulted in a large undiagnosed population and late-stage diagnosis.
“We’re hugely encouraged by the study’s findings, which demonstrate the viability of our approach as an accurate, fast and accessible alternative to traditional spirometry, especially in primary care settings.”
‘Diagnosis and Severity Assessment of COPD using a Novel Fast-Response Capnometer And Interpretable Machine Learning’ set out to examine whether TidalSense’s technology, combined with machine learning, could offer a fast, reliable and precise alternative diagnostic test to traditional spirometry, which is technique-dependent, non-specific, and requires administration by a trained healthcare professional.
The study, which will be published in COPD: Journal of Chronic Obstructive Pulmonary Disease, concluded that the N-Tidal device could be used alongside interpretable machine learning as an accurate, point-of-care diagnostic test for COPD, particularly in primary care, as a rapid and accurate rule-in or rule-out test.
Machine learning algorithms were trained using data from 999 patients, including 294 with varying COPD severities and 705 without the condition.
The innovative machine learning model accurately differentiated COPD patients from non-COPD individuals. Notably, it achieved a positive predictive value (PPV) of 93 per cent for high-confidence predictions, suggesting its potential for rapid and reliable COPD diagnosis at the point of care.
The study also developed a model that accurately distinguished between GOLD 1 (mild) and GOLD 4 (very severe) COPD with an AUC of 0.980. This ability to assess severity could be further developed to help clinicians tailor treatment plans and, in the future, monitor disease progression more effectively.
The N-Tidal device, when used alongside the algorithms developed in this paper, could offer several benefits compared to conventional spirometry methods, including:
-
High usability: Requires only 75 seconds of normal tidal breathing, eliminating the need for forceful exhalation, which can be difficult for some patients and can cause an aerosol-generating cough.
-
Rapid: Results are available within minutes, allowing for faster diagnosis and treatment initiation (c.f. Spirometry which requires an appointment of between 15 and 45 minutes).
-
Accessible: The portable and easy-to-use device can be deployed in various healthcare settings, including primary care clinics.
-
Potentially cost-effective: The technology is simpler, has a shorter training time and does not require a specialist clinician to operate it. For these reasons, along with its significantly shorter testing time, it could reduce healthcare costs compared to spirometry.
This pivotal study suggests that the N-Tidal could significantly alleviate the burden on existing healthcare systems which are routinely struggling to provide spirometry in primary care settings.
By offering a faster, more accessible, and potentially more cost-effective alternative, N-Tidal could lead to earlier diagnosis, improved disease management, and better patient outcomes.
Professor Anoop Chauhan, Group Chief Research Officer at Portsmouth Hospitals University NHS Trust, said: “This research has the potential to be a game-changer for COPD diagnosis.
“The ease of use and rapid results offered by this new approach could significantly improve access to COPD diagnosis, particularly in primary care settings where access to spirometry can be limited.
“Early and accurate diagnosis is crucial for effective COPD management, and this technology holds great promise for improving patient outcomes.”
