AI-Powered ECG Analysis Enables Early COPD Detection

Chronic obstructive pulmonary disease (COPD) is a major global cause of illness and death, yet it is often diagnosed late because early symptoms are vague and standard diagnostic tools such as spirometry can be resource-intensive. Delayed diagnosis limits opportunities for early intervention and worsens long-term outcomes. Researchers have now shown that a widely available, low-cost test routinely used for heart assessment can help identify COPD much earlier. Their findings demonstrate that subtle disease-related signals can be detected before a formal clinical diagnosis is made.

In the study conducted by Mount Sinai Health System (New York City, NY, USA), researchers developed a deep learning approach using a convolutional neural network to analyze standard 10-second, 12-lead electrocardiograms. Although ECGs are primarily designed to evaluate heart rhythm, COPD can cause structural and physiological changes in the heart that subtly alter ECG waveforms, changes that artificial intelligence (AI) models are able to detect.

The team trained and tested the model using ECGs extracted from the GE MUSE system, which stores raw waveform data as XML files. Data spanning 2006 to 2023 were analyzed from five Mount Sinai hospitals serving a demographically diverse population. Additional validation was performed using ECGs from an external hospital system and from patients with COPD enrolled in the UK Biobank to assess robustness across settings.

In total, more than 208,000 ECGs were analyzed, including records from over 18,000 patients with COPD matched to more than 49,000 controls by age, sex, and race. The model achieved strong diagnostic performance, with an area under the curve of 0.80 in internal testing, 0.82 in external validation, and 0.75 in the UK cohort. The results, published in eBioMedicine, were consistent across diverse populations.

Further analysis linked model predictions to spirometry measurements, while explainability methods highlighted P-wave changes as a key signal associated with COPD. These findings suggest that AI-powered ECG analysis could support earlier identification of at-risk patients, enabling earlier management and potentially slowing disease progression. The researchers note that this approach could be particularly valuable in remote or under-resourced settings where access to specialized pulmonary testing is limited.

“By demonstrating that AI can enhance the diagnostic utility of ECGs for COPD, a pathway is opened for earlier intervention and management of this disease, potentially reducing the severity of its progression and associated financial cost burdens. The use of such AI-enhanced diagnostic tools can be expanded to remote or under-resourced areas where access to specialized diagnostic facilities might be limited,” said Mount Sinai's Dr. Girish Nadkarni. “Additionally, this study lays the groundwork for future research into the integration of AI technologies with other routine diagnostic tools—possibly improving the diagnostic accuracy and timeliness for a range of chronic conditions and ultimately enhancing prevention and early intervention.”

Related Links:
Mount Sinai Health System