A lower-cost test could widen access to earlier cardiac screening
A study led by researchers at UT Southwestern Medical Center suggests that artificial intelligence can make one of medicine’s simplest heart tests far more useful in places where advanced imaging is hard to access. In work published in JAMA Cardiology, the team found that an AI system applied to routine electrocardiograms, or ECGs, accurately screened patients in Kenya for left ventricular systolic dysfunction, a major precursor to heart failure.
The finding matters because ECGs are relatively inexpensive and broadly available compared with echocardiograms, which are considered the gold standard for identifying this kind of underlying heart dysfunction. In many lower-resource health systems, access to echocardiography is limited by equipment costs, infrastructure, and specialist availability. That leaves many patients undiagnosed until heart failure becomes more advanced and harder to treat.
The new results point to a practical alternative: use a widely available test, then add AI analysis to identify which patients are most likely to need follow-up care. If validated and deployed at scale, that approach could help shift diagnosis earlier, when intervention may be more effective.
Why the study matters in sub-Saharan Africa
Heart failure is increasing globally, but the burden is especially severe in sub-Saharan Africa. According to the researchers, patients in the region often develop heart failure at younger ages and experience worse outcomes even though they may have fewer complicating conditions than patients in wealthier countries. That combination makes earlier detection especially important.
Before full heart failure develops, many patients first experience precursor conditions such as left ventricular systolic dysfunction. In that condition, the heart’s left ventricle does not pump blood effectively. Detecting it early can help clinicians intervene sooner, but that usually requires ultrasound-based heart imaging.
The UT Southwestern-led team argues that this is exactly the gap AI-ECG could help address. Rather than replacing echocardiography, the system could serve as a front-end screening layer in clinics and hospitals that cannot perform imaging on every patient. That would allow scarce diagnostic resources to be directed toward those at highest risk.
What researchers found
The source report describes the AI-augmented ECG analysis as accurately screening patients for underlying impairment in heart function in Kenya. The authors frame that performance as evidence that AI-ECG could become a scalable and affordable way to identify people at risk for heart failure in resource-limited settings.
Ambarish Pandey of UT Southwestern said the results support AI-ECG as a practical screening tool where access to echocardiography is constrained. That is an important distinction. The study is not presenting the ECG alone as a definitive replacement for more advanced imaging. Instead, it suggests that combining standard ECGs with AI interpretation can improve case finding in environments where the traditional diagnostic pathway is difficult to scale.
The research was led jointly with partners including Bernard Samia of M.P. Shah Hospital in Kenya and the Kenya Cardiac Society. That collaboration matters because the value of AI systems in medicine depends heavily on whether they perform in the health systems where they are intended to be used. Evidence from real-world use in Kenya gives the findings more relevance than a purely theoretical or lab-based validation would.
Why ECG plus AI is an attractive model
ECGs are already common in clinical care because they are fast, comparatively cheap, and simple to administer. Their weakness is that they do not directly provide the anatomical detail clinicians get from an echocardiogram. AI may help close part of that gap by detecting subtle electrical patterns associated with structural or functional heart problems that would otherwise be missed in routine interpretation.
That creates a potentially powerful model for emerging health systems. Instead of waiting for widespread expansion of expensive imaging capacity, providers may be able to improve screening by upgrading the intelligence applied to existing tests. In practical terms, that means software, workflows, and validation could become as important as new hardware in expanding care access.
It also aligns with a broader trend in medical AI: using algorithms not only in top-tier academic hospitals, but also in settings where the main benefit is triage, screening, and more efficient use of scarce expertise. If a routine ECG can help flag patients who most need echocardiography, specialist care can be focused where it will have the greatest impact.
What comes next
The study adds momentum to the idea that AI can make established medical tools more useful in lower-resource settings. But translation into care delivery will depend on several next steps, including validation across different patient populations, integration into clinic workflows, and evidence that screening leads to better treatment decisions and outcomes.
Even so, the core result is notable. Heart failure is often diagnosed too late, and the diagnostic bottleneck is frequently tied to cost and capacity rather than lack of need. By showing that AI-enhanced ECG interpretation can identify a key precursor condition in Kenya, the study offers a concrete example of how digital tools may help narrow that gap.
For health systems facing rising cardiovascular disease burdens, the appeal is straightforward: use what is already available, improve what it can reveal, and catch high-risk patients earlier. That does not solve every infrastructure problem in global cardiac care, but it may provide a realistic and scalable step forward.
This article is based on reporting by Medical Xpress. Read the original article.
Originally published on medicalxpress.com
