A team of researchers at UT Southwestern Medical Center has found that pairing a standard electrocardiogram with artificial intelligence can reliably catch early signs of heart failure in patients in Africa.
The study tested AI-augmented electrocardiogram analysis – known as AI-ECG – across eight health care facilities and hospitals in Kenya, and nearly 6,000 patients.
Currently, the world’s leading cause of death, heart disease, continues to claim more lives each year with no signs of slowing down. And that burden falls especially hard on sub-Saharan Africa, where patients tend to develop the disease at younger ages and face worse outcomes, even though they typically present with fewer complicating conditions than patients in wealthier countries.
A key reason for those worse outcomes is the difficulty of catching pre-existing heart conditions before they become full-blown heart failure.
According to UTSW, many patients first develop “left ventricular systolic dysfunction” – a state in which the heart’s main pumping chamber loses power. The standard tool for diagnosing this is an echocardiogram, which uses ultrasound to produce images of the heart. But echocardiograms can be expensive, and the equipment and expertise required to perform them are largely unavailable across much of Africa.
A problem that the UTSW team believes AI could help fix.
“These findings support AI-ECG as a practical, scalable screening tool that can effectively identify individuals at risk for heart failure in resource-limited settings where access to echocardiography is constrained, addressing a critical gap in global cardiovascular care,” said Dr. Ambarish Pandey, associate professor of Internal Medicine at UT Southwestern.
A standard ECG measures the heart’s electrical activity and costs a fraction of what an echocardiogram does. The AI layer added to the ECG in this study searches the results for patterns associated with dysfunction and other precursors to heart failure – essentially squeezing more information out of a test that hospitals and clinics already have on hand.
Among the 1,444 patients who received both an AI-ECG and a confirmatory echocardiogram, the algorithm identified left ventricular systolic dysfunction in around 14% of cases. Its negative predictive value reached 99.1%, meaning that when the AI found no sign of dysfunction, echocardiography confirmed that finding in nearly every case.
Additionally, the algorithm correctly identified over 95% of patients with precursor conditions and accurately cleared over 79% of those without.
If validated at scale, AI-ECG could provide clinicians in under-resourced settings with an affordable way to identify patients who need closer cardiac monitoring before a silent condition becomes fatal.
A Look At AI In Medicine
As previously covered by The Dallas Express, AI is already changing how doctors work in the United States, helping health systems run more efficiently and catch problems earlier.
A DX report found that most Dallas-area women were open to AI helping read their mammograms – but only with a radiologist overseeing the process. That study, also out of UT Southwestern, found that nearly 74% of patients wanted to be told if AI was involved in their care, and more than 80% had at least one concern about the technology, ranging from data privacy to accuracy.
However, implementing AI in healthcare hasn’t been entirely seamless.
Texas’ medical and legal communities have been watching carefully how AI performs in high-stakes medical environments. Attorney General Ken Paxton secured a settlement in 2024 against an AI healthcare company after his office found it had made deceptive claims about the accuracy of its systems, potentially jeopardizing patient safety – a reminder that promising AI results must be matched by honest, transparent reporting.
“AI companies offering products used in high-risk settings owe it to the public and to their clients to be transparent about their risks, limitations, and appropriate use. Anything short of that is irresponsible and unnecessarily puts Texans’ safety at risk,” Paxton said at the time.
