For people with a structural problem in their heart but no symptoms, detecting the disorder has been challenging. However, a researcher at the Yale School of Medicine has developed a way to identify these seemingly invisible heart problems using artificial intelligence (AI).
Dr. Rohan Khera, clinical director of the medical school’s Center for Health Informatics and Analytics, has found that AI can analyze a simple electrocardiogram (ECG) and detect what a basic reading cannot. Khera’s research was published in the journal Circulation.
Left ventricular systolic dysfunction is a structural heart disorder that reduces the heart’s ability to pump blood. It often develops before symptoms arise, leading to an increased risk of heart failure and premature death. According to Khera’s study, there are effective and affordable treatments available for this disorder.
Traditionally, cardiac imaging techniques such as an ultrasound or MRI are used to diagnose heart disorders. However, these tests are not feasible for widespread use in the community. With the use of AI and deep learning, Khera’s team has developed a technology that can identify signatures of structural heart disorders from ECG data.
ECGs are routinely done during physical exams and can now even be taken on wearable devices like an Apple Watch. With approximately 100 million ECGs done in the United States each year, analyzing these data using AI has the potential to identify individuals with heart disorders accurately.
The Yale study has been validated in multiple locations, including California, Missouri, Texas, and a Brazilian longitudinal study. Khera’s approach to identifying heart disease has shown promising results.
Using computer vision algorithms, Khera and his team process ECG photos to identify areas within the heart that may suggest a diagnosis of low heart function. AI can analyze the structural abnormalities in the heart that human clinicians may not be able to detect.
As more advancements are made in AI and ECG technology, screening for heart disorders using ECGs is expected to become more common. The capability of AI to identify heart conditions that humans may miss is an exciting development in the field of cardiology.
