A recent study conducted by Eko Health has unveiled an innovative tool that utilizes artificial intelligence (AI) to detect heart diseases. This AI-powered tool, available on the Sensora platform, combines Eko’s digital stethoscopes with an AI algorithm trained to recognize […]
A recent study conducted by Eko Health has unveiled an innovative tool that utilizes artificial intelligence (AI) to detect heart diseases. This AI-powered tool, available on the Sensora platform, combines Eko’s digital stethoscopes with an AI algorithm trained to recognize structural heart murmurs that may indicate valvular heart disease. Approved by the FDA last year, this AI technology goes a step further than simply identifying signs of abnormal heart sounds by analyzing their timing and severity to differentiate between benign murmurs, absent murmurs, and murmurs occurring during systole and diastole.
The Sensora platform also includes Eko’s software for care pathway analysis, allowing clinicians to monitor the progress of their patients throughout the healthcare system.
According to the study published in the journal Circulation, and presented at the scientific sessions of the American Heart Association, Eko discovered that this AI tool significantly improves the detection of heart murmurs and thus the diagnosis of valvular heart disease.
The study involved a total of 369 patients over the age of 50 who had never been diagnosed with valvular heart disease or heart murmurs. Each patient was examined using a standard analog stethoscope and one of Eko’s digital stethoscopes equipped with the AI algorithm.
After comparing the results of each examination with echocardiogram data, it was found that the stethoscope with artificial intelligence could identify signs of valvular heart disease with a sensitivity of 94%, compared to just over 41% for the standard stethoscopes. In total, the AI tool identified 22 patients with “moderate or more severe” cases of the disease that had previously gone unnoticed, while the analog method only detected eight new cases.
The results were less pronounced in terms of specificity, or the classification of negative cases of valvular heart disease. In this metric, Eko’s digital device achieved slightly below 85%, while the non-AI device achieved over 95%.
The study authors concluded that Eko’s method “demonstrated a significant impact on the detection of valvular heart disease compared to conventional practice” and proposed that integrating this technology into routine primary patient examinations can improve diagnosis rates, patient treatment pathways, and overall outcomes.
Eko argues that these improvements are indeed necessary, citing data showing that about half of all adults over the age of 65 have some form of valvular heart disease. While the majority of these cases are mild, with only around 10% of this age group carrying clinically significant cases of the disease, valvular heart disease can rapidly progress if left undiagnosed and untreated due to the frequent absence of symptoms or the nonspecific nature of the symptoms. This can lead to heart failure, stroke, and even death.
“The consequences of undiagnosed or late-diagnosed valvular heart disease are serious and costly to our healthcare system,” said Dr. Mose Rancier, the lead author and principal investigator of the study, in a statement by Eko. “This study shows that patients can be more effectively assessed for valvular heart disease in primary care using AI-based technology.”
In the future, researchers plan to continue enrolling study participants to gather more evidence and will also follow patients for 12 months after the initial examinations to assess their clinical outcomes.