Mayo Clinic researchers have developed and externally validated an artificial intelligence model that can detect potentially significant obstruction of blood flow from routine heart ultrasound videos, without requiring specialised Doppler imaging. The findings could help clinicians identify patients who may need additional testing for hypertrophic cardiomyopathy.
The technology focuses on left ventricular outflow tract (LVOT) obstruction, a condition that can occur in people with hypertrophic cardiomyopathy (HCM), a genetic disorder in which the heart muscle becomes abnormally thick. Mayo Clinic says about two-thirds of people with HCM develop LVOT obstruction, which can restrict blood leaving the heart and contribute to symptoms including chest pain and shortness of breath.
The study evaluated 1,833 patients in the Mayo Clinic cohort. Researchers tested the model in 275 patients and then externally validated it using data from 46 patients at a hospital in South Korea. The AI system analysed resting, non-Doppler ultrasound videos and used three standard ultrasound views to improve its ability to distinguish patients with elevated LVOT gradients.
Researchers also found that the model could identify obstruction that might emerge when the heart is under stress. In a subset of cases, its performance in identifying the condition exceeded that of two expert echocardiographers reviewing the same non-Doppler images.
The model is not intended to replace Doppler echocardiography. Instead, it could serve as an early screening tool to flag patients for confirmatory Doppler measurements, stress testing or referral to an HCM specialty centre.
The researchers said the technology may also have applications with portable ultrasound systems and in settings where comprehensive Doppler assessment is not readily available. However, further prospective validation is planned across different clinical environments, ultrasound platforms and patient populations.
The study, published in Circulation: Cardiovascular Imaging, received no external funding.
Source: This article is based on an official press release issued by Mayo Clinic
Mayo Clinic AI model helps clinicians detect heart obstruction using routine ultrasound images – Mayo Clinic News Network
