The study reveals a new way to assess an important measure of heart function

Coronary heart disease is the leading cause of death for adults worldwide. Performing coronary angiography provides a standard clinical diagnostic assessment for virtually all relevant clinical decisions, from medication to coronary bypass surgery. In many cases, quantification of the left ventricular ejection fraction (LVEF) at the time of coronary angiography is critical to improving clinical decision-making and treatment decisions, particularly when angiography is performed for potentially life-threatening acute coronary syndromes (ACS).

Because the left ventricle is the pumping center of the heart, measuring the chamber ejection fraction provides important information about the percentage of blood leaving the heart each time it contracts. Currently, measuring LVEF during angiography requires an additional invasive procedure called left ventriculography — in which a catheter is inserted into the left ventricle and contrast dye is injected — which carries additional risks and increases exposure to contrast.

In a study published May 10 in the heart gamma, senior author and UCSF cardiologist Jeff Tyson, MD, MPH, and first author Robert Avram, MD, of the Montreal Heart Institute set out to determine whether deep neural networks (DNNs) can be used, It is a class of artificial intelligence algorithm, for predicting heart pump (systolic) function from standard angiographic videos. They have developed and tested a DNN called CathEF to estimate LVEF from coronary vessel images of the left side of the heart.

CathEF introduces a new approach that takes advantage of data collected routinely during each angiogram to provide information that is not currently available to clinicians during angiography, effectively expanding the usefulness of medical data using AI and providing real-time LVEF information that informs clinical decision-making. “

Jeff Tyson, MD, assistant professor of medicine and cardiology at the University of California, San Francisco

The authors performed a cross-sectional study on 4,042 adult angiograms matched with corresponding transthoracic echocardiograms (TTEs) from 3,679 UCSF patients and trained a video-based neural network to estimate low LVEF (less than or equal to 40%) and to predict (continuous) a proportion of LVEF from Standard angiographic videos of the left coronary artery.

The results showed that CATHIF accurately predicted LVEF, with strong correlations with echocardiographic measurements of LVEF, which is the standard noninvasive clinical approach. The model was also externally validated in real-life vascular images from the Ottawa Heart Institute. The algorithm performed well across different patient demographics and clinical conditions, including acute coronary syndromes and different levels of renal function—groups of patients that may be less appropriate to receive a standard left ventricular diagram procedure.

“This study provides a new method for assessing LVEF, an important measure of heart function, during any routine coronary angiogram without the need for additional procedures or increased cost,” said Avram, MD, an interventional cardiologist and former research fellow at the University of California, San Francisco. “LVEF is essential for making decisions during the procedure and for managing patient care.”

Although the algorithm was trained on a large dataset of vascular images from the University of California, San Francisco, and then separately validated on a dataset from the Ottawa Heart Institute, the researchers are conducting further research to test this algorithm in the point of care. and determine their clinical impact. Workflow in patients with heart attacks. To this end, a multicenter validation study in patients with ACS is underway to compare the performance of cardiac and left ventricular catheterizations with TTEs performed within 7 days of ACS.

“This work demonstrates that AI technology has the potential to reduce the need for invasive testing and improve cardiologists’ diagnostic capabilities, ultimately improving patient outcomes and quality of life,” said Tyson.