UCSF MedConnectionCan Artificial Intelligence Reduce Invasive Testing and Improve Cardiac Diagnostics?

The Evolution of Cardiac Care: AI’s Role in Enhancing Heart Function Assessment

In an era where coronary heart disease stands as the foremost cause of adult mortality worldwide, the quest for improving diagnostic precision and patient care quality has never been more critical. A study published on May 10 in JAMA Cardiology by a team from UCSF, led by cardiologist Geoff Tison, MD, MPH, and Robert Avram, MD, from the Montreal Heart Institute, unveils a transformative approach leveraging artificial intelligence (AI) to predict left ventricular ejection fraction (LVEF) from standard coronary angiogram videos. This innovation proposes a significant shift in cardiac diagnostics, aiming to optimize clinical decision-making and patient management without the need for additional invasive procedures or increased costs.

Key Points:

• Coronary heart disease is the leading cause of adult death globally, with left ventricular ejection fraction (LVEF) being a critical measure in diagnosing and managing the condition.
• Traditionally, measuring LVEF during angiography involves an invasive procedure known as left ventriculography, posing additional risks and contrast exposure to patients.
• The study introduces CathEF, a deep neural network (DNN) designed to estimate LVEF from coronary angiograms of the left heart, eliminating the need for additional procedures.
• CathEF was developed through a cross-sectional study of 4,042 angiograms matched with transthoracic echocardiograms (TTEs) from 3,679 UCSF patients, aiming to accurately predict reduced LVEF (≤40%) and continuous LVEF percentage.
• Results demonstrated that CathEF could predict LVEF with strong correlations to echocardiographic measurements, offering a noninvasive, cost-effective alternative for real-time clinical decision-making.
• The algorithm was externally validated using real-world angiograms from the Ottawa Heart Institute, showcasing its effectiveness across diverse patient demographics and clinical conditions, including acute coronary syndromes and varying levels of renal function.
• Further research is ongoing to test CathEF at the point-of-care and evaluate its impact on clinical workflow and patient outcomes, particularly in those suffering from acute coronary syndromes (ACS).

“This work demonstrates that AI technology has the potential to reduce the need for invasive testing and improve the diagnostic capabilities of cardiologists, ultimately improving patient outcomes and quality of life.”
– UCSF Cardiologist Geoff Tison, MD, MPH

More in Cardiology