NIH Funding to Support Cardiovascular Disease Detection Using AI
Financial support from the NIH will allow researchers to enhance cardiovascular disease detection through machine- and deep-learning models.
Source: Getty Images
By - After receiving $6.2 million in funding from the National Institutes of Health (NIH), researchers from Case Western Reserve University and University Hospitals Cleveland Medical Center (UH) aim to enhance cardiovascular disease detection using artificial intelligence (AI) techniques and computed tomography (CT) calcium scoring.
The Centers for Disease Control and Prevention (CDC) noted that in 2020, about 697,000 people in the US died from heart disease. Conditions related to heart disease are also the leading causes of death for men, women, and most racial groups in the US, the CDC noted.
Using two grants from the NIH totaling $6.2 million, UH and Case Western Reserve University researchers aim to determine how AI approaches, particularly machine- and deep-learning models, can help clinicians in predicting cardiovascular disease.
According to the press release, the funding is based on the CLARIFY Registry. This UH effort provides community members with risk factors for heart disease with CT scan calcium score assessments, which are accurate predictors of cardiovascular issues.
“With the help of AI and machine-learning, these grants will help usher a new era of predictive health analytics to automate risk assessment and, more importantly, empower patients to seek care to reduce their cardiovascular risk. The multidisciplinary team that we have assembled here between engineering and medicine is absolutely essential to solve difficult problems,” said Sanjay Rajagopalan, MD, chief of the division of cardiovascular medicine, chief academic and scientific officer of UH Harrington Heart & Vascular Institute, and co-leader of the study, in the news release.
CT calcium scoring occurs through a CT scan of a patient’s chest, allowing researchers to collect information regarding the level of calcium in coronary arteries. This data can then assist professionals in making well-informed decisions regarding suggestions for treatment, which may vary from prescribing aspirin to cholesterol-lowering drugs.
The new grants will support research examining all aspects of CT scan-obtained calcium data, including calcifications in the heart and fat deposits around the heart, to enhance predictions of cardiovascular disease.
In 2017, Rajagopalan, Daniel Simon, MD, and Robert Gilkeson, MD, began a free CT calcium scoring program at UH. They found that the program raised patient awareness regarding heart risks, encouraging self-care and the use of medicines.
“The clinical colleagues at University Hospitals were foresighted in developing the CT calcium score program and the CLARIFY registry, which was beneficial for the clinical community,” said David Wilson, PhD, the Robert Herbold Professor of Biomedical Engineering (BME) at Case Western Reserve, and co-leader of the study, in the news release. “We have a substantial amount of data that has been gathered throughout the years that AI can examine.”
Using AI to enhance the detection and treatment of various conditions is growing more common.
In August, for example, researchers from the AI in Medicine Program of Brigham and Women’s Hospital created a deep-learning algorithm to improve lung cancer radiation therapy treatment. Through this AI model, researchers aimed to identify non-small cell lung cancer tumors within CT scans.
After training and testing the model, they asked radiation oncologists to perform segmentation processes. They then asked these physicians to rate and edit segmentations produced by other providers or the AI model.
They concluded that there were no major differences in performance between segmentations produced through human-AI collaborations and human-only segmentations.
