AI helps cardiologists deliver personalized healthcare—but there is still plenty of work to do

The rapid rise of artificial intelligence (AI) has helped cardiologists, radiologists, nurses and other healthcare providers embrace precision medicine in a way that ensures more heart patients are receiving personalized care. AI models can use electronic health record (EHR) data to spot early signs of infection, for example, or identify patients who may benefit from at-home cardiac monitoring.

But this is just the beginning when it comes to using AI and machine learning (ML) to improve patient care, according to a new scientific statement from the American Heart Association (AHA).[1] Published in Circulation, the paper examines both the progress  made and the untapped potential of these technologies.

“Despite enormous academic interest and industry financing, AI-based tools, algorithms and systems of care have yet to improve patient outcomes at scale,” wrote writing group chair Antonis A. Armoundas, PhD, an associate professor of medicine with Harvard Medical School, and colleagues.

Armoundas et al. added that they hoped to “identify best practices, gaps and challenges that may improve the applicability of AI tools.”

How advanced AI is already improving care for heart disease patients

AI and ML technologies have already gone a long way toward improving the diagnosis, treatment and management of cardiovascular disease.

In cardiac imaging, for instance, advanced algorithms aim to improve accuracy, decrease acquisition times, reduce radiation exposure and even estimate patient outcomes in the worlds of cardiac CT, echocardiography, cardiac MRI and nuclear imaging.

Entry way to the RSNA artificial intelligence showcase. Photo by Dave Fornell. #RSNA #RSNA23 #RSNA2023 orthopedic imaging

The artificial intelligence showcase at RSNA 2023 in Chicago.

AI and ML are also improving care for stroke patients.

“AI/ML applications on CT of the head can automatically detect early ischemic changes of the brain, without the need for diffusion-weighted MRI,” the writing group explained. “AI/ML algorithms have improved quantification of CT or MR brain perfusion imaging and enhanced their ability to predict recovery of cerebral function during the time taken to transport patients for reperfusion therapies. Other applications include neurointerventional planning for the management of acute ischemic stroke and cerebral aneurysms, and for patients recruitment in clinical trials for acute stroke.”

Finding the appropriate patient data to use, however, represents a challenge for care teams working to use AI in cardiac imaging. Obtaining the necessary data is hard enough, but then specialists often find that the data they want to use are unstructured and unlabeled.

AI and ML have already “dramatically affected” the use of electrocardiography (ECG), according to the writing group. AI models can detect structural heart issues in ECG results, for instance, which is crucial as the need for high-quality interpretations continues to grow.

Validating AI ECG models in a way that limits bias does remain a challenge, though, and “the limited availability of digitized and well-labeled electrocardiographic data and open-source datasets may limit research and development of AI/ML algorithms.”

AI models have made cardiac care more precise by improving the monitoring of heart patients in the hospital and at home. Algorithms can track hospitalized patients for signs of clinical deterioration, sepsis, hypotension, cardiac arrest, atrial fibrillation and even drug-related proarrhythmia, the writing group noted. In addition, AI-powered consumer wearables can help cardiologists monitor patients for a wide variety of complications. These devices open “specific ethical issues” because private patient data are often being stored and shared using smartphones and web-based applications; going forward, this is a growing concern among providers and patients alike that will need to be addressed.

Tracking and evaluation of EHR data is another potential use of AI in today’s healthcare space. It’s still early for such algorithms, but research teams around the world are working to make this a reality.

“In principle, appropriate analysis of the EHR could improve disease detection, stratify patients into treatable disease types (novel 'phenotypes') and identify novel clinical workflows … On the other hand, AI/ML applied to EHR could simulate sequential decision-making at different time points, enrolling every patient who has been treated or not treated with little exclusion criteria and with less patient dropout,” according to the writing group.

Data found in EHRs “are only as good as their curation and consistency,” the authors warned. Without clean and accurate information, in other words, even the most advanced algorithm will struggle to make accurate predictions.

What care teams needs to improve AI implementation

Hospitals and health systems must be careful when integrating advanced algorithms into day-to-day patient care, Armoundas and colleagues explained. 

“Robustly designed AI/ML systems can identify informative and hidden patterns in complex clinical data to personalize cardiovascular medicine from screening and diagnosis, to find novel classification and phenotypes, to predict adverse outcomes, to guide therapy and to guide trial design,” they wrote. “AI/ML should augment and support clinical decision-making rather than replace clinical judgement needed for evidence-based practice. However, to realize this potential, AI/ML analytics must be presented to clinicians through intuitive and interpretable human-computer interfaces that enhance user trust and integrate with existing clinical workflows.”

Health equity, bias, fairness, public trust, liability and cybersecurity are some of the factors that must be considered when implementing advanced AI and ML models. Cybersecurity is an especially strong concern in today’s healthcare landscape.

“Theft of medical records allows access to financial services and healthcare for criminals,” the group wrote.  

Read the full AHA scientific statement for more

The 23-page statement represents the views of both the AHA as well as several subgroups, including its Institute for Precision Cardiovascular Medicine, Council on Cardiovascular and Stroke Nursing, Council on Lifelong Congenital Heart Disease and Heart Health in the Young, Council on Cardiovascular Radiology and Intervention, Council on Hypertension, Council on the Kidney in Cardiovascular Disease and Stroke Council.

Click here to read the full document in Circulation.

Michael Walter
Michael Walter, Managing Editor

Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.

Around the web

Ron Blankstein, MD, professor of radiology, Harvard Medical School, explains the use of artificial intelligence to detect heart disease in non-cardiac CT exams.

Eleven medical societies have signed on to a consensus statement aimed at standardizing imaging for suspected cardiovascular infections.

Kate Hanneman, MD, explains why many vendors and hospitals want to lower radiology's impact on the environment. "Taking steps to reduce the carbon footprint in healthcare isn’t just an opportunity," she said. "It’s also a responsibility."

Trimed Popup
Trimed Popup