Thought Leadership

Machine learning is coming for our hearts: How AI can help the health care industry fight heart disease

We have screening guidelines for diseases like breast and colon cancer, and they help save many lives. We need a concerted effort to pursue an evidence-guided path for cardiovascular disease screening.
July 28, 2022

Recent breakthroughs in algorithm-guided medical technologies and the accessibility of point-of-care devices are challenging the status quo of cardiac care. Proactively screening for cardiovascular disease has the potential to become a standard practice that can identify millions of patients with silent heart disease.

Cardiovascular disease continues to be the leading cause of death in America. Yet, there aren’t concrete national guidelines for cardiovascular disease screening, even though asymptomatic people would likely benefit from early intervention with existing treatments. Evidence has consistently shown that early detection plays a role in preventing heart disease from progressing further and provides more time for intervention.

However, variability in clinician training, workflows, and reimbursement could make knowing when and where to screen patients challenging. For cardiac screening to become standard practice in primary care and in the community, well-informed guidelines and standards should be created. Clinical guidelines established by national health care societies substantially impact how health care is practiced, and without them, adoption may be inconsistent.

We have screening guidelines for diseases like breast and colon cancer, and they help save many lives. We need a concerted effort to pursue an evidence-guided path for cardiovascular disease screening.


Screening plays an important role in detecting health conditions in people with non-specific or no symptoms. However, screening has a mixed reputation. Many recognize the challenges of detecting disease early but may be skeptical about its actual benefits.

The prevention of heart disease is focused mainly on lifestyle modifications, with a more narrow focus on proactive heart screenings. While reasons for the absence of concrete cardiac screening guidelines are nuanced, much of it can point to an insufficient body of research exploring the harms and rewards of cardiovascular screening.

For many, heart disease care means getting annual bloodwork to check LDL cholesterol or triglycerides. And while routine blood work can provide some insight into one’s risk of heart disease, it’s unable to identify if there is disease present. Many people have high cholesterol but haven’t developed heart disease.

Testing to identify real structural or rhythm issues with the heart is highly variable. Large gaps in cardiovascular care emerge when looking at patients by age, race, gender, and income. While more focused on detecting structural heart abnormalities, the traditional stethoscope is also limited, as medical professionals make decisions based on what they can hear. Many clinicians lack confidence in distinguishing clinically significant heart murmurs from healthy sounds.

Costs and time constraints often disincentivize cardiovascular screening in common settings, like the annual physical exam, if there’s no reason to suspect disease. A common way to detect cardiovascular disease is with an echocardiogram, an expensive, time-consuming test requiring trained technicians and specialists to capture and interpret the data. To test every person over 65 would be an unfeasible burden. Cardiology, like much of the U.S. health care system, is reactive. As such, many cardiologists are trained to treat acute heart events, not how to manage the disease when it’s moderate or mild.

While there are barriers to screening, accurate AI technology in portable devices creates new opportunities to overcome them.


In just the last decade, advancements in medical technology have been profound. Medical-grade devices and equipment and consumer products like smartwatches, wearables, and smartphones are opening new opportunities for reimaging health care. Many of these devices are powered by machine-learning algorithms fueled with vast amounts of data to improve the accuracy and accessibility of disease detection.

The emergence of these portable tools provides a cost-effective way to screen patients with reliable results. Because these devices are more affordable, fast, and portable, they are ideal for the primary care office, community screening event, or pharmacy, unlike the traditional invasive diagnostic tests. These products can make the annual physical or office visit a powerful cardiovascular disease screen.

The pace of AI in disease detection is continuing to increase. Large medical record datasets are being cleaned and processed to help algorithms spot new patterns and correlations. Google’s AI can help prevent blindness with the early detection of diabetic retinopathy. Clinicians continue to find new insights hidden in ECG signals, including detecting pulmonary hypertension, heart murmurs, and weakened heart pumps. Other algorithm efforts are being deployed in imaging departments around the world to improve the detection of abnormalities.

In addition, a consumer-focused shift in health care is playing out in the rising popularity of retail health clinics. These clinics are deeply interested in improving workflows and patient outcomes through technology, while offering care models that appeal to consumers, like cardiac screenings.


National and society guidelines for cardiovascular screenings should be the first step in helping standardize how these new technologies can improve the early detection of heart disease. Guidelines ultimately can influence payers and spur the creation of reimbursement codes. Integrating payment models can also ease the transition to using these technologies in the primary care setting. By leveraging portable cardiovascular devices, a physician or nurse during a routine visit can easily and affordably identify early signs of disease, resulting in more time for patients to change the course of their disease. The possibilities are profound.

According to the CDC, a person dies every 36 seconds from cardiovascular disease. Despite all the recent advances in therapeutics and interventions, heart disease deaths are increasing. Clearly, the lack of coherent screening guidelines isn’t due to a lack of need. And the pace of innovation will not slow down. I believe breakthrough therapies and innovations in cardiovascular care will undoubtedly continue, and the health care system will wrestle with how to implement these new workflows and technologies.

New tides of digital innovations in cardiovascular-disease care are getting into clinicians’ hands. Still, the healthcare industry is underprepared to create thoughtful recommendations for how these should be used to screen patients. With a disease so prevalent in our society and patients’ lives on the line, action on the development of national cardiac-screening guidelines is vital.

This article originally appeared on July 28, 2022 on Fast Company.