Dr. Charley Taylor

Lessons from the pandemic: How embracing AI can revolutionize healthcare

October 12, 2021
By Charley Taylor

There are times when medical innovations are slow to take root. When the COVID-19 pandemic hit, all attention was focused on getting the virus in check and caring for those afflicted, and out of fear, lack of medical staff capacity, and quarantine protocols, people ignored other aspects of their health. As a result, there was then an urgency to figure out ways to care for patients in acute, non-COVID related instances, like heart attacks, strokes, and everyday maintenance of chronic conditions like diabetes, high blood pressure and cancer care. The need to treat patients while decreasing their and the medical staff’s exposure to the virus, rejuvenated interest in advancing solutions like telemedicine and had physicians thinking differently about the tools at hand - pushing technologies that were often used as second or third options to the forefront.

The key to caring for patients outside of COVID-19 infections was to help physicians determine when a hospital visit was necessary. If a visit or procedure was necessary, getting patients in and out as quickly as possible while providing their physicians with the most accurate information for a treatment path was a primary concern.

30% of all patient visits during the first six months of the pandemic were provided via telemedicine. That’s a jaw-dropping statistic for a healthcare tool that was hardly used pre-pandemic. Telehealth is helping to close gaps in access to care for many populations and should remain a mainstay of modern healthcare, and it’s refreshing to see that policymakers are working to secure expanded coverage for telehealth after the pandemic. Treating patients from afar does present challenges and in some scenarios is simply impossible; yet telehealth has proliferated, from a simple virtual check-up to complex robotic surgery performed through remote access.

While this is all incredibly exciting, it is not without its own hurdles. Alignment between clinical and administrative leaders is needed to advance beyond the status quo; doing so will help to save countless lives and lift the burdens on healthcare providers while helping to reduce enormous costs. Trust in medical professionals to handle data responsibly will also remain a key issue.

During COVID-19, heart disease related deaths increased to nearly 700,000 cases in the U.S. in 2020, primarily due to the decrease in people not seeking medical care. Medical professionals struggled to find efficient and accurate ways for stress testing. Stress tests are the most commonly used heart test to diagnose coronary artery disease (CAD), but they have an alarmingly low rate of accuracy. Based on the uncertain findings of a stress test, research shows 20 to 30% of patients with CAD are falsely reassured that they do not have significant disease, and 55% end up having invasive procedures, which in hindsight are unnecessary. Stress testing is outdated and its performance is simply unacceptable for diagnosing the world’s number one killer.

During the pandemic, some medical professionals are proactively moving from the traditional stress test to a cardiac CT scan first approach. Using cardiac CT scans, patients can now be diagnosed with CAD within a matter of hours and within one hospital visit. Initial clinical visit discussions can be conducted virtually and the CT scan, which takes approximately 90 minutes at the hospital and requires minimal patient interaction with the clinical team, provides the physician the data they need to rule out CAD. Additionally, after ruling out CAD, physicians are able to investigate other cardiovascular concerns (e.g., myocarditis).

For patients with CAD, layering the CT scan with AI-powered technology enables clinicians to obtain additional actionable patient information from a brief, non-invasive cardiac test. There are AI-enabled algorithms which accurately analyze cardiac CT images and provide physicians with information about the extent to which a blockage is limiting blood flow. This information helps physicians determine a definitive treatment plan, and identify which patients can be managed medically, and which require more invasive treatment.

Navigating questions such as clinical appropriateness, patient impact and financial value will be vital to the adoption of potentially life-saving, AI-enabled services. Strong clinical evidence demonstrates that innovations using advanced technology can help deliver better patient outcomes and lower healthcare costs. Not only that, but using hard, clear data will allow more open conversations between patients and physicians, to help communicate and improve understanding surrounding diagnosis and treatment. It also helps to address historic biases that have kept many patients from reaching out to receive care.

All of this is to say that COVID-19 has necessitated our acceptance of some technologies faster than ever before to better care for patients. Hopefully this experience will help increase adoption of future technologies quicker, allowing us to live healthier, happier lives moving forward.

About the author: Dr. Charley Taylor is a co-founder, chief technology officer (CTO), and member of the board of directors of HeartFlow Inc. Previously, he was an associate professor in the Department of Bioengineering and Surgery at Stanford University with courtesy faculty appointments in the Departments of Mechanical Engineering and Radiology, and is internationally recognized for the development of computer modeling and imaging techniques for cardiovascular disease research, device design and treatment planning.