Wearable technology has been used to remotely monitor patient activity in cardiac rehabilitation programs, but as volumes of data grow, artificial intelligence (AI) will play a bigger role in monitoring these patients and pulling out actionable insights. Research by the NYU Grossman School of Medicine was presented at the American Heart Association (AHA) 2023 meeting on this topic, which showed the AI could easily pick out phenotype step-count trajectories in older adults.
"AI can look at really fine-grain data. So think about the Fitbit. It actually collects data on a minute-by-minute level, and after 90 days, it kind of becomes a humongous data set. But what AI can do is, it can look at all that data. We looked at step count trajectories, so you have a baseline physical activity level as measured by step counts on day one. And then the AI looked at the trajectory of how many steps they take every day evolved over time," explained Souptik Barua, PhD, assistant professor with the department of medicine at NYU Grossman School of Medicine, in an interview with Cardiovascular Business.
Data for this study came from the Rehabilitation Using Mobile Health for Older Adults With Ischemic Heart Disease in the Home Setting (RESILIENT) trial led by John Dodson, MD, MPH, associate professor in the Department of Medicine at NYU Grossman School of Medicine. The study was supported by a National Institute of Aging (NIA) grant 1R01AG062520-01.
Unveiling wearable activity tracker insights through AI
Ideally, the numbers of steps should go up over the course of time, but researchers found the AI could identify different exercise patterns to better classify the types of activity and compliance each patient had. These fell into into four specific groups.
"For example, what we saw was that while around 32 of 75 individuals that we analyzed did show an increased incline showing improvement, we also found roughly around 20% of the participants whose physical activity was kind of steady," Barua said. They also saw groups that started out with a decrease and then an increase in activity, and another where they started out strong and then activity decreased.
He highlighted AI's role in handling the massive amount of data collected by wearables and avoiding the need for human researchers or physicians to perform a time-intensive manual sift through extensive data to pull out insights. The AI algorithms can analyze minute-by-minute data collected over a span of three months, offering a comprehensive understanding of patients' physical activity in minutes.
Personalized insights for rehabilitation planning
Barua emphasized the importance of these distinct patterns in guiding rehabilitation strategies. He said the AI can show adherence to an exercise regimen and help show patient engagement levels. Using these personalized insights, patients showing a decline might require additional support or adjustments in their rehabilitation plans.
He said wearables and AI provide nuanced insights into patients' activities at home throughout the rehabilitation period. This contrasts with traditional methods, where in-person assessments were conducted sporadically, offering only snapshots of progress.
"These patterns can help us determine for someone whose usage has dropped right from the start of the intervention, just giving them a Fitbit would not be the right way to go. Maybe there are some extra incentives or more relaxed usage protocols that might benefit them better," Barua explained.
The future of cardiac rehabilitation
Barua believes that precision medicine, facilitated by rich wearable data and AI, will play a pivotal role in the future of cardiac rehabilitation. The vision involves a continuous refinement of rehabilitation strategies based on personalized insights. The iterative process allows for the optimization of mobile health-based cardiac rehabilitation, creating a more effective and tailored approach for each individual.
Dave Fornell has covered healthcare for more than 17 years, with a focus in cardiology and radiology. Fornell is a 5-time winner of a Jesse H. Neal Award, the most prestigious editorial honors in the field of specialized journalism. The wins included best technical content, best use of social media and best COVID-19 coverage. Fornell was also a three-time Neal finalist for best range of work by a single author. He produces more than 100 editorial videos each year, most of them interviews with key opinion leaders in medicine. He also writes technical articles, covers key trends, conducts video hospital site visits, and is very involved with social media. E-mail: dfornell@innovatehealthcare.com