Research News

Through advancements in artificial intelligence (AI) and health monitoring devices, a ring or a watch can help track important information about our physical health, including identifying difficult-to-detect patterns and predicting emerging risks. Now, researchers and clinicians are working together to discover how this technology can help our smallest patients.  

Animesh (Aashoo) Tandon, MD, MS, Pediatric Cardiologist, is exploring the possibility of using digital health technologies for infants with congenital heart disease (CHD). For years, wearables have promised clinicians continuous physiologic and activity data on high-risk adult patients, but with limited success. Now, Dr. Tandon and collaborators are combining wearables with other measures and cutting-edge analytic techniques to continuously gain deeper insights into infant health. 

How does digital health support clinical care? 

Vital signs are one of the most significant indicators of a patient's overall health. In high-risk patients, small changes in vital signs can indicate that something is starting to go wrong in the body. To collect data on a patient’s vital signs, clinicians usually rely on self-reported results, medical appointments and hospital stays. The problem with these controlled environments is that they do not show the whole picture, Dr. Tandon says. Clinicians cannot see how a patient’s vitals change as they go about their daily lives. 

Digital health solutions, including wearable technology, allow clinicians to see how patients' vitals fluctuate throughout their day. They can even detect incremental changes a patient or parent might not notice or know to report. This data is critical to clinician assessment of when a patient’s health may be at risk.  

How can digital health help infants? 

Dr. Tandon, Vice Chair for Innovation and Director of Cardiovascular Innovation at Cleveland Clinic Children’s, works with infants with CHD, which occurs when there is a problem with how the heart forms before birth. In all cases, normal blood flow through the heart is prevented. CHD is a term representing many forms of heart disease, but infants born with more complex forms of CHD are at high risk of suffering from heart failure, liver disease, kidney disease and blood clots.  

“When we are looking at potential complications from CHD, the best way to treat the complications is to catch them early,” Dr. Tandon says. “We monitor infants with CHD very closely, but it is difficult to predict when some of the complications are going to happen. With the large amounts of continuous data analyzed by new technologies, we have the potential to catch these issues much earlier and ensure their impact is not as bad.” 

Why are we only starting to explore wearable solutions for infants now?  

For nine years Dr. Tandon has wanted to explore how wearables can advance the care of infants, but companies only recently started developing devices suitable for children and infants. In parallel, there have been significant advances in analysis tools, including AI, to pull out the subtle features needed to understand health problems early. 

"It was only a year ago that the first few sensors got FDA clearance for monitoring oxygen saturation and heart rate in infants,” Dr. Tandon says. “The sensors we currently have require infants to be hooked up with wires to a machine, which leads to us only having intermittent monitoring data.” 

How will we determine if digital health can help infants with CHD?  

The first stage is focusing on the inpatient setting. When children arrive at Cleveland Clinic Children’s intensive care unit, clinicians have them wear hospital-grade monitors to capture an electrocardiogram and pulse oximeter waveforms. These devices help to measure vitals and display the results as waveforms and numbers that indicate functions, including heart rate and oxygen saturation.  

Through the Discovery Accelerator partnership, IBM researchers are helping Dr. Tandon and the multidisciplinary, collaborative team to combine wearable data with other data and sensor signals that they provide to custom AI to analyze the characteristics and shape of these waveforms to identify factors that can predict a patient's deterioration. These AI models have the potential to detect negative changes to vital signs faster and more accurately than current methods and may be able to predict emerging risks before they can be detected.  

Once the AI models are validated, the team will use them to help analyze data from infants who are at home (outpatients). The team will explore what vitals are best suited for wearables, what additional information is needed, and how often they need to be collected to identify potential issues.  

To support this work, Dr. Tandon was recently the inaugural awardee of the Johnson Family Innovation Award, a philanthropic award which aims to increase the impact of Cleveland Clinic and IBM’s Discovery Accelerator partnership. 

“Predicting major health events before they occur is critical for the long-term health of infants,” Dr. Tandon says. “With these wearable solutions we now have the potential to proactively provide treatments and prevent the worst effects of CHD.”