Machine Learning Model Helps Predict Clinical Lab Test Results
Using wearable device data from smartwatches, machine learning may enable physicians to predict health measurements that otherwise would require invasive tests.
Source: Getty Images
By Jill McKeon
- Applying machine learning to wearable device data could help predict clinical laboratory measurements without a visit to the doctor’s office, a new study published in Nature Medicine reveals. The long-term data collection that wearable devices enable provide a more holistic view of a patient’s health.
“There is a circadian (daily) variation in heart rate and in body temperature, but these single measurements in clinics don’t capture that natural variation,” said Jessilyn Dunn, one of the study’s lead investigators, in a May 24th press release.
“But devices like smartwatches or Fitbits have the ability to track these measurements and natural changes over a prolonged period of time and identify when there is variation from that natural baseline.”
The study cohort consisted of 54 patients, studied over three years. Patients wore a smartwatch that collected data on heart rate, skin temperature, sweat gland activation, and movement. Each patient also participated in regular doctor’s visits, where their vitals were taken and tests administered.
Researchers found that the smartwatch data provided important insights into a patient’s overall health. If a patient consistently had low sweat gland activation, it signaled that they were dehydrated. In comparing watch-based measurements to clinic-based measurements, researchers concluded that the wearable device data largely matched up with clinical test results and could help predict abnormalities.
“If you think about someone just showing up in an emergency room, it takes time to check them in, to get labs going, and to get results back,” Dunn explained in the press release.
“But if you were to show up in an ER and you’ve got an Apple Watch or a Fitbit, ideally you’d be able to pull the long-term data from that device and use algorithms to say, ‘this may be what’s going on.’”
While vital signs taken at a clinic provide a snapshot of the patient’s health at that given time, smartwatch data may actually help physicians diagnose or predict health issues. In addition, the results found that measurements like red and white blood cell count, measured in a clinic, was correlated with the smartwatch data. For example, a higher body temperature and low movement recorded on the smartwatch lined up with a higher white blood cell count result from the blood test.
“This experiment was a proof-of-concept, but our hope for the future is that physicians will be able to use wearable data to immediately get valuable information about the overall health of a patient and know how to treat them before the clinical labs are returned,” Dunn said in the press release. “There is a potential for life-saving intervention there if we can get people the right care faster.”
There has been a recent rise in both machine learning applications in healthcare and the use of wearable devices. Stanford Medicine and Fitbit recently partnered to study a wearable device’s ability to track infectious diseases among college athletes. In addition, a recent study revealed the use of a predictive analytics model that can predict where and how quickly a disease outbreak can spread. This research shows promise for a future with a focus on predictive analytics and better health outcomes.
