Researchers Use mHealth to Track, Assess Parkinson's Movements

mHealth researchers are reporting that wearables combined with an AI platform can help healthcare providers track the severity of Parkinson’s disease.

As reported in a recent issue of Nature, a team of researchers from IBM, the Pfizer Innovation Research Lab, Boston University’s Spivack Center for Clinical & Translational Bioscience and Tufts Medical Center conducted a study using wearable devices and an AI algorithm. They found that the connected health platform could be used to measure the severity of tremors – which, in turn, could help providers adjust care management, particularly medication use.

Their study compared the actions of 45 healthy participants to 35 patients living with Parkinson’s, who were given a wrist-worn accelerometer and asked to perform certain hand and walking movements. Researchers compared the movements and broke them down into segments and sequences that could be measured and analyzed. An AI algorithm was then used to identify movements affected by Parkinson’s and rate the severity of the symptoms.

In addition, 26 patients living with Parkinson’s were given a GeneActive mHealth wearable, which they were asked to wear at home, to ensure that data captured at home compares with date recorded in a clinic environment.

“By transforming continuous signals from wearables into a statistical distribution of movement syllables (the symbolic movement representation) we captured the increasing disorder in motion between healthy and disease state,” the researchers noted. “This statistical representation also correlated with increased motor impairment among people with PD as quantified by their MDS-UPDRS (the Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale) scores.”

Researchers needed about 10 minutes of activity to create a database that could be used to effectively measure Parkinson’s severity,

According to Parkinson’s News Today, the study is part of the Blue Sky Project, a collaborative effort by IBM Research and Pfizer to create AI algorithms that will improve clinical trials. Data from three separate studies was included in the Parkinson’s project to ensure that information gathered from mHealth devices in the home setting were as accurate as lab-based studies.

The connected health platform fine-tuned by the study could someday be used for research that goes beyond Parkinson’s.

“We have shown how the stationary representation of movement syllables captures the overall quality of motor behavior,” the researchers concluded. “The proposed method differentiates motor impairment accompanying various PD related states both in clinical settings when scripted activities are performed, and in unconstrained everyday life. The proposed model was explicitly tested on movement differences associated with PD but we see no reasons why it should not be generalizable to detecting other neurological states with characteristic movement signatures.”