Role of data measurement characteristics in the accurate detection of Parkinson's disease symptoms using wearable sensors.
Nicholas Shawen, Megan K O'Brien, Sanjeev Venkatesan and 6 others
PMID 32312287WHAT IT FOUND
Wearable sensors can detect Parkinson's tremor and slowness using simpler data.
A 30 Hz sampling rate and basic features captured symptoms as well as high-resolution settings, allowing devices to run longer without losing accuracy.
Key findings
01A 30 Hz sampling rate was sufficient to detect both tremor and bradykinesia without significant loss of accuracy compared to higher rates.
02Accelerometer data alone, such as from a smart watch, was sufficient to detect tremor, but detecting bradykinesia required combining accelerometer and gyroscope data.
03Using features based only on signal magnitude achieved similar performance to using all three axes, reducing the computational load required for real-time monitoring.
STILL TO COME
How it was doneWhat they foundWhat it means for PTsWhat it means for OTsWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
The study included only 13 participants, which is a small sample size for training and validating machine learning models. The data was collected during standardized tasks in a clinical setting, so it is unclear if the models would perform as well during unstructured, free-living activities at home. The study focused only on upper limb symptoms (tremor and bradykinesia) and did not test other Parkinson's symptoms like gait or rigidity. The authors used a specific machine learning approach (random forest), and results may differ with other algorithms or feature sets.
Declared interests
The study was supported by the Michael J. Fox Foundation for Parkinson's Research and the National Institutes of Health. The authors did not declare any competing financial interests.
The easy way to misread this
Do not assume these findings apply to all Parkinson's symptoms or real-world monitoring. The study only tested upper limb tremor and bradykinesia in a controlled clinical setting, so accuracy may differ for gait, rigidity, or daily activities at home.