US researchers develop algorithms to monitor Parkinson’s tremors

31 October 2019 (Last Updated October 31st, 2019 11:14)

Researchers from Florida Atlantic University’s College of Engineering and Computer Science have collaborated with the Icahn School of Medicine and the University of Rochester Medical Center to develop machine-learning algorithms to track tremors in patients suffering from Parkinson’s Disease.

US researchers develop algorithms to monitor Parkinson’s tremors
With this method, patients with Parkinson’s disease can be monitored at home and elsewhere and provide clinicians with vital information to effectively manage and treat their patients with this disorder. Credit:Florida Atlantic University

Researchers from Florida Atlantic University’s College of Engineering and Computer Science have collaborated with the Icahn School of Medicine and the University of Rochester Medical Center to develop machine-learning algorithms to track tremors in patients suffering from Parkinson’s Disease.

Around seven to ten million people worldwide suffer from Parkinson’s, the second-most age-related neurodegenerative disorder.

The most significant problem in Parkinson’s is tremor, an involuntary movement disorder that affects the daily activities of  patients.

Typically, neurologists measure the tremors using the Unified Parkinson’s Disease Rating Scale (UPDRS).

However, UPDRS is an onsite physical examination method that offers only a glimpse of the tremors that patients experience in their everyday activities.

FAU department of computer and electrical engineering and computer science assistant professor Behnaz Ghoraani said: “A single, clinical examination in a doctor’s office often fails to capture a patient’s complete continuum of tremors in his or her routine daily life.

“Wearable sensors, combined with machine-learning algorithms, can be used at home or elsewhere to estimate a patient’s severity rating of tremors based on the way that it manifests itself in movement patterns.”

Researchers have investigated the application of two machine-learning algorithms, gradient tree boosting and LSTM-based deep learning.

Results from the study showed that the gradient tree boosting method estimated total tremor and resting tremor sub-score with high accuracy and in majority cases with the same results determined through the UPDRS.

Patients also experienced a decline in tremors after taking medication, even in cases where results did not match with the total tremor sub-scores recorded through the UPDRS assessments.

Out of all existing UPDRS task-dependent methods and task-independent tremor estimation methods, this process is claimed to have provided the highest performance results.

Meanwhile, in a separate development, Medtronic has launched its advanced Patient Programmer technology for Deep Brain Stimulation (DBS) therapy at the Samsung Developers Conference in San Jose, California.