A standardized accelerometry method for characterizing tremor: Application and validation in an ageing population with postural and action tremor

Etienne Gauthier-Lafreniere; Meshal Aljassar; Vladimir V Rymar; John Milton; Abbas F Sadikot

doi:10.3389/fninf.2022.878279

A standardized accelerometry method for characterizing tremor: Application and validation in an ageing population with postural and action tremor

Front Neuroinform. 2022 Aug 4:16:878279. doi: 10.3389/fninf.2022.878279. eCollection 2022.

Authors

Etienne Gauthier-Lafreniere^{1

2}, Meshal Aljassar¹, Vladimir V Rymar¹, John Milton³, Abbas F Sadikot¹

Affiliations

¹ Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University Health Centre, McGill University, Montreal, QC, Canada.
² Department of Psychiatry, Montreal Neurological Institute, McGill University Health Centre, McGill University, Montreal, QC, Canada.
³ W.M. Keck Science Department, Claremont Colleges, Claremont, CA, United States.

Abstract

Background: Ordinal scales based on qualitative observation are the mainstay in the clinical assessment of tremor, but are limited by inter-rater reliability, measurement precision, range, and ceiling effects. Quantitative tremor evaluation is well-developed in research, but clinical application has lagged, in part due to cumbersome mathematical application and lack of established standards.

Objectives: To develop a novel method for evaluating tremor that integrates a standardized clinical exam, wrist-watch accelerometers, and a software framework for data analysis that does not require advanced mathematical or computing skills. The utility of the method was tested in a sequential cohort of patients with predominant postural and action tremor presenting to a specialized surgical clinic with the presumptive diagnosis of Essential Tremor (ET).

Methods: Wristwatch accelerometry was integrated with a standardized clinical exam. A MATLAB application was developed for automated data analysis and graphical representation of tremor. Measures from the power spectrum of acceleration of tremor in different upper limb postures were derived in 25 consecutive patients. The linear results from accelerometry were correlated with the commonly used non-linear Clinical Rating Scale for Tremor (CRST).

Results: The acceleration power spectrum was reliably produced in all consecutive patients. Tremor frequency was stable in different postures and across patients. Both total and peak power of acceleration during postural conditions correlated well with the CRST. The standardized clinical examination with integrated accelerometry measures was therefore effective at characterizing tremor in a population with predominant postural and action tremor. The protocol is also illustrated on repeated measures in an ET patient who underwent Magnetic Resonance-Guided Focused Ultrasound thalamotomy.

Conclusion: Quantitative assessment of tremor as a continuous variable using wristwatch accelerometry is readily applicable as a clinical tool when integrated with a standardized clinical exam and a user-friendly software framework for analysis. The method is validated for patients with predominant postural and action tremor, and can be adopted for characterizing tremor of different etiologies with dissemination in a wide variety of clinical and research contexts in ageing populations.

Keywords: Essential Tremor; Parkinson; accelerometry; clinical rating scale; focused ultrasound (MRgFUS); power spectrum; sensor; thalamotomy.