Novel Objective Tool to Assess Tremor Reveals Unilateral Focused Ultrasound Improves Tremor Bilaterally

Abstract

Background: Tremor in essential tremor and in tremor-dominant Parkinson's disease is assessed by subjective observations in patients undergoing focused ultrasound thalamotomy, a minimally invasive procedure intended to alleviate tremor in these patients.

Objective: To develop an objective tool for tremor analysis to be used before and after focused ultrasound thalamotomy treatment in the treated hand (contralateral to ablation) and non-treated (ipsilateral to ablation).

Methods: Using image processing and signal processing that utilized images of a Archimedes spiral drawing, we created a tool to analyze tremor. First, we showed that the proposed tool reproduces known clinical dynamics on the treated hand, and then we used it to evaluate the clinical dynamics on the non-treated hand.

Results: Using the tool we developed, we were able to demonstrate a significant reduction in tremor following focused ultrasound thalamotomy among 132 essential tremor and 26 tremor-dominant Parkinson's disease patients in the treated hand using drawings of Archimedes spirals up to 1 year following the procedure. Thus, we reproduced known clinical data and therefore validated the proposed tool. In addition, we were able to demonstrate a significant improvement in the non-treated hand as well as a significant deterioration in the efficacy of focused ultrasound thalamotomy over time.

Conclusion: Our objective method, which incorporated image processing and signal processing, provided a quantitative measure of tremor reduction following focused ultrasound thalamotomy. It demonstrated significant improvement in tremors in the treated and non-treated hands following focused ultrasound thalamotomy as well as deterioration in the efficacy of treatment over time. If replicated in other studies, this method may complement current subjective assessments.

Keywords: Essential tremor; Focused ultrasound thalamotomy; Parkinson’s disease; Signal processing; Spirals.

Fig. 1

An example of a scanned template and drawings completed by a patient. The dominant hand and the hand used for the drawing are marked at the top of the template by the clinician. Below this information are the two spiral templates, headed by the texts “Drawing A”—a wider spiral—and “Drawing B”—a narrower spiral, and the straight lines headed by the text “Drawing C.” Additional information is sometimes hand-written by the clinician, such as the date at the lower left side of the page. Scanning artifacts of sporadic black dots on the page and tilt of the template can be seen

Fig. 2

An example of a dataset spiral following the automated preprocessing of the template in Fig. 1. The spiral under “Drawing A” was cropped out, its tilt corrected and all black dots induced by scanning removed

Fig. 3

A schematic description of the image processing method that computed the change in a patient’s performance in the spiral-drawing task. In the preprocessing phase, depicted in the rectangle, a template spiral and its reference line—in green—are prepared. For each scanned and cropped patient’s drawing in a spiral, the template spiral is removed. Both the reference line (green) and the patient’s drawn line are sampled (red dots), and the performance score featuring the L1 distance between the two sampled lines is calculated

Fig. 4

Change from baseline (pre-treatment) in patients as a function of time following the treatment for Archimedes spiral drawing. The average ± standard deviation of the population are drawn. The horizontal axis is the days elapsed from treatment. The vertical axis is the percent change from the baseline for the patient population