Objective: To examine the specificity and sensitivity of complexity metric derived from hand tremor fluctuation in the identification of Parkinsonian tremor (PT) from essential tremor (ET).

Background: PT and ET are two most common movement disorders, and are often misdiagnosed in clinics. The spontaneous fluctuation of tremor is regulated by numerous physiological interactions over multiple scales of space and time. The dynamics of the tremor is thus “complex”, containing meaningful information pertaining to the underlying physiological function. We previously showed that the complexity of tremor fluctuation is significantly different between PT and ET, but the specificity and sensitivity of such complexity metric to identify PT from ET is unclear.

Method: Seventy participants (48 in training set and 22 in testing set) with clinically-diagnosed PT and seventy age-matched participants with ET completed this study. Participants completed two 30-second tests to measure the acceleration of tremor in both left and right hand in the conditions that sitting while arms were at resting state (i.e., resting tremor) and while arms were outstretched horizontally (i.e., postural tremor). The multiscale entropy (MSE) was then used to quantify the complexity of the acceleration time series.  The Receiver operating characteristic (ROC) curve was used to obtain the diagnostic threshold of the complexity tremor.

Results: Compared to PT group, ET group had lower complexity of both hands across conditions (F>34.2, p<0.001). Moreover, the ROC curves revealed that the complexity metric can distinguish ET from PT (area-under-the-curve=0.77~0.88, cut-off value=48 (postural), 49 (resting)). Using the cut-off score of 49 of averaged complexity metric in resting condition, 18 out of 22 participants with ET were identified correctly and 5 of 22 participants with PT were identified as ET incorrectly, showing an 80% accuracy to identify ET from PT. Using the cut-off score of 48 in postural condition, 20 out of 22 participants with ET were identified correctly and only 2 of the participants with PT were identified as ET incorrectly, revealing a high accuracy of 90% to identify ET from PT.

Conclusion: We here demonstrate first-of-its-kind evidence that the physiological complexity of hand tremor captures different pathology in PT and ET, and may serve as a novel marker to help the classification of these pathological conditions in clinical practice.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/the-complexity-metric-of-hand-tremor-identifies-parkinsonian-tremor-and-essential-tremor/