Objective: To develop a new easy-to-use wearable device to estimate the muscular contraction pattern of rest tremor (RT) using inertial data.

Background: The RT pattern is the best electrophysiological feature to distinguish parkinsonian (alternating) from non-parkinsonian (synchronous) RT disorders, in the absence of dopamine imaging.[1,2] RT pattern evaluation requires surface electromyography (EMG), which is cheaper than dopamine imaging but needs expertise and is usually limited to research centers. Many wearable devices for tremor analysis based on simple inertial sensors have been recently developed, but none of them is able to distinguish alternating from synchronous tremors.^[3]

Method: We enrolled 37 tremulous Parkinsonian patients (with dopaminergic deficit) and 33 non-parkinsonian RT patients (with normal dopamine imaging). We developed a new wearable device including a 6-axis ST LSM6DSL IMU sensor and an Adafruit Feather nRF52840 development board, connected to a smartphone app by Bluetooth technology. Five recording segments of RT of 10 sec each were acquired from each patient. Surface EMG was simultaneously performed to assess the RT pattern, used as reference standard. The dataset of 390 segments was split into training and testing sets (70% and 30% of data); a Random Forest model was trained on a combination of features extracted from inertial data for the classification of RT pattern.

Results: The dataset included 211 recording segments with alternating RT pattern and 179 segments with synchronous pattern. The ML model based on inertial data showing the best performance had 0.94 sensitivity, 0.94 specificity and 0.94 accuracy in distinguishing alternating from synchronous tremor in the k-fold Cross-Validation performed in the training set. Similar performances (0.95 sensitivity, 0.92 specificity and 0.93 accuracy) were obtained in the validation set.

Conclusion: We developed a new low-cost device to estimate the RT pattern using a combination of inertial features, which may represent a first-level diagnostic test to be used in large populations by general practitioners or neurologists in ambulatory settings without using EMG. This may improve the early diagnosis of tremulous Parkinson’s disease worldwide, especially in rural areas or low-income countries with limited access to care or economical resources.

References: 1. Quattrone A, Nisticò R, Morelli M, Arabia G, Crasà M, Vescio B, Mechelli A, Cascini GL, Quattrone A. Rest Tremor Pattern Predicts DaTscan (123 I-Ioflupane) Result in Tremulous Disorders. Mov Disord. 2021; 36: 2964-2966. doi: 10.1002/mds.28797.
2. Nisticò R, Pirritano D, Salsone M, Novellino F, Del Giudice F, Morelli M, Trotta M, Bilotti G, Condino F, Cherubini A, Valentino P, Quattrone A. Synchronous pattern distinguishes resting tremor associated with essential tremor from rest tremor of Parkinson’s disease. Parkinsonism Relat Disord. 2011; 17: 30-3. doi: 10.1016/j.parkreldis.2010.10.006.
3. Vescio B, Quattrone A, Nisticò R, Crasà M, Quattrone A. Wearable Devices for Assessment of Tremor. Front Neurol. 2021; 12: 680011. doi: 10.3389/fneur.2021.680011.

« Back to 2023 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/development-and-validation-of-a-new-wearable-device-for-the-differential-diagnosis-of-resting-tremor/