Table 1.

Literature review of the recent works in ET/PD classification.

References	Goal	Dataset	Method	Results
Hossen et al.23	ET/PD classification	Accelerometer data, [19 PD, 21 ET] for training and [20 PD, 20 ET] for testing	Statistical Signal Characterization performed on the spectral domain of tremor signals	Accuracy = 90%
Ghassemi et al.31	ET/PD classification	Electromyogram and accelerometer data, [13 PD, 11 ET] for training and testing	Classification of Wavelet features with Support Vector Machines (SVM)	Accuracy = 83%
Brzan et al.49	ET/PD Classificclassificationation	Electromyogram data [27 PD, 27 ET] for training and testing	A set of statistical and physiological features classified with decision tree	Accuracy = 94%
DiBiase et al.15	ET/PD classification	Accelerometer data, [16 PD, 20 ET] for training and [55] for testing	Analysis in spectral domain	Accuracy = 92%, Sensitivity = 95%, Specificity = 95%
Barrantes et al.50	ET/PD/Healthy classification	Accelerometer data, [17 PD, 16 ET, 12 healthy, 7 unknown]	Spectral analysis of the signals	Accuracy=84.38%
Molparia et al.51	ET/PD classification	Accelerometer data and genetic profiles, [40 PD, 27 ET] for training and testing	Statistical properties of signal along with genomics data	Sensitivity = 76%, Specificity = 65%
Locatelli et al.52	ET/PD classification	Low power wearable device, [17 PD, 7 ET] for training and testing	Various machine learning techniques	Accuracy=$95.8\%$
Moon et al.53	ET/PD classification	Gain and balance characteristics, [524 PD, 43 ET] for training and testing	Hand-crafted features and classical ML	Accuracy =$92\%$
Dugue et al.54	ET/PD classification	Accelerometer data, [17 PD, 16 ET, 12 Healthy, 7 inconclusive]	Spectral features and various ML techniques	Accuracy = $84.4\%$