Tremor Morphology Toolkit: Waveform Modulation-Based Tremor Phenotyping

Clinicians typically characterize tremor by its frequency, amplitude, and body region affected. Other aspects also provide critical clues to the type of tremor and its underlying pathophysiology including tremor waveform morphology, consistency across cycles, and variability in amplitude and timing. We introduce the Tremor Morphology Toolkit (TMT), an eight metric, waveform derived panel that quantifies these features from upper limb inertial signals. Using accelerometer and gyroscope recordings from 34 individuals with upper-extremity tremor across four standardized tasks (rest, outstretched, wing beating, and kinetic), we computed: (i) rSHAPE, cycle shape consistency via pairwise similarity of normalized cycles; (ii) rSINE, sinusoidality, the mean correlation of each cycle with its best fit sinusoid; (iii) rCVA, the coefficient of variation of cycle amplitude; (iv) rCVD, the coefficient of variation of cycle duration; (v) AWI, amplitude warping index; (vi) DWI, duration warping index; (vii) FreqVar, a spectral bandwidth indexing power weighted frequency spread; and (viii) Amplitude Breathing indices, which quantify slow waxing and waning of tremor amplitude over time from fluctuations in the signal envelope, representing the time varying amplitude of the tremor waveform. We detail algorithms, parameterizations, and implementation defaults to support reproducible use. TMT is designed to complement conventional frequency amplitude measures by capturing tremor morphology and modulation in a simple, interpretable framework. It enables objective quantification of features often assessed qualitatively at the bedside, facilitates comparisons across recording conditions, and provides a foundation for validation against clinical labels and disease phenotypes. Future studies can integrate these parameters with electrophysiological measures to investigate mechanisms underlying tremor generation and modulation.