Archimedes spirals have been widely used to quantify clinically meaningful action tremor along with being a part of The Essential Tremor Rating Assessment Scale (TETRAS) scale. It is common to see a variation in tremor between the beginning and end of a spiral as noted during data analysis. Understanding the cause of these separate spiral components could lead to more specific clinical use and interpretation of spirals. 

Patients with Essential Tremor (ET) drew multiple spirals on a digitizing tablet in our clinic along with performance of TETRAS. Time-frequency analysis was performed using wavelet analysis, comparing the first loop of the spiral (short segment) with the remaining segment of the spiral (long segment). The amplitude and frequency of the largest peak was extracted for analysis.  The results were correlated with TETRAS scale. In addition, some subjects were studied with EMG and accelerometers.

There was a high correlation in the power of the tremor at peak frequency between the short segment vs/ the rest of the spiral (r=0.98, p <0.001). Both the short and the long segment of the spirals show a significant correlation with total TETRAS scale, (R=0.59, p <0.001, R=0.53, p<0.001). Additionally, 63.4% of the subjects show two non-harmonic peaks in spirals for both short and long segments, with median peaks 5Hz and 7Hz. 

There is a significant correlation between the TETRAS score and the power of the peak of the tremor of both the initial segment and the rest of the spiral, with no significant difference between them. The two peaks observed on most of the spirals in the frequency domain may correspond to distinct presentations of action tremor, namely postural as well as kinetic tremor. We will also use EMG and accelerometry results to further study the two peaks.