Evaluation of the Motor Performance of People with Parkinson’s Disease Through the Autocorrelation Function Estimated from Sinusoidal Drawings

The objective of this work was to use the autocorrelation function to assess the fine motor performance of a healthy control group (CG) and people with Parkinson’s disease (PD), whose fluctuations caused by the disease compromise motor skills, performed mainly by the hands and fingers essential for daily activities such as dressing, carrying objects and taking care of personal hygiene. For this, accelerometer signals were collected while the volunteers drew a sinusoidal pattern. The correlogram was estimated and two features were calculated from it: (i) the normalized area under the curve (\( \varvec{AuC} \)); (ii) the difference between matching peaks and valleys from the reference and estimated correlograms (\( \varvec{v}_{{\varvec{mm}}} \)). These features allowed for the discrimination between drawing patterns of CG (\( \varvec{v}_{{\varvec{mm}}} \) = 0.154 ± 0.063, \( \varvec{AuC} \)  = 0.204 ± 0.040) and PD (\( \varvec{v}_{{\varvec{mm}}} \) = 0.239 ± 0.085, \( \varvec{AuC} \) = 0.179 ± 0.033) groups. Following the verification of normality (Shapiro-Wilk, p > 0.05), the t-test was applied to confirm the significant differences between groups (p < 0.05). The Cohen’s d effect size (p < 0.05) was medium (0.730) for \( \varvec{AuC} \) and large (1.252) for \( \varvec{v}_{{\varvec{mm}}} \). This further confirms the differences between the features extracted from the groups. Therefore, the features \( \varvec{AuC} \) and \( \varvec{v}_{{\varvec{mm}}} \) estimated from the autocorrelation function are effective for the assessment of the motor performance of healthy individuals and people with PD.