Wristbands Containing Accelerometers for Objective Arm Swing Analysis in Patients with Parkinson's Disease

Abstract

In patients with Parkinson's disease (PD), arm swing changes are common, even in the early stages, and these changes are usually evaluated subjectively by an expert. In this article, hypothesize that arm swing changes can be detected using a low-cost, cloud-based, wearable, sensor system that incorporates triaxial accelerometers. The aim of this work is to develop a low-cost, assistive diagnostic tool for use in quantifying the arm swing kinematics of patients with PD. Ten patients with PD and 11 age-matched, healthy subjects are included in the study. Four feature extraction techniques were applied: (i) Asymmetry estimation based on root mean square (RMS) differences between arm movements; (ii) posterior-anterior phase and cycle regularity through autocorrelation; (iii) tremor energy, established using Fourier transform analysis; and (iv) signal complexity through the fractal dimension by wavelet analysis. The PD group showed significant (p < 0.05) reductions in arm swing RMS values, higher arm swing asymmetry, higher anterior-posterior phase regularities, greater "high energy frequency" signals, and higher complexity in their XZ plane signals. Therefore, the novel, portable system provides a reliable means to support clinical practice in PD assessment.

Keywords: Parkinson’s disease; accelerometer; age-related chronic diseases/syndromes; gait; mobile health; wearables.

Figure 1

Smart wristbands used to measure the linear and angular acceleration of the subject’s arms in three axes. Data are uploaded through a network consisting of a smartphone that receives the data and then uploads it to a database on the cloud for subsequent analysis using a MATLAB^® script.

Figure 2

Acceleration signal and RMS for an PD patient and a control. (A) Control right arm swing. (B) Control left arm swing. (C) PD patient right arms swing. (D) PD patient left arm swing. Note that the patient exhibits very different RMS, due to waveform and amplitude between limbs.

Figure 3

Regularity based on autocorrelation of accelerometer signals. (A) Control regularity. (B) Patient Regularity. (Red) left arm. (Blue) Right arm. The first peak (A) reveals the regularity of the step, whereas the second peak (B) reveals the regularity of the stride. Patient presents a lower A peak than control one. The B peaks in both cases are similar.

Figure 4

1–2–3–4–1 represents all phases of a complete gait cycle. 1–2–3 are representations of gait states in which the feet and arms have completed a half gait cycle. The peaks in Y are presented in the changes of the direction of arm movement (1, 3, and 1), and their valleys appear when the arm crosses the frontal plane (2 and 4).

Figure 4

1–2–3–4–1 represents all phases of a complete gait cycle. 1–2–3 are representations of gait states in which the feet and arms have completed a half gait cycle. The peaks in Y are presented in the changes of the direction of arm movement (1, 3, and 1), and their valleys appear when the arm crosses the frontal plane (2 and 4).

Figure 5

Examples of the spectral frequency for PD patients and controls. (A) Control frequency spectrum, step and strike peaks are in 1Hz and 2Hz, respectively. (B) Patient frequency spectrum, step and strike peaks are lower than tremor peak around 5H. (Red indicates the left arm, and blue indicates the right arm).

Figure 6

Variance at the wavelet level of detail j (in log₂ scale) for an PD patient and a control. (Red indicates the left arm, and blue indicates the right arm). (A) Control signal variance. Both arms variance through wavelet decomposition present linear trend, which causes a slightly different slope in the line (β). (B) PD patient signal variance. The less affected arm has a normal slope (β), but the affected arm presents a lower slope (β = 2.01).

Figure 7

In patients with Parkinson’s disease, a tremor is present in the XZ plane, and the Y-axis acts as the axis of rotation for this movement.