Wearable Sensors System for an Improved Analysis of Freezing of Gait in Parkinson's Disease Using Electromyography and Inertial Signals

Abstract

We propose a wearable sensor system for automatic, continuous and ubiquitous analysis of Freezing of Gait (FOG), in patients affected by Parkinson's disease. FOG is an unpredictable gait disorder with different clinical manifestations, as the trembling and the shuffling-like phenotypes, whose underlying pathophysiology is not fully understood yet. Typical trembling-like subtype features are lack of postural adaptation and abrupt trunk inclination, which in general can increase the fall probability. The targets of this work are detecting the FOG episodes, distinguishing the phenotype and analyzing the muscle activity during and outside FOG, toward a deeper insight in the disorder pathophysiology and the assessment of the fall risk associated to the FOG subtype. To this aim, gyroscopes and surface electromyography integrated in wearable devices sense simultaneously movements and action potentials of antagonist leg muscles. Dedicated algorithms allow the timely detection of the FOG episode and, for the first time, the automatic distinction of the FOG phenotypes, which can enable associating a fall risk to the subtype. Thanks to the possibility of detecting muscles contractions and stretching exactly during FOG, a deeper insight into the pathophysiological underpinnings of the different phenotypes can be achieved, which is an innovative approach with respect to the state of art.

Keywords: Parkinson’s disease; gait analysis; inertial signal; sensor fusion; surface electromyography; telemedicine; wearable sensors.

Figure 1

Sketch of the paper flow.

Figure 2

Back-side of the Bio2Bit Move (top left); patch electrodes (top right); and bio-potential system block diagram used for sEMG acquisition (bottom).

Figure 3

Sketch of the device positioning on: (a) the tibialis anterior; and (b) the gastrocnemius of the right leg. The axis orientation is reported.

Figure 4

Stages of a conventional preprocessing of the raw sEMG signal.

Figure 5

An example of the normalized angular velocity around the z-axis recorded on the TA in correspondence of the distinct phases of a single step: (A) toe-off; (B) initial swing; (C) mid-swing; (D) heel strike; and (E) mid-stance.

Figure 6

(a) An example of the sEMG (blue) and gyro (red) traces recorded on the TA; (b) zoom of the trace in a regular gait interval; (c) zoom of the trace in the FOG interval; (d) zoom of a single regular step; (e) zoom of a single shuffled step; and (f) zoom of step attempts during trembling FOG.

Figure 7

Flowchart of: (a) the FOG detection algorithm; and (b) the improved FOG detection algorithm.

Figure 8

(a) An example of the normalized gyro trace of a single regular step recorded on the TA; (b) absolute value of the angular velocity; and (c) zoom around the threshold T₁ = 0.01.

Figure 9

(a) An example of the sEMG (blue) and gyro (red) traces recorded on the TA (coinciding with a portion of Figure 6a, re-reported here for the sake of clarity); (b) the corresponding R values on the same time scale; (c) a scatter plot of all the tests; and (d) a box plot of all the tests.

Figure 10

Sketch of the algorithm blocks implemented in the frequency domain for the offline distinction of the FOG phenotypes.

Figure 11

Shuffling (left) and trembling (right) FOG: (a) product of the normalized sEMG and gyro traces (inset: power spectral density); (b) low pass filtering at 2 Hz; (c) left and right leg traces; and (d) result of the subtraction of the right and left leg traces.

Figure 12

Power spectral density of the trace displayed in Figure 11d during: (a) shuffling FOG; and (b) trembling FOG.

Figure 13

Muscle activity type traces relative to TA (red) and GC (black) of a single patient (P1): (a) outside FOG; (b) during a shuffling FOG; and (c) during a trembling FOG. The insets show the intensity in the same time interval. The timescale is the same in the three plots.

Figure 14

sEMG traces for the GC and TA recorded in the time interval $\Delta$t = 1 s referring to: (a) a single regular step of patient P1 (top plot) and the ensemble average of single regular steps (bottom plot); (b) a single shuffled step of patient P1 (top plot) and the ensemble average of single shuffled steps (bottom plot); and (c) a trembling episode of patient P1 (top plot) and the ensemble average of trembling episodes (bottom plot).