Gait Parameters Measured from Wearable Sensors Reliably Detect Freezing of Gait in a Stepping in Place Task

Abstract

Freezing of gait (FOG), a debilitating symptom of Parkinson's disease (PD), can be safely studied using the stepping in place (SIP) task. However, clinical, visual identification of FOG during SIP is subjective and time consuming, and automatic FOG detection during SIP currently requires measuring the center of pressure on dual force plates. This study examines whether FOG elicited during SIP in 10 individuals with PD could be reliably detected using kinematic data measured from wearable inertial measurement unit sensors (IMUs). A general, logistic regression model (area under the curve = 0.81) determined that three gait parameters together were overall the most robust predictors of FOG during SIP: arrhythmicity, swing time coefficient of variation, and swing angular range. Participant-specific models revealed varying sets of gait parameters that best predicted FOG for each participant, highlighting variable FOG behaviors, and demonstrated equal or better performance for 6 out of the 10 participants, suggesting the opportunity for model personalization. The results of this study demonstrated that gait parameters measured from wearable IMUs reliably detected FOG during SIP, and the general and participant-specific gait parameters allude to variable FOG behaviors that could inform more personalized approaches for treatment of FOG and gait impairment in PD.

Keywords: Parkinson’s disease; freezing of gait; inertial measurement unit; sensors; wearables.

Figure 1

Gait parameters extracted from wearable inertial measurement unit sensors (IMUs). Participants performed the stepping in place task on dual force plates (dark gray). Two IMUs were mounted on the lateral side of the shanks, and shank angular velocities in the sagittal plane from the left (blue) and right (red) legs were measured. Gait parameters, such as peak shank angular velocity (purple), swing time (pink), stride time (brown), and swing angular range (green), were extracted from shank angular velocity data.

Figure 2

Detection of freezing of gait (FOG) during the stepping in place (SIP) task. FOG detection during SIP from kinetics derived from vertical force measured by dual force plates (A) and from kinematics derived from shank angular velocity measured by wearable inertial measurement unit sensors (B). Blue and red traces correspond to the left and right legs, respectively. Swing angular range (C) and arrhythmicity (D) calculated from shank angular velocity show visually detectable differences between freezing and non-freezing episodes.

Figure 3

All steps from all participants mapped according to the top three general predictors of freezing of gait (FOG) during the stepping in place (SIP) task. The general model revealed that arrhythmicity, swing time coefficient of variation (CV), and swing angular range together were the most robust predictors of FOG during SIP. There is a clear separation between freeze (red) and non-freeze (green) steps during SIP.

Figure 4

All steps from participants 1 and 2 mapped according to their top three participant-specific predictors of freezing of gait (FOG) during the stepping in place (SIP) task. Participant-specific models revealed that varying sets of gait parameters best predicted FOG during SIP for each participant. For example, participant 1′s most robust predictors of FOG were a combination of peak shank angular velocity, swing time, and swing angular range (left), while participant 2′s most robust predictors of FOG were a combination of arrhythmicity, swing time, and swing time coefficient of variation (CV) (right). For both participants, there is a clear separation between freeze (red) and non-freeze (green) steps during SIP.