Differential Gait Decline in Parkinson's Disease Enhances Discrimination of Gait Freezers from Non-Freezers

Machine learning model comparison for freezing of gait prediction in advanced Parkinson's disease

Introduction

Freezing of gait (FOG) is a paroxysmal motor phenomenon that increases in prevalence as Parkinson's disease (PD) progresses. It is associated with a reduced quality of life and an increased risk of falls in this population. Precision-based detection and classification of freezers are critical to developing tailored treatments rooted in kinematic assessments.

Methods

This study analyzed instrumented stand-and-walk (SAW) trials from advanced PD patients with STN-DBS. Each patient performed two SAW trials in their OFF Medication-OFF DBS state. For each trial, gait summary statistics from wearable sensors were analyzed by machine learning classification algorithms. These algorithms include k-nearest neighbors, logistic regression, naïve Bayes, random forest, and support vector machines (SVM). Each of these models were selected for their high interpretability. Each algorithm was tasked with classifying patients whose SAW trials MDS-UPDRS FOG subscore was non-zero as assessed by a trained movement disorder specialist. These algorithms' performance was evaluated using stratified five-fold cross-validation.

Results

A total of 21 PD subjects were evaluated (average age 64.24 years, 16 males, mean disease duration of 14 years). Fourteen subjects had freezing of gait in the OFF MED/OFF DBS. All machine learning models achieved statistically similar predictive performance (p < 0.05) with high accuracy. Analysis of random forests' feature estimation revealed the top-ten spatiotemporal predictive features utilized in the model: foot strike angle, coronal range of motion [trunk and lumbar], stride length, gait speed, lateral step variability, and toe-off angle.

Conclusion

These results indicate that machine learning effectively classifies advanced PD patients as freezers or nonfreezers based on SAW trials in their non-medicated/non-stimulated condition. The machine learning models, specifically random forests, not only rely on but utilize salient spatial and temporal gait features for FOG classification.

Development and implementation of the frog-in-maze game to study upper limb movement in people with Parkinson's disease

Upper-limb bradykinesia occurs early in Parkinson's disease (PD) and bradykinesia is required for diagnosis. Our goal was to develop, implement and validate a game "walking" a frog through a maze using bimanual, alternating finger-tapping movements to provide a salient, objective, and remotely monitorable method of tracking disease progression and response to therapy in PD. Twenty-five people with PD and 16 people without PD participated. Responses on 5 different mazes were quantified and compared to spatiotemporal gait parameters and standard disease metrics in these participants. Intertap interval (ITI) on maze 2 & 3, which included turns, was strongly inversely related to stride-length and stride-velocity and directly related to motor UPDRS scores. Levodopa decreased ITI, except in maze 4. PD participants with freezing of gait had longer ITI on all mazes. The responses quantified on maze 2 & 3 were related to disease severity and gait stride-length, were levodopa responsive, and were worse in people with freezing of gait, suggesting that these mazes could be used to quantify motor dysfunction in PD. Programming our frog-in-maze game onto a remotely distributable platform could provide a tool to monitor disease progression and therapeutic response in people with PD, including during clinical trials.

Levodopa responsive gait dynamics in OFF- and ONOFF-state freezing of gait in Parkinson's disease

Background

In people with Parkinson's disease (PwPD), Freezing of Gait (FOG) episodes can be levodopa responsive (OFF-FOG) or levodopa unresponsive (ONOFF-FOG). Steady-state gait abnormalities, outside of the freezing episodes themselves also exist and the response to levodopa in these different groups has not been previously documented.

Objectives

To define the levodopa responsiveness in steady-state gait in OFF-FOG and ONOFF-FOG individuals.

Methods

Steady-state gait was collected in both the effective levodopa OFF-state (doses withheld > 8 h) and ON-state (1 h after levodopa dosing) in 32 PwPD; 10 with OFF-FOG and 22 with ONOFF-FOG. Levodopa response was compared between the two groups in the mean and variability (CV) of 8 spatiotemporal gait parameters.

Results

Both OFF-FOG and ONOFF-FOG participants showed improvement in mean stride-length and stride-velocity with levodopa. Improvement was seen in the OFF-FOG but not the ONOFF-FOG groups in mean stride-width and CV Integrated pressure with levodopa.

Discussion

In this study we show that steady-state gait deficits improve with levodopa in PwPD with OFF-FOG and ONOFF-FOG, even though episodes of FOG did not resolve in the ONOFF-FOG group. Lowering levodopa in people with ONOFF-FOG, or levodopa-unresponsive freezing of gait, should be undertake with caution and objective gait titration at different levodopa doses may be beneficial. Further work is needed to elucidate the pathophysiologic mechanisms of these differences.

Gait performance and non-motor symptoms burden during dual-task condition in Parkinson's disease

INTRODUCTION

Impaired gait is observed in patients with Parkinson's disease (PD) in both single-task (ST) and dual-task (DT) conditions. Non-motor symptoms (NMSs), another vital symptom future experienced along the PD disease trajectory, contribute to gait performance in PD. However, whether DT gait performance is indicative of NMS burden (NMSB) remains unknown. This study investigated correlation between NMS and DT gait performance and whether NMSB is reflected in the DT effects (DTEs) of gait parameters in PD.

METHODS

Thirty-three idiopathic PD participants were enrolled in this study; the median H-Y staging was 2.5. NMSB was assessed by Non-motor Symptoms Scale (NMSS). Spatiotemporal gait parameters under ST and DT conditions were evaluated by wearable sensors. Gait parameters under ST and DT conditions and DTEs of gait parameters were compared across NMSB groups. The associations between NMS and DTEs of gait parameters were analyzed by correlation analysis and linear regression models.

RESULTS

Compared to PD patients with mild-moderate NMSB, the severe-very severe NMSB group showed slower gait speed and shorter stride length under both ST and DT conditions (p < 0.05). DT had significantly negative effect on gait parameters in PD patients, including gait speed, stride length, and gait cycle duration (p < 0.05). PD patients with mild-moderate NMSB showed larger DTEs of cadence and bilateral gait cycle duration (p < 0.05). DTEs of bilateral gait cycle duration and swing phase on the more affected (MA) side were significantly correlated with NMSS scores (∣rSp∣ ≥ 0.3, p < 0.05). Gait cycle duration on the less affected (LA) side explained 43% of the variance in NMSS scores, when accounting for demographic and clinical confounders (β =  - 1.095 95% CI - 4.061 ~  - 0.058, p = 0.044; adjusted R² = 0.434).

CONCLUSION

DT gait performance could reflect NMSB in PD patients at early stage, and gait cycle duration is a valuable gait parameter to further investigate and to provide more evidence for PD management.

Gait Declines Differentially in, and Improves Prediction of, People with Parkinson's Disease Converting to a Freezing of Gait Phenotype

BACKGROUND

Freezing of gait (FOG) is a debilitating, variably expressed motor symptom in people with Parkinson's disease (PwPD) with limited treatments.

OBJECTIVE

To determine if the rate of progression in spatiotemporal gait parameters in people converting from a noFOG to a FOG phenotype (FOGConv) was faster than non-convertors, and determine if gait parameters can help predict this conversion.

METHODS

PwPD were objectively monitored longitudinally, approximately every 6 months. Non-motor assessments were performed at the initial visit. Steady-state gait in the levodopa ON-state was collected using a gait mat (Protokinetics) at each visit. The rate of progression in 8 spatiotemporal gait parameters was calculated. FOG convertors (FOGConv) were classified if they did not have FOG at initial visit and developed FOG at a subsequent visit.

RESULTS

Thirty freezers (FOG) and 30 non-freezers were monitored an average of 3.5 years, with 10 non-freezers developing FOG (FOGConv). FOGConv and FOG had faster decline in mean stride-length, swing-phase-percent, and increase in mean total-double-support percent, coefficient of variability (CV) foot-strike-length and CV swing-phase-percent than the remaining non-freezers (noFOG). On univariate modeling, progression rates of mean stride-length, stride-velocity, swing-phase-percent, total-double-support-percent and of CV swing-phase-percent had high discriminative power (AUC > 0.83) for classification of the FOGConv and noFOG groups.

CONCLUSION

FOGConv had a faster temporal decline in objectively quantified gait than noFOG, and progression rates of spatiotemporal gait parameters were more predictive of FOG phenotype conversion than initial (static) parameters Objectively monitoring gait in disease prediction models may help define FOG prone groups for testing putative treatments.

Feasibility of telemedicine research visits in people with Parkinson's disease residing in medically underserved areas

Introduction

Gait, balance, and cognitive impairment make travel cumbersome for People with Parkinson's disease (PwPD). About 75% of PwPD cared for at the University of Arkansas for Medical Sciences' Movement Disorders Clinic reside in medically underserved areas (MUAs). Validated remote evaluations could help improve their access to care. Our goal was to explore the feasibility of telemedicine research visits for the evaluation of multi-modal function in PwPD in a rural state.

Methods

In-home telemedicine research visits were performed in PwPD. Motor and non-motor disease features were evaluated and quantified by trained personnel, digital survey instruments for self-assessments, digital voice recordings, and scanned and digitized Archimedes spiral drawings. Participant's MUA residence was determined after evaluations were completed.

Results

Twenty of the fifty PwPD enrolled resided in MUAs. The groups were well matched for disease duration, modified motor UPDRS, and Montreal Cognitive assessment scores but MUA participants were younger. Ninety-two percent were satisfied with their visit, and 61% were more likely to participate in future telemedicine research. MUA participants traveled longer distances, with higher travel costs, lower income, and education level. While 50% of MUA participants reported self-reliance for in-person visits, 85% reported self-reliance for the telemedicine visit. We rated audio-video quality highly in approximately 60% of visits in both groups. There was good correlation with prior in-person research assessments in a subset of participants.

Conclusions

In-home research visits for PwPD in MUAs are feasible and could help improve access to care and research participation in these traditionally underrepresented populations.

Levodopa ONOFF-state freezing of gait: Defining the gait and non-motor phenotype

Background

Freezing in the levodopa-medicated-state (ON-state) is a debilitating feature of Parkinson’s disease without treatment options. Studies detailing the distinguishing features between people with freezing of gait that improves with levodopa and those whose freezing continues even on levodopa are lacking.

Objective

To characterize the gross motor, gait, and non-motor features of this phenotype.

Methods

Instrumented continuous gait was collected in the levodopa-medicated-state in 105 patients: 43 non-freezers (no-FOG), 36 with freezing only OFF-levodopa (OFF-FOG) and 26 with freezing both ON- and OFF-levodopa (ONOFF-FOG). Evaluation of motor and non-motor disease features was undertaken using validated scales. A linear mixed model with age, sex, disease duration, and motor UPDRS scores as covariates was used to determine differences in spatiotemporal gait and non-motor disease features among the groups.

Results

Compared to OFF-FOG, the ONOFF-FOG group had greater disease severity (on the Unified Parkinson’s disease Rating Scale) and worse cognition (on the Montreal Cognitive Assessment, Frontal Assessment Battery and Scales for Outcome in Parkinson’s disease-Cognition scales) and quality of life (on the PDQ-39), but similar mood (on the Hamilton depression and anxiety scales) and sleep quality (on Epworth sleepiness scale and RBD questionnaire). For several gait features, differences between the ONOFF-OFF groups were at least as large and in the opposite direction as differences between OFF-no groups, controlling for disease severity. Variability in ONOFF-FOG was greater than in other groups. Using results from our study and others, a power analysis for a potential future study reveals sample sizes of at least 80 ONOFF and 80 OFF-FOG patients would be needed to detect clinically meaningful differences.

Conclusions

Intra-patient variability in spatiotemporal gait features was much greater in ONOFF-FOG than in the other two groups. Our results suggest that multifactorial deficits may lead to ONOFF-FOG development.

Amplitude setting and dopamine response of finger tapping and gait are related in Parkinson's disease

Movement amplitude setting is affected early in Parkinson’s disease (PD), clinically manifesting as bradykinesia. Our objective was to determine if amplitude setting of upper limb bimanual movements and bipedal gait are similarly modulated in PD. 27 PD and 24 control participants were enrolled. Participants performed a bimanual anti-phase finger tapping task wearing gloves with joint angular sensors, and an instrumented gait assessment. Participants performed normal and fast paced assessments to vary motor load. PD participants were evaluated OFF (PD-OFF) and ON (PD-ON) levodopa. PD-OFF participants had smaller tap amplitude, and greater tap amplitude variability than controls in the more affected hands (all p < 0.05). Tap amplitude and stride length (p = 0.030) were correlated in PD-OFF. Tap amplitude was also correlated with motor UPDRS (p < 0.005) and bradykinesia motor (p < 0.05) and ADL (p < 0.005) UPDRS subscores. The relative amount of improvement in tap amplitude and stride length with levodopa was correlated. In PD, upper limb and gait amplitude setting are similarly scaled with motor demand and dopamine supplementation. This suggests these automated motor functions are subserved by common functional networks.