Effectiveness of Center of Pressure Trajectory as Anticipatory Postural Adjustment Measurement in Parkinson's Disease With Freezing of Gait History

Deep Learning-Based Prediction of Freezing of Gait in Parkinson's Disease With the Ensemble Channel Selection Approach

PURPOSE

A debilitating and poorly understood symptom of Parkinson's disease (PD) is freezing of gait (FoG), which increases the risk of falling. Clinical evaluations of FoG, relying on patients' subjective reports and manual examinations by specialists, are unreliable, and most detection methods are influenced by subject-specific factors.

METHOD

To address this, we developed a novel algorithm for detecting FoG events based on movement signals. To enhance efficiency, we propose a novel architecture integrating a bottleneck attention module into a standard bidirectional long short-term memory network (BiLSTM). This architecture, adaptable to a convolution bottleneck attention-BiLSTM (CBA-BiLSTM), classifies signals using data from ankle, leg, and trunk sensors.

FINDING

Given three movement directions from three locations, we reduce computational complexity in two phases: selecting optimal channels through ensemble learning followed by feature reduction using attention mapping. In FoG event detection tests, performance improved significantly compared to control groups and existing methods, achieving 99.88% accuracy with only two channels.

CONCLUSION

The reduced computational complexity enables real-time monitoring. Our approach demonstrates substantial improvements in classification results compared to traditional deep learning methods.

Swing limb detection using a convolutional neural network and a sequential hypothesis test based on foot pressure data during gait initialization in individuals with Parkinson's disease

Objective. Start hesitation is a key issue for individuals with Parkinson's disease (PD) during gait initiation. Visual cues have proven effective in enhancing gait initiation. When applied to laser-light shoes, swing-limb detection efficiently activates the laser on the side of the stance limb, prompting the opposite swing limb to initiate stepping.Approach. This paper presents the development of two models for this purpose: a convolutional neural network that predicts the swing limb's side using center of pressure data, and a swing onset detection model based on sequential hypothesis test using foot pressure data.Main results. Our findings demonstrate an accuracy rate of 85.4% in predicting the swing limb's side, with 82.4% of swing onsets correctly detected within 0.05 s.Significance. This study demonstrates the efficiency of swing-limb detection based on foot pressures. Future research aims to comprehensively assess the impact of this method on improving gait initiation in individuals with PD.

Investigating spatiotemporal and kinematic gait parameters in individuals with Parkinson's disease with a history of freezing of gait and exploring the effects of dopaminergic therapy on freezing of gait subtypes

Introduction

Freezing of Gait (FOG) is a prevalent and debilitating symptom in idiopathic Parkinson's disease (PD). This study evaluated spatiotemporal and kinematic gait parameters in individuals with PD with a history of FOG and explored the effects of dopaminergic therapy on FOG subtypes.

Methods

One hundred and nine individuals with PD underwent clinical assessments and quantitative biomechanical measures during walking cycles before and after dopaminergic therapy. Individuals with FOG were classified into levodopa-responsive and levodopa-unresponsive groups.

Results

Individuals with FOG displayed longer disease duration and higher Unified Parkinson's Disease Rating Scale (UPDRS) II, III, IV scores, and total scores and levodopa equivalent dose, than those without FOG (all p < 0.0001). Following propensity score matching of 15 pairs based on UPDRS total score and disease duration during the off-medication state, the analysis comparing the FOG and non-FOG groups revealed no significant differences in spatiotemporal and kinematic parameters. In 39 cases of FOG, dopaminergic therapy improved gait performance in individuals with PD, enhancing spatiotemporal parameters (speed, stride length, step length, step variability) and kinematic parameters (shoulder and elbow flexion/extension range of motion (ROM), pelvic rotation, and hip abduction/adduction ROM) regardless of FOG responsiveness to dopaminergic therapy. A significant difference in trunk sway ROM (p = 0.029) remained before and after dopaminergic therapy, even after adjusting for disease duration and clinical severity.

Discussion

Dopaminergic therapy had varying effects on PD with FOG, improving several spatiotemporal and kinematic gait parameters but being less effective in levodopa-unresponsive cases. Quantitative biomechanical measures offer detailed insights into gait performance, aiding personalized fall risk assessment and guiding individualized rehabilitation programs.

Exploring the role of anticipatory postural adjustment duration within APA2 subphase as a potential mediator between clinical disease severity and fall risk in Parkinson's disease

Introduction

People with Parkinson's Disease (PD) often show reduced anticipatory postural adjustments (APAs) before voluntary steps, impacting their stability. The specific subphase within the APA stage contributing significantly to fall risk remains unclear.

Methods

We analyzed center of pressure (CoP) trajectory parameters, including duration, length, and velocity, throughout gait initiation. This examination encompassed both the postural phase, referred to as anticipatory postural adjustment (APA) (APA1, APA2a, APA2b), and the subsequent locomotor phases (LOC). Participants were instructed to initiate a step and then stop (initiating a single step). Furthermore, we conducted assessments of clinical disease severity using the Unified Parkinson's Disease Rating Scale (UPDRS) and evaluated fall risk using Tinetti gait and balance scores during off-medication periods.

Results

Freezing of gait (FOG) was observed in 18 out of 110 participants during the measurement of CoP trajectories. The Ramer-Douglas-Peucker algorithm successfully identified CoP displacement trajectories in 105 participants (95.5%), while the remaining 5 cases could not be identified due to FOG. Tinetti balance and gait score showed significant associations with levodopa equivalent daily dose, UPDRS total score, disease duration, duration (s) in APA2a (s) and LOC (s), length in APA1 (cm) and APA2b (cm), mediolateral velocity in APA1 (X) (cm/s), APA2a (X) (cm/s), APA2b (X) (cm/s) and LOC (X) (cm/s), and anterior-posterior velocity in APA2a (Z) (cm/s) and APA2b (Z) (cm/s). Multiple linear regression revealed that only duration (s) in APA2a and UPDRS total score was independently associated with Tinetti gait and balance score. Further mediation analysis showed that the duration (s) in APA2a served as a mediator between UPDRS total score and Tinetti balance and gait score (Sobel test, p = 0.047).

Conclusion

APA2 subphase duration mediates the link between disease severity and fall risk in PD, suggesting that longer APA2a duration may indicate reduced control during gait initiation, thereby increasing fall risk.

Predicting slight freezing of gait in Parkinson's disease with anticipatory postural adjustments and limits of stability

INTRODUCTION

Gait initiation (GI) includes automatic and voluntary movements. However, research on their impact on the first step in patients with Parkinson's disease (PD) and their relationship to freezing of gait (FOG) is lacking. We examined the effects of automatic movements (anticipatory postural adjustments [APAs]) and voluntary movements (limits of stability [LOS]) on the first step (first-step duration and first-step range of motion), along with their early recognition and prediction of slight FOG.

METHODS

Twenty-three patients with PD and slight freezing (PD + FOG) and 25 non-freezing patients with PD (PD-FOG) were tested while off medications and compared with 24 healthy controls (HC). All participants completed a 7-m Stand and Walk Test (7 m SAW) and wore inertial sensors to quantify the APAs and first step. LOS was quantified by dynamic posturography in different directions using a pressure platform. We compared differences among all three groups, analysed correlations, and evaluated their predictive value for slight FOG.

RESULTS

In PD + FOG, APAs and LOS were worse than those in the PD-FOG and HC groups (p < 0.001), and the first step was worse than that in HC (p < 0.001). APAs were correlated mainly with the first-step duration. APAs and LOS were correlated with the first-step range of motion. APAs have been recognized as independent predictors of FOG, and their combination with LOS enhances predictive sensitivity.

CONCLUSION

APAs and LOS in patients with PD directly affect the first step during GI. In addition, the combination of APAs and LOS helped predict slight FOG.