Prediction of Freezing of Gait in Patients with Parkinson's Disease Using EEG Signals.

Freezing of gait (FOG) is an episodic gait disturbance affecting initiation and continuation of locomotion in many Parkinson's disease (PD) patients, causing falls and a poor quality of life. FOG can be experienced on turning and start hesitation, passing through doorways or crowded areas dual tasking, and in stressful situations. Electroencephalography (EEG) offers an innovative technique that may be able to effectively foresee an impending FOG. From data of 16 PD patients, using directed transfer function (DTF) and independent component analysis (ICA) as data pre-processing, and an optimal Bayesian neural network as a predictor of a transition of 5 seconds before the impending FOG occurs in 11 in-group PD patients, we achieved sensitivity and specificity of 85.86% and 80.25% respectively in the test set (5 out-group PD patients). This study therefore contributes to the development of a non-invasive device to prevent freezing of gait in PD.

An EEG study of turning freeze in Parkinson's disease patients: The alteration of brain dynamic on the motor and visual cortex.

Freezing of gait is a very debilitating symptom affecting many patients with Parkinson's disease, leading to a reduced mobility and increased risk for falls. Turning is known to be the most provocative trigger for freezing of gait. However, the underlying brain dynamic changes associated with a turning freeze remain unknown. This study therefore used ambulatory EEG to investigate the brain dynamic changes associated with freezing of gait during turning. In addition, this study aimed to determine the most suitable EEG sensor location to detect freezing of gait during turning using our classification system. Data from four Parkinson's disease patients with freezing of gait was analysed using power spectral density and brain effective connectivity, comparing periods of successful turning with freezing of gait during turning. Results showed that freezing of gait during turning is associated with significant alterations in the high beta and theta power spectral densities across the occipital and parietal areas. Furthermore, brain effective connectivity showed that freezing during turning was associated with increased connectivity towards the visual area, which also had the highest accuracy to detect freezing episodes in the O1 regions by using power spectral density in our classification analyses. This is the first study to show cortical dynamic changes associated with freezing of gait during turning, providing valuable information to enhance the performance of future freezing of gait detection systems.

Analysis and Prediction of the Freezing of Gait Using EEG Brain Dynamics.

Freezing of Gait (FOG) is a common symptom in the advanced stages of Parkinson's disease (PD), which significantly affects patients' quality of life. Treatment options offer limited benefit and there are currently no mechanisms able to effectively detect FOG before it occurs, allowing time for a sufferer to avert a freezing episode. Electroencephalography (EEG) offers a novel technique that may be able to address this problem. In this paper, we investigated the univariate and multivariate EEG features determined by both Fourier and wavelet analysis in the confirmation and prediction of FOG. The EEG power measures and network properties from 16 patients with PD and FOG were extracted and analyzed. It was found that both power spectral density and wavelet energy could potentially act as biomarkers during FOG. Information in the frequency domain of the EEG was found to provide better discrimination of EEG signals during transition to freezing than information coded in the time domain. The performance of the FOG prediction systems improved when the information from both domains was used. This combination resulted in a sensitivity of 86.0%, specificity of 74.4%, and accuracy of 80.2% when predicting episodes of freezing, outperforming current accelerometry-based tools for the prediction of FOG.

Using EEG spatial correlation, cross frequency energy, and wavelet coefficients for the prediction of Freezing of Gait in Parkinson's Disease patients.

Parkinson's Disease (PD) patients with Freezing of Gait (FOG) often experience sudden and unpredictable failure in their ability to start or continue walking, making it potentially a dangerous symptom. Emerging knowledge about brain connectivity is leading to new insights into the pathophysiology of FOG and has suggested that electroencephalogram (EEG) may offer a novel technique for understanding and predicting FOG. In this study we have integrated spatial, spectral, and temporal features of the EEG signals utilizing wavelet coefficients as our input for the Multilayer Perceptron Neural Network and k-Nearest Neighbor classifier. This approach allowed us to predict transition from walking to freezing with 87 % sensitivity and 73 % accuracy. This preliminary data affirms the functional breakdown between areas in the brain during FOG and suggests that EEG offers potential as a therapeutic strategy in advanced PD.

The detection of Freezing of Gait in Parkinson's disease patients using EEG signals based on Wavelet decomposition.

Freezing of Gait (FOG) is one of the most disabling gait disturbances of Parkinson's disease (PD). The experience has often been described as "feeling like their feet have been glued to the floor while trying to walk" and as such it is a common cause of falling in PD patients. In this paper, EEG subbands Wavelet Energy and Total Wavelet Entropy were extracted using the multiresolution decomposition of EEG signal based on the Discrete Wavelet Transform and were used to analyze the dynamics in the EEG during freezing. The Back Propagation Neural Network classifier has the ability to identify the onset of freezing of PD patients during walking using these features with average values of accuracy, sensitivity and specificity are around 75 %. This results have proved the feasibility of utilized EEG in future treatment of FOG.

Optimization of head movement recognition using Augmented Radial Basis Function Neural Network.

For people with severe spine injury, head movement recognition control has been proven to be one of the most convenient and intuitive ways to control a power wheelchair. While substantial research has been done in this area, the challenge to improve system reliability and accuracy remains due to the diversity in movement tendencies and the presence of movement artifacts. We propose a Neural-Network Configuration which we call Augmented Radial Basis Function Neural-Network (ARBF-NN). This network is constructed as a Radial Basis Function Neural-Network (RBF-NN) with a Multilayer Perceptron (MLP) augmentation layer to negate optimization limitation posed by linear classifiers in conventional RBF-NN. The RBF centroid is optimized through Regrouping Particle Swarm Optimization (RegPSO) seeded with K-Means. The trial results of ARBF-NN on Head-movement show a significant improvement on recognition accuracy up to 98.1% in sensitivity.