Neuro-degenerative diseases classification based on blind source separation and Multifractal Detrended Fluctuation analysis of gait dynamics

Neurodegenerative diseases (NDD) including Amyotrophic Lateral Sclerosis (ALS), Parkinson’s disease (PD) and Huntington disease (HD) can be defined as the degeneration in the structure of neurons in human body. It is mentioned in the related literature that NDD may cause various clinical symptoms disrupting gait dynamics. The characterization of gait analysis is crucial for early diagnosis, efficient treatment planning and monitoring of ALS progression and other NDD. The database consisting of 64 five-minute recordings of Compound Force Signal (CFS) obtained from 13 ALS, 15 PD, 20 HD and 16 healthy subjects was used in the study. a five-stage structure is used. In the first step, a data group recorded by force sensitive sensors was used to analyze the walking dynamics that is underneath. In the second step, the signal filtered by the filter bank of wavelet transform with the default coefficients of the noise reduction and improve it. In third Step a set of feature is extracted from recorded data. In the fourth step, the extracted features are considered as inputs of a feature dimension reduction structure (principal component analysis). The reduced dimensional attributes are considered as inputs of linear classification structures (SVM) and nonlinear (KNN, D-Tree and MLP). The goal of finding a grade tag is the type of disease based on walking signal analysis. All simulations were implemented under MATLAB software and validation of the proposed method was done by analyzing the confusion matrix and calculating the accuracy, sensitivity and specificity index. The results of the simulation showed that the perceptron multi-layered neural network has a precision accuracy of 92% higher in the diagnosis of neurodegenerative complication based on dynamic walking analysis.

کلیدواژه ها:

Neurodegenerative ، ICA ، Genetic ، algorithm ، AdaBoost ، fractal