Determination of the type and level of multiple sclerosis disease using the signal of gait based on extraction of chaostic and statistical features of synergy patterns and the application of intelligent neural network

سال انتشار: 1398
نوع سند: مقاله کنفرانسی
زبان: انگلیسی
مشاهده: 454

فایل این مقاله در 5 صفحه با فرمت PDF قابل دریافت می باشد

این مقاله در بخشهای موضوعی زیر دسته بندی شده است:

استخراج به نرم افزارهای پژوهشی:

لینک ثابت به این مقاله:

شناسه ملی سند علمی:

CARSE04_082

تاریخ نمایه سازی: 17 اسفند 1398

چکیده مقاله:

The myelin sheath is a lipoprotein layer that forms on many long dendrites and axons, and its role is to create more disruption on the neural surface of the nerve, which speeds up the electrical conduction of long-distance electrical messages. To be. Loss of myelin sheath leads to failure in the delivery of nerve messages and therefore to neurological diseases such as multiple sclerosis. Multiple sclerosis consists of three levels of relapsing-remitting, primary progressive, secondary progressive. In general, the purpose of this study was to analyze the gait signal of MS patients using nonlinear extraction of synergy coefficients to classify different disease levels. In this study, we used gait data of 50 MS patients aged 43 10 10 years with EDSS score of around 3 with standard deviation of 2 degrees for hip and 2.7 degrees for knee and 1.4 degrees. For ankle walking compared to healthy people are the control group. They were asked to travel a distance of 10 meters. After collecting the required data, first the MATLAB default filter coefficients based on the wavelet transform bank filter are used to pre-process the signal of MS patients to eliminate noise. Then, the gait signal of patients with nonlinear features such as fractal and entropy dimension and Lyapunov view and coherence dimension is extracted and these features of MS patient gait during walking will be examined. Then, using the neural network clusters, different levels of the disease are differentiated to help prevent further complications by quickly and automatically detecting the disease. Nonlinear dimensionality reduction methods such as LLE are also used to reduce dimensionality. Finally, to evaluate the performance of the class, we will use the parameters of sensitivity, accuracy and diagnostic power. The results of classification of different levels of MS disease with MLP classification showed that the accuracy, sensitivity and diagnostic power of MLP classification were 93.58%, 92.92% and 90.45%, respectively.

نویسندگان

Hamidreza Mirzaei

Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran,

Fereydoon Nowshiravan Rahatabad

Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran,

Nader Jafarnia Dabanloo

Department of Biomedical Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran