Diagnosis of Multiple Sclerosis Based on Image Processing of Patient Motion

In recent years, brain injuries, often known as the silent epidemic, are among the growing diseases that require special attention. The purpose of this study was to provide an efficient and low- cost method for the automatic diagnosis of multiple sclerosis. According to the importance of this disease, its gravity, and the influence of it on the function of Fred's life, and the decrease in the quality of Wilmore's life, we are looking for the less aggressive methods of communication, and the less aggressive, and the computer help we have in helping the most valuable methods of communication. To do this, ۳۰ individuals including ۱۸ healthy individuals and ۱۲ individuals with multiple sclerosis were used to record video sequences that performed ۸ exercises recommended by the British Multiple Sclerosis Patients Association three times. After recording video sequences, the desired time and location features are extracted. After selecting the appropriate features, they are used for classification by neural network and decision tree. The results show that proposed method in this paper has at best ۹۶.۷۳ % accuracy. The accuracy percentage is the evidence of good categorization of experimental data them.

Keywords:

Multiple Sclerosis , Automatic Diagnosis , Decision Tree Classification , Neural Network Classification