Brain Image Analysis by Using Support Vector Machine (Svm) Method

Abstract

Research on image analysis of the brain with Support Vector Machine method had been carried out through image processing using matlab software. The CT Scan image data was entered using the “imread” command then pre-processing was performed to eliminate noise with the mean method (multiplication operation between matrix components). After the pre-processing, image histogram analysis was carried out to find out the distribution of pixel intesity values on the image. The final step was calculating the mean and standard deviations specified with all pixel value calculation on every brain image. Value calculation the mean and standar deviation were carried out to 7 normal images and 7 brain tumor images. Mean and standar deviation calculation were used as a parameter to determine normal class and brain tumor class with SVM classification. The calculation results of he normal images have mean values with the range from 65.45 until 88.50 and standard deviation values with the range from 44.41 until 49.50, where as the brain tumor images have mean values with range from 69.46 until 93.29 and standard deviation values with range from 50.05 until 61.60. Therefore, the results of this study show that the mean and standar deviation of brain tumor images is greater than the normal images.