1M.Tech Student, Department of Computer Science Engineering, UIT, Rabindranath Tagore University, Bhopal
2Associate Professor, Department of Computer Science Engineering, UIT, Rabindranath Tagore University, Bhopal
3HoD, & Asst. Professor, Department of Computer Science Engineering, UIT, Rabindranath Tagore University, Bhopal
The unlimited and uncontrollable growth of cells can cause human brain tumors. Correct treatment and early diagnosis of brain tumors are essential to avoid permanent damage to the brain. In medical image diagnosis, tumor segmentation and classification schemes are used to identify tumor and non-tumor cells in the brain. The automatic classification is a challenging task which utilizes the traditional methods due to its more execution time and ineffective decision making. To overcome this problem, this research proposes an automatic tumor classification method named as Hybrid Kernel based Fuzzy C-Means clustering -Convolution Neural Network (Hybrid KFCM-CNN) method. The algorithm identifies the position of tumor in brain MRI as they are mostly preferred for tumor diagnosis in clinic. The proposed method also crops tumor region from segmented image and way growth of tumor and help in treatment planning. It also provides important information about location, dimension and shape of brain tumor region with no exposing the enduring to a high ionization radiation. The size of tumor is calculated in term of number of pixels. Similarly the primary brain tumor is considered into Benign and malignant type on MRI brain images, based on accuracy, sensitivity, specificity in MATLAB simulation
KFCM, CNN, Feature Extraction, MRI Image, GLCM
