Detecting brain tumor in computed tomography images using Markov random fields and fuzzy C-means clustering techniques

• Main Authors: Abdulbaqi, Hayder Saad, Department of Physics, College of Education, University of Al-Qadisiya, Al-Qadisiya, Jafri, Mohd Zubir Mat, Omar, Ahmad Fairuz, Mustafa, Iskandar Shahrim Bin, Abood, Loay Kadom
• Format: Conference Proceeding
• Language: English
• Subjects: ANIMAL TISSUES; BRAIN; Brain cancer; CAT SCANNING; CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSICS; Clustering; COMPARATIVE EVALUATIONS; Computed tomography; DETECTION; DIAGNOSIS; Fields (mathematics); FUZZY LOGIC; HUMAN POPULATIONS; Image detection; IMAGE PROCESSING; Image segmentation; MARKOV PROCESS; Markov processes; Medical imaging; NEOPLASMS; PATIENTS; RANDOMNESS; STANDARD OF LIVING; Tumors
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/1.4915191

Brain tumors, are an abnormal growth of tissues in the brain. They may arise in people of any age. They must be detected early, diagnosed accurately, monitored carefully, and treated effectively in order to optimize patient outcomes regarding both survival and quality of life. Manual segmentation of brain tumors from CT scan images is a challenging and time consuming task. Size and location accurate detection of brain tumor plays a vital role in the successful diagnosis and treatment of tumors. Brain tumor detection is considered a challenging mission in medical image processing. The aim of this paper is to introduce a scheme for tumor detection in CT scan images using two different techniques Hidden Markov Random Fields (HMRF) and Fuzzy C-means (FCM). The proposed method has been developed in this research in order to construct hybrid method between (HMRF) and threshold. These methods have been applied on 4 different patient data sets. The result of comparison among these methods shows that the proposed method gives good results for brain tissue detection, and is more robust and effective compared with (FCM) techniques.