Conference Proceeding

A Computer-Aided Brain Tumor Detection Approach with Learning Vector Quantization (Lvq) Method

Mr. Berkan URAL ,
Research Assistant in Kafkas University and Gazi University, Turkey

Mr. Berkan URAL, completed his B.Sc. in TOBB ETU in the department of Electrical and Electronics Engineering. Then, he has been working as a Research Assistant in Kafkas University and Gazi University since September 2014. He is generally in the department of Circuits and Systems (Biomedical Systems) and he completed the master-M.Sc.- program in Gazi University in January 2016. He continues his doctorate-Ph.D.- program in Gazi University and he will complete the program soon.

Brain tumor is one of the major problems in transience among children and adults. Early diagnostic procedures are become crucial for detecting a brain tumor on time. Indeed, the detection of a brain tumor is a challenging problem, due to high diversity in tumor appearances and tumor boundaries. This study proposes a well-organized method that is based on Magnetic Resonance Imaging (MRI), brain image segmentation and classification approach for the diagnosis of a brain tumor non- invasively. The aim of the study is to detect and to identify the tumor formation in the brain with using image processing techniques and Learning Vector Quantization (LVQ) method, respectively. In this application, firstly, K-means clustering technique which is integrated with Fuzzy C-means algorithm is used on the MRI images. Secondly, with using thresholding and level-set segmentation, tumor areas are detected with a high accuracy ratio. Thirdly, automatic brain tumor classification and analysis are achieved with using LVQ method.

Published: 11 May 2017