Brain tumor detection using GLCM features and classifiers

• Main Authors: Sanjayprabu, S., Kumar, R. Sathish, Somasundaram, K., Karthikamani, R.
• Format: Conference Proceeding
• Language: English
• Subjects: Abnormalities; Brain; Brain cancer; Classifiers; Diagnostic systems; Feature extraction; Magnetic resonance imaging; Medical imaging; Parameter sensitivity; Principal components analysis; Tumors
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0125260

The brain is one of the important organs in the human body. It controls functions such as vision, hail, memory, knowledge, personality, and difficulty at work. Numerous health associations have rated brain tumors as an alternative major imbalance that causes the highest number of deaths worldwide. Diagnosis of an undiagnosed brain tumor provides an opportunity for effective medical treatment. At diagnosis, a brain tumor shows the growth of extra cells in the brain, some of which can lead to cancer. The usual system for describing brain tumors is magnetic resonance imaging (MRI). Information about the development of abnormal cells in the brain is obtained from MRI images. The brain abnormalities are detected using computer-aided diagnostic systems. The paper aims to classify the brain tumors from MRI images using the GLCM feature extraction technique and five classifiers. The standard parameters like sensitivity, selectivity, and accuracy are used to compare the classifier performance. In this work, the hybrid classifier Expectation with principle component analysis classifier obtained the best accuracy of 97.66 % compared to other classifiers.