Performance Evaluation of an Intelligent BFGS-Based Neuroclassifier for Brain Tumour Decision Support System

Abstract

Brain Tumour is a disease that leaves its tolling effect on patients as well as relatives. Developing as a result of cancerous or non-cancerous mass or growth of abnormal cells in the brain, every efficient technique for the detection and management of the disease is worth being explored. Intelligent systems are therefore not left out of this quest. In this work, an artificial neural network trained and optimized using the modified Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm was developed to classify brain tumours as meningioma, glioma or pituitary tumour. After training and deployment, the network attained an accuracy of 98.6% for glioma classification, 97.10 % for meningioma and 97.89% for pituitary tumour classification. The performance of the network shows our proposed classifier performed satisfactorily and further illustrates the importance of the deployment of intelligent techniques in image processing and medicine in general.