Performance Improved PSO based Modified Counter Propagation Neural Network for Abnormal MR Brain Image Classification

Abnormal Magnetic Resonance (MR) brain image classification is amandatory but challenging task in the medical field. Accurate identification ofthe nature of the disease is highly essential for the successful treatmentplanning. Automated systems are highly preferred for image classificationbecause of its high accuracy. Artificial neural networks are one of the widelyused automated techniques. Though they yield high accuracy, most of theneural networks are computationally heavy due to their iterative nature. Lowspeed neural classifiers are least preferred since they are practically nonfeasible. Hence, there is a significant requirement for a neural classifier whichis computationally efficient and highly accurate. To satisfy these criterions, amodified Counter Propagation Neural Network (CPN) is proposed in this workwhich proves to be much faster than the conventional network. For furtherenhancement of the performance of the classifier, Particle Swarm Optimization (PSO) technique is used in conjunction with the modified CPN. Experimentsare conducted on these classifiers using real-time abnormal images collectedfrom the scan centres. These three types of classifiers are analyzed in terms ofclassification accuracy and convergence time period. Experimental results showpromising results for the PSO based modified CPN classifier in terms of theperformance measures