Paper Details: Downloads: 1157
Serial Number: P1121526388
Title: Expert System Development for the Fuzzy ANN Based Diagnosis of Brain Tumor
Authors: Nandita Pradhan and A.K. Sinha
Abstract: A novel method for the development of an Expert System for the diagnosis of Brain Tumor has been presented in this paper. The knowledge has been acquired from clinical symptoms, neurological tests, cerebrospinal fluid (CSF) tests and feature vectors extracted from fluid attenuated inversion recovery (FLAIR) brain magnetic resonance (MR) images of the brain tumor patients from sizeable number of samples and data collected from different hospitals and nursing homes. Further each symptom case has nine most significant symptoms; each neurological examination case has seven most significant tests results; each CSF test has five parameter tests results and for MR brain image processing three empirically developed higher order wavelet and statistical functions are used and fuzzy C means algorithm is used for segmentation of intracranial image. Knowledge base thus acquired from symptoms, neurological tests, CSF tests results and from image processing are used for machine learning using fuzzy artificial neural network and expert system for brain tumor diagnosis is developed with an unique integrated approach. The results of proposed system are very promising and it is found that it can be effectively applied for automated tumor diagnosis. Final result depicted that 91.11% of correct diagnosis is obtained for brain tumor.
Keywords: Brain Tumor, Clinical Symptoms and Neurological Tests, CSF Tests, Magnetic Resonance Images, Fuzzy Artificial Neural Network
Journal/Conference: International Journal of Artificial Intelligence and Machine Learning
Volume: 15
Issue: 1
Submission Date: 6/22/2015 12:00:00 AM
Review Date: 7/20/2015 12:00:00 AM
Publishing Date: 9/26/2015 12:00:00 AM
Article Downloads: 1157
Download:

Facebook