Meeting Banner
Abstract #3722

An MR Brain Image Classifier System Via Particle Swarm Optimization and Kernel Support Vector Machine

Yudong Zhang1, 2, Chuanmiao Xie, 13, Bradley S. Peterson1, 2, Zhengchao Dong1, 2

1Columbia University, New York, NY, United States; 2New York State of Psychiatric Institute, New York, NY, United States; 3Department of Medical Imaging & Interventional Radiology, Sun Yat-Sen University, Guangzhou, Guangdong, China

We proposed a novel hybrid system to classify an MR brain image as either normal or abnormal. The method employed digital wavelet transform to extract features and used principal component analysis to reduce the dimensionality of the feature space. Afterwards, we constructed a kernel support vector machine with Radial Basis Function kernel, using particle swarm optimization to optimize the parameters in the training function. We tested the method with a dataset of 90 brain images consisting of 17 diseases. A 5-fold cross validation showed that our method achieved 97.78% classification accuracy, higher than 86.22% by BP-NN and 91.33% by RBF-NN.

Abstract PDF

Keywords

abnormal abnormality accuracy achieved addition aids altos applicable applications applying approximation arrhythmia automated axial back basis best bootstrap brain brains carcinoma cerebral chaos check china chose classification classifier classifiers classify clinical common component computational conditions confusion consists constraint construct correctly cross dataset dementia detection determine diagnosis diagnostic digital dimensionality disease diseases distinguish dong effectively efficiency either encephalopathy excellent expert extends extract fault feature features fitness flowchart fold folded form function grossly herpes hidden home hybrid hypothesize improve improved independently institute introduced kernel layer lesions machine machines made matched matrix medical metastatic methodology misclassified motor multipliers network networks neural neuron nonlinear novel numerous optimization original paper park particle performance pick plus principal problem proceedings product programming programs propagation proposed psychiatric quadratic radial radiology reduce replace resolution sarcoma scalar school sclerosis sensitivity sensor simplified solve specificity submitted summarized support swarm system systems tools train trained training universal validation vector visible wavelet website widely