论文部分内容阅读
Aiming to feature relevant and redundancy problem in MRI Prostate Tumor ROI high dimension representation,a model,Prostate Tumor CAD Model Based on NN with PCA feature-level fusion in MRI,is proposed in this paper.Firstly,102 dimension features are extracted form MRI prostate tumor ROI,including 6 dimension geometry features,6 dimension statistical features,7 dimension Hu invariant moment features,56 dimension GLCM texture features,3 dimension TAMURA texture features and 24 dimension frequency features;Secondly,8 dimension features are obtained by PCA in Cumulative contribution rate 89.62%,and reducing the dimension of the feature vectors;Thirdly neural network is regarded as Classifier to classify with BFGS,Levenberg-Marquardt,BP and GD training algorithm,Finally,180 MRI images of prostate patients are regarded as original data,prostate tumor CAD model based on NN with feature-level fusion are utilized to aid diagnosis.Experiment results illustrate that the ability to identify benign and malignant prostate tumor are improved at least 10%through Neural network with PCA feature-level fusion,and the strategy is effective,relevant and redundancy among features are reduces in some degree.There is positive significance for MRI prostate Tumor CAD.