Lung Nodule Detection And Classification Using Image Processing Techniques,Principal Component Analysis And Artificial Neural Network

The detection of lung cancer disease in early time increases the survival rate for lung cancer patients.Effective ways for predict and treat lung cancer remain challenges due to lack of effective ways of detection the lung nodules which causes by their arbitrariness in shape,size and texture.In this paper,image processing is used for image pre-processing,image segmentation and feature extraction.Artificial neural network(ANN)has been employed to learn extracted feature for nodule detection such as shape,size and volume.While principal component analysis was employed for multivariate data processing,it used to detect the complexity of interrelationships between diverse patient,disease and treatment variables.MATLAB have been used for all procedure in image processing and artificial neural network.XLSTART software was used for principal component analysis.The lung cancer database which contains the images divide the lung cancer into two kinds:1)Normal with no nodule and 2)nodule image such as benign or malignant.Therefore,by using the proposed method the accuracy obtained was 97%.