Lung Nodule Classification on Computed Tomography Using Neural Networks
Lung cancer is a leading health issue and the major cause of death among all types of cancers. CT scanning is the popular method for lung cancer diagnosis detection. Manual processing of tomograms take long time for diagnosis. It is not an easy task. This complex work can also reduce the quality of diagnosis. Machine learning and neural network algorithm can be used to automatically process X-ray pictures, tomograms and PET images to detect diseases. The goal of the proposed work is to find any abnormal thing in lungs. Convolutional neural network is trained to classify abnormal area from the normal cells. The detection algorithm is designed to determine the existence of cancer in tomography images and validation, training and testing using CT images. The proposed work investigates the performance of classifier by training algorithm with morphological feature extraction. The performance results shows that proposed method achieves higher accuracy than existing methods.