Early Studies of Lung Tumor Detection and Diagnosis Using Neural Network Backpropagation and Fuzzy Logic Systems

Abstract

Malignant lung tumor is a deadly disease for both men and women, but if lung tumor identification results are known correctly in the early stages it can be treated easily without major risk because the tumor is still small. In general, the chest x-ray method is highly believed to be the initial detection process of lung tumors but serious errors in the case of the diagnosis give bad results and can lead to death. In order to know the type of tumor early and correctly required an application program of lung tumor detection system. The purpose of this study was to obtain a result of identification of early lung tumors using backpropagation neural networks and fuzzy logic systems to assist radiologists in diagnosing patient illnesses. This study used lung x-ray image object, 40 samples for training and 5 samples for artificial neural network test. The method used is the process of digital image processing from RGB image conversion to gray level; enhancement through spatial filtration and frequency; and segmentation through thresholding and morphological techniques, then continued identification and diagnosis using backpropagation neural networks and fuzzy logic systems. The backpropagation neural network has been successfully educated with 40 samples using the Matlab application program. Based on the comparison between the calculated value of artificial neural network backpropagation with the measurement results with Matlab program, it can be concluded artificial neural network backpropagation has successfully identified 5 samples well. Then from the process of fuzzy logic system of 5 identified samples, has been known malignant type and early stages of each of the 5 samples of the cancer object.