Predicting Lung Cancer Disease Using Optimized Weighting-Based Enhanced Neural Network Classification

Lung cancer is a highly lethal disease, claiming the lives of approximately 1 million individuals annually. Cancer is characterized by abnormal and rapid cell growth, making it difficult to control, though early diagnosis through Computed Tomography (CT) scan images can significantly improve chances of survival. However, detecting lung cancer early is challenging and can increase the risk of complications such as infection, inflammation, and tumor growth in the lungs. Moreover, the current methods for analyzing lung cancer prognosis have low accuracy. To solve this problem, we proposed an Optimized Weighting-Based Enhanced Neural Network (OWENN) method to classify patients based on attributes accurately. Furthermore, they utilize the Adaptive Median Filter (AMF) technique during image pre-processing to calculate the mean value of each pixel and remove noise. Moreover, Improved Particle Swarm Optimization (IPSO) can be implemented to extract tumors from lung images efficiently. Finally, the OWENN method improves the accuracy of cancer or non-cancer detection by classifying patients based on their selected attributes. Afterward, the experimental results indicate that the suggested OWENN approach can attain lung cancer prediction by evaluating sensitivity, precision, accuracy, and time delay.