Segmentation and Classification of Lung Tumor Analysis using LU-Net with BBH Optimizer

Lung cancer is a significant worldwide health issue that requires the development of efficient and rapid diagnostic techniques to improve patient outcomes and overall health. Significant progress has been made in the field of medical image processing in recent years, owing to the use of advanced deep learning techniques such as Convolutional Neural Networks (CNNs). The paper used CNNs in the work to investigate a novel strategy for identifying and classifying lung cancer, processing on two large datasets: the Lung Image Database Consortium Infectious Disease Research Institute (LIDC-IDRI) and the Lung Nodule Analysis (LUNA 16). The study began with image preprocessing to improve image quality, which included Bayesian threshold-based Taylor series (BT-based TS) denoising. The major emphasis was on lung tumor segmentation, The Behavior-based Herding Optimization (BBHO) approach was used to fine-tune classification parameters, resulting in a remarkable Accuracy rate of 0.994 and a decrease in false negatives.