Improved Butterfly Optimization Algorithm Based Conventional Neural Network for Lung Cancer Detection in Internet of Things

Abstract

The rapid expansion of the Internet of Things (IoT) and cloud computing in the previous decade has propelled healthcare services to new heights. Simultaneously, lung cancer is recognized as a serious illness that contributes to an annual increase in worldwide mortality rates. Using IoT-based lung disease data, this research presents a lung cancer prediction module through a conventional neural network (CNN) classifier. The importance of the proposed module is based on the optimum tuning of the CNN classifier's weights utilizing the proposed Improved Butterfly optimization algorithm (IBOA), which has better convergence criteria in the global optimal solution. The efficiency of the proposed model is evaluated using performance measures, such as accuracy, precision, recall, and F1-measure, which yield 93.8563 %, 95.3264 %, 94.4036 %, and 94.419 %, respectively. When compared to traditional approaches, the experimental results obtained utilizing the proposed IBOA-CNN module confirms the model's effectiveness in effective lung cancer diagnosis.