Hybrid Computational Intelligence Techniques for Automated Lung Cancer Detection

lung diseases like pneumonia, tuberculosis, and lung cancer remain a major global health problem. Early and precise diagnosis is vital to improving patient outcomes and ensuring effective treatment. Thus, we are proposing a unique hybrid approach that combines swarm intelligence-based optimization techniques and attention-based deep learning to achieve better identification of lung disease. The hybrid framework that we proposed contains a Contrast Multiscale Binary Pattern for preprocessing, an Augmented Graph Bee Colony algorithm for spatial relationships in medical images, and Particle Swarm Optimization with ResNet R-network to find small and subtle abnormality in lung structures. Elitist Ant Colony Optimization is then used to achieve optimization for feature selection and features optimization. InceptionNet is finally used to classify lung cancer. The experimental results show that our proposed deep CNN models represent a huge improvement over the accuracy and computational efficacy than the more conventional deep learning method or model for lungs disease identification purposes.