Hy_PSO: Hybrid Algorithm for Lung Cancer Diagnosis and Prognosis

As one of the primary causes of cancer-related death globally, lung cancer demands the development of sophisticated diagnostic and prognostic instruments in order to enhance patient outcomes. Lung cancer stands as one of the malignancies with significant morbidity and mortality rates. Imaging serves as a crucial component throughout the entire spectrum of lung cancer management, facilitating everything from initial detection to evaluating response to treatment. Innovations in medical imaging and image processing have created new opportunities for enhancing the diagnosis and prognosis of lung cancer. This study introduces a hybrid algorithm, termed the Hy_PSO algorithm, which combines the strengths of particle swarm optimization and convolutional neural networks (CNNs) for classification tasks. Feature selection optimization significantly influences the accuracy of the classification model. Comparative analysis with other classification techniques, including support vector machines, artificial neural networks (ANNs), and CNNs, demonstrates the superiority of the proposed algorithm. Experimental results indicate its effectiveness in automatic lung tumor detection, surpassing contemporary methods in terms of accuracy (98.4%), sensitivity (99%), and specificity (97.65%). Hy_PSO combines many Machine Learning (ML) methods, including support vector machines (SVM), decision trees, and neural networks, to create prediction models after determining the ideal feature set. To find promising biomarkers for targeted medicines, estimate patient survival rates, and categorize lung cancer stages, these models are trained using the optimal characteristics. By utilizing the advantages of both PSO and machine learning, the hybrid technique makes sure that the prediction models are more robust and generalizable.