Leveraging CNN and SVM for Lung Cancer Prediction: A Dual-Class Classification Approach

This study gives a detailed analysis of lung cancer detection using a hybrid model that combines support vector machines (SVM) and convolutional neural networks (CNN). This research aims to categorize lung scans into two categories: healthy and ill. This will provide early detection and personalized therapy options for individuals with lung cancer. The model is trained and tested for prediction performance using a large dataset of X-ray images and computed tomography (CT) scans. The results show how well the suggested CNN-SVM hybrid model distinguishes between healthy and sick lungs. With accuracy rates of 92.68% and 94.43%, recall rates of 94.41% and 92.71%, and F1-scores of 93.54% and 93.56% for healthy and sick lungs, respectively, the model achieves exceptional precision, recall, and F1-score values for both classes. The model accurately and consistently differentiates between malignant and non-malignant lung tissues. The weighted and macro average performance measures performed well throughout the dataset, with average accuracy, recall, and F1-score values of 93.56%, 93.56%, and 93.55%, respectively. The micro average accuracy, recall, and F1-score of 93.55% provide further evidence of the model's stability and usefulness. The study's overall results highlight the efficacy of CNN-SVM hybrid models for lung cancer detection. It aims to support the development of enhanced diagnostic tools to improve the prognosis and outcomes of lung cancer patients by combining deep learning and traditional machine learning methodologies.