Integrating Domain Knowledge into Deep Networks for Lung Cancer Prediction

Lung cancer is one of the leading causes of cancer-related deaths worldwide. Early detection plays a crucial role in improving survival rates; however, traditional manual diagnosis using CT scan images is time-consuming and prone to human error. This paper presents an automated lung cancer detection system using ensemble machine learning algorithms and deep neural networks. The proposed system integrates preprocessing, lung segmentation, feature extraction, and classification to improve diagnostic accuracy. Ensemble models combining Decision Tree, AdaBoost, and Multi-Layer Perceptron (MLP) classifiers were implemented along with an RBF-based classifier for CT image prediction. Experimental results demonstrate improved accuracy and reduced diagnosis time compared to conventional methods. The system assists radiologists in making accurate and reliable clinical decisions.