Lung Cancer Prediction Using CNN Algorithm

Abstract

Lung cancer remains one of the leading causes of cancer-related mortality worldwide, making early diagnosis 
essential for improving patient survival rates and treatment effectiveness. Traditional diagnostic procedures rely 
heavily on radiological expertise and manual interpretation of Computed Tomography (CT) scan images, which 
can be time-consuming and susceptible to human error. This paper presents an intelligent deep learning-based 
framework for automated lung cancer detection and classification using CT scan images. The proposed system 
incorporates image preprocessing techniques to enhance image quality and reduce noise, followed by lung region 
segmentation to identify relevant areas of interest. Feature extraction is performed to capture critical texture and 
morphological characteristics associated with pulmonary nodules. 
The extracted features are analysed using machine learning and deep learning techniques, including Support 
Vector Machine (SVM), Random Forest (RF), Artificial Neural Network (ANN), and Convolutional Neural 
Network (CNN), to classify lung nodules as benign or malignant. The framework is trained and evaluated using 
publicly available lung cancer datasets, ensuring reliable and reproducible results. Performance evaluation is 
conducted using standard metrics such as Accuracy, Precision, Recall, F1-Score, and Confusion Matrix analysis. 
Experimental results demonstrate that the proposed CNN-based approach achieves superior classification 
performance compared to conventional machine learning methods, providing more accurate and consistent 
predictions. The proposed framework offers a scalable and efficient computer-aided diagnostic solution that can 
assist healthcare professionals in the early detection of lung cancer, ultimately supporting timely clinical decision
making and improving patient outcomes.