Towards Real-Time Lung Cancer Diagnosis: A Deep Learning Approach with Optimized CNN for CT Scan Analysis

Lung cancer continues to be among the deadliest illnesses globally, calling for early and precise diagnosis for better patient outcomes. This research introduces an automated system of lung cancer diagnosis via deep learning and medical imaging. A Convolutional Neural Network (CNN) model was constructed and trained using a balanced data set of 200 CT scan images, with 100 malignant and 100 benign cases. The dataset was preprocessed by reducing noise, enhancing contrast, and augmenting to enhance robustness. The CNN model consisted of four layers of convolution for feature extraction, max-pooling layers to reduce dimensions, and fully connected layers for classification. The model was trained with the Adam optimizer and trained for more than 30 epochs. Performance on the test set achieved accuracy of 92.5 %, precision of 91.2 %, recall of 93.8 %, and an F1-score of 92.5 %, validating its performance in identifying malignant versus benign cases. Comparison with VGG16 and ResNet50 verified better performance, and mean processing time of 2.5 seconds per image assured applicability in real-time systems. Error analysis revealed slight misclassifications from overlapping features and tiny tumor sizes. The research emphasizes the promise of AIbased diagnostic instruments in medical imaging and recommends improvement in the future, such as training on large datasets and adding explainable AI for better interpretability and clinician adoption.