<p>The timely detection of lung cancer is vital to enhance the outcome of patients, which causes high mortality rates. This study presents a novel deep learning model, Dual Convolutional Autoencoder Transformer encoder with Swin Transformer, designed for efficient lung cancer detection. The framework involves data acquisition from five various datasets, followed by pre-processing techniques including grayscale conversion, normalization, and resizing. The proposed model integrates Convolutional Autoencoders for effective feature extraction and Swin Transformers for capturing extended-range dependencies. A dual-channel attention mechanism is utilized by the proposed model to improve attention to critical features in medical images. The inclusion of Swin Transformer and dual channel attention enhanced both feature extraction and attention mechanisms, resulting in better model robustness. The performance of the proposed model is analyzed by employing various metrics such as accuracy, precision, recall, specificity, and F1 score. The proposed model achieves an average accuracy of 98.75% and a precision of 98.44% across five datasets, demonstrating robust and constant performance. The proposed model provides a considerable improvement in the detection of lung cancer, which provides a reliable and efficient tool for early diagnosis.</p>