One of the main reasons for mortality in the globe is lung cancer. CT scans are mostly used in early detection of lung cancer, which is crucial for helping individuals survive. In this study, we provide a deep learning model that automatically detects lung cancer and determines its stage from CT scan images using deep learning. To create a hybrid deep learning model that increases accuracy and can identify both image features and data changes over time, we combine CNN and CNN-LSTM models. This hybrid model outperforms previous techniques enhancing detection precision while minimizing diagnostic errors. The basic CNN model already gives excellent results for two-category (binary) classification, but when we add sequence learning with LSTM, the CNN-LSTM model performs even better in terms of accuracy. The technique not only classifies but also segments tumors and measures aberrant areas to categorize cancer into Stage 1, 2, or 3 according to the quantity and size of lesions found. The Tkinter package in Python is used to create a comprehensive graphical user interface (GUI) that allows end users to inspect confusion matrices, train models, visualize accuracy/loss trends, upload datasets, and forecast cancer from previously unseen test images. A Python program called Tkinter is used to create a straightforward and intuitive graphical user interface (GUI). It was selected due to its ease of use, minimal setup requirements, and ability to quickly create functional models. This research requires a simple and functional interface without the use of bulky or complicated software can benefit from Tkinter. Visual interpretability and stage prediction are incorporated for clinical relevance. This study presents a deep learning-based technique for early detection of lung cancer and demonstrates the potential of deep learning in medical imaging.