Hybrid Deep Learning Methodology for the Detection of Lung Cancer: A Four-Step Methodology

Lung cancer is a significant health concern affecting both males and females, leading to a rise in mortality rates. Computer-aided detection systems have emerged as valuable tools for assisting in the diagnosis and management of lung cancer. This research presents a novel hybrid deep learning model that combines DenseNet and Bidirectional Long Short-Term Memory (BiLSTM) networks for improved lung cancer detection. The model was evaluated using a publicly accessible CT imaging dataset (IQ-OTH), demonstrating exceptional performance with a success rate of 97%, sensitivity of 95%, and specificity of 96%. The proposed approach offers a promising solution for enhancing the accuracy and efficiency of lung cancer diagnosis. By developing and validating such algorithms, we can contribute to the development of software suites for routine pathological procedures, alleviating the burden on healthcare professionals.