Lung Nodule Classification Using MobileNet Transfer Learning

Lung cancer is one of the deadliest forms of cancer, responsible for millions of deaths worldwide each year. The detection and classification of lung nodules can be challenging due to the small size and obscured locations of nodules, especially in early-stage cancers. To address this challenge, this study proposes an automatic lung nodule classification method that utilizes deep learning (DL) techniques to improve accuracy and diagnosis time. The study’s primary goal is to outperform current methodologies in detecting malignant lung nodules and predicting lung cancer. To achieve this objective, different DL models, including CNN, VGG16, and MobileNet, were evaluated for their efficacy in a classification task. The results showed that using transfer learning with MobileNet achieved the highest accuracy rate of 99.11%. The proposed technology can help medical personnel locate and diagnose lung nodules early, allowing for faster intervention and potentially better patient outcomes. Overall, this research emphasizes the potential of MobileNet transfer learning in facilitating the diagnostic process of lung nodules, which can be critical for accurate diagnosis and treatment planning.