Leveraging MobileNet V2 for swift and precise identification of lung and colon cancers

Lung and colon cancers pose a significant global health challenge. Early detection is crucial for improved patient outcomes. This study investigates Applying deep learning to differentiate lung and colon cancer subtypes using the LC25000 dataset. The MobileNetV2 architecture was selected due to its efficiency and accuracy. The dataset was divided into five categories: Colon adenocarcinoma, non-neoplastic colon tissue, Malignant lung adenocarcinoma, Malignant lung squamous cell carcinoma, and non-neoplastic lung tissue. Data augmentation techniques were applied to enhance model performance. The model achieved impressive accuracies of 97.57%, 98.00%, and 98.3% on the training, validation, and testing sets, respectively. These results highlight the potential of deep learning for rapid, accurate, and minimally invasive cancer detection, surpassing traditional methods. Our findings contribute to the development of more efficient and accessible cancer screening approaches.