Research on Lung Cancer Classification Based on Deep Learning

Lung cancer (LC), ranking first in both incidence and mortality rates globally among malignant tumors, underscores the significance of its prevention and treatment. As medical imaging becomes increasingly crucial in the diagnosis of lung cancer, there is an urgent need for efficient and precise image classification methods. To address resource-constrained scenarios, convolutional autoencoders are employed to deeply mine abstract feature representations of images, capturing feature dimensions that contain critical information. Subsequently, the features extracted by the autoencoder are input into a lightweight MobileNet V3-Small model for image classification tasks. Experimental results demonstrate that this method outperforms traditional methods in terms of accuracy, especially in resource-limited environments.