EGP-Net: a lung nodule segmentation network integrating edge guidance and pyramidal multi-scale contextual attention mechanisms

1

EGP-Net is designed for accurate segmentation of pulmonary nodules in CT images, addressing challenges like blurred boundaries and complex structures.
2

The network combines a Res2Net-50 encoder with edge guidance and multi-scale contextual attention techniques to enhance feature representation.
3

EGP-Net achieved an IoU of 88.32% and a Dice coefficient of 92.65% on the LIDC dataset, outperforming existing segmentation methods.
4

The model was evaluated on both public and private datasets, demonstrating robust performance in pulmonary nodule segmentation.
5

Ablation experiments confirmed the effectiveness of each component within the EGP-Net architecture for improved segmentation accuracy.

1