M3Net: A Macro-to-Meso-to-Micro Clinical-inspired Hierarchical 3D Network for Pulmonary Nodule Classification

2026-05-12

AI Summary

This study addresses the challenge of classifying benign and malignant pulmonary nodules in CT images by proposing a three-stage 3D deep network inspired by radiologists’ hierarchical diagnostic workflow. The architecture sequentially models macroscopic, mesoscopic, and microscopic nodule characteristics, integrating local structural details, semantic cues, and global spatial relationships. Innovatively embedding clinical hierarchical reasoning into the network design, the method employs multi-scale inputs, scale-specific encoders, and mutual information maximization across cross-scale latent spaces to enforce semantic consistency, thereby enhancing model interpretability and reliability. Evaluated on the LIDC-IDRI and USTC-FHLN datasets, the proposed approach achieves classification accuracies of 86.96% and 84.24%, respectively, outperforming current state-of-the-art methods.

Abstract

The accurate classification of benign and malignant pulmonary nodules in CT scans is critical for early lung cancer screening, yet remains challenging due to the multi-scale and heterogeneous nature of pulmonary nodules. While deep learning offers potential for auxiliary diagnosis, most existing models act as "black boxes", lacking the transparency and explainability required for trustworthy clinical integration. To address this issue, we propose M3Net, a novel 3D network for pulmonary nodule classification inspired by the hierarchical diagnostic workflow of radiologists, which integrates multi-scale contextual information from fine-grained structures to global anatomical relationships. Our framework constructs a progressive multi-scale input, from fine-grained nodule structures to local semantics and global spatial relationships. M3Net employs scale-specific encoders and ensures cross-scale semantic consistency through latent space projection and mutual information maximization. Extensive experiments on the public LIDC-IDRI dataset and a self-collected clinical dataset (USTC-FHLN) demonstrate that our method achieves state-of-the-art performance, with accuracies of 86.96% and 84.24% respectively, outperforming the best baseline by 3.26% and 2.17%. The results validate that M3Net provides a more robust and clinically relevant solution for pulmonary nodule classification. The code is available at https://github.com/jylEcho/M3-Net.

Problem

• pulmonary nodule classification
• lung cancer screening
• multi-scale heterogeneity
• model explainability
• clinical trustworthiness

Innovation

• multi-scale learning
• hierarchical 3D network
• cross-scale consistency
• pulmonary nodule classification
• clinical interpretability