Intelligent phenotyping system for chronic obstructive pulmonary disease based on three-dimensional reconstruction of airway tree
By constructing a gas-blood coupling topology graph and utilizing dual graph neural networks and implicit neural field generation techniques, we have achieved a synchronous quantitative assessment of airway and vascular lesions, solving the phenotypic confusion problem caused by ignoring the vascular tree in existing systems, and providing a more accurate phenotypic classification of chronic obstructive pulmonary disease.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE FOURTH AFFILIATED HOSPITAL OF GUANGZHOU MEDICAL UNIV (GUANGZHOU ZENGCHENG DISTRICT PEOPLES HOSPITAL)
- Filing Date
- 2026-04-27
- Publication Date
- 2026-07-17
AI Technical Summary
Existing intelligent phenotyping systems for chronic obstructive pulmonary disease (COPD) only model and analyze the airway tree separately, ignoring the closely coupled pulmonary vascular tree. This results in the classification results failing to accurately reflect the full picture of the disease and the potential risk of progression, thus affecting the accuracy of individualized treatment decisions.
A gas-blood coupling topology map is constructed, and dual graph neural networks and implicit neural field generation technology are used to achieve joint characterization and synchronicity quantification of the spatial co-occurrence patterns of airway and vascular lesions, generating coupled enhanced chronic obstructive pulmonary disease phenotypic classification results.
By integrating pathological information from the airway and vascular tree, the system accurately distinguishes between synchronous and decoupled remodeling regions, providing multi-dimensional phenotypic classification results. This overcomes the phenotypic confusion of existing systems and provides quantitative assessment data for clinical practice.
Smart Images
Figure CN122417367A_ABST