Journal article
Human-Airway Surface Mesh Smoothing Based on Graph Convolutional Neural Networks
Computer methods and programs in biomedicine, Vol.246, 108061
04/2024
DOI: 10.1016/j.cmpb.2024.108061
PMID: 38341897
Abstract
•We propose an airway-mesh-smoothing learning method (AMSL).•AMSL trains two graph convolutional neural networks to smooth vertex positions and face normal vectors.•Compared to the existing methods, the proposed smoothing yields improved results on public datasets.•AMSL reproduces the airway branch diameter accurately, allowing a better estimate of flow properties.
A detailed representation of the airway geometry in the respiratory system is critical for predicting precise airflow and pressure behaviors in computed tomography (CT)-image-based computational fluid dynamics (CFD). The CT-image-based geometry often contains artifacts, noise, and discontinuities due to the so-called stair step effect. Hence, an advanced surface smoothing is necessary. The existing smoothing methods based on the Laplacian operator drastically shrink airway geometries, resulting in the loss of information related to smaller branches. This study aims to introduce an unsupervised airway-mesh-smoothing learning (AMSL) method that preserves the original geometry of the three-dimensional (3D) airway for accurate CT-image-based CFD simulations.
The AMSL method jointly trains two graph convolutional neural networks (GCNNs) defined on airway meshes to filter vertex positions and face normal vectors. In addition, it regularizes a combination of loss functions such as reproducibility, smoothness and consistency of vertex positions, and normal vectors. The AMSL adopts the concept of a deep mesh prior model, and it determines the self-similarity for mesh restoration without using a large dataset for training. Images of the airways of 20 subjects were smoothed by the AMSL method, and among them, the data of two subjects were used for the CFD simulations to assess the effect of airway smoothing on flow properties.
In 18 of 20 benchmark problems, the proposed smoothing method delivered better results compared with the conventional or state-of-the-art deep learning methods. Unlike the traditional smoothing, the AMSL successfully constructed 20 smoothed airways with airway diameters that were consistent with the original CT images. Besides, CFD simulations with the airways obtained by the AMSL method showed much smaller pressure drop and wall shear stress than the results obtained by the traditional method.
The airway model constructed by the AMSL method reproduces branch diameters accurately without any shrinkage, especially in the case of smaller airways. The accurate estimation of airway geometry using a smoothing method is critical for estimating flow properties in CFD simulations.
Details
- Title: Subtitle
- Human-Airway Surface Mesh Smoothing Based on Graph Convolutional Neural Networks
- Creators
- Thao Thi Ho - Kyungpook National UniversityMinh Tam Tran - Kyungpook National UniversityXinguang Cui - Huazhong University of Science and TechnologyChing-Long Lin - University of IowaStephen Baek - University of VirginiaWoo Jin Kim - Kangwon National University HospitalChang Hyun Lee - Seoul National University HospitalGong Yong Jin - Jeonbuk National UniversityKum Ju Chae - Jeonbuk National UniversitySanghun Choi - Kyungpook National University
- Resource Type
- Journal article
- Publication Details
- Computer methods and programs in biomedicine, Vol.246, 108061
- Publisher
- Elsevier B.V
- DOI
- 10.1016/j.cmpb.2024.108061
- PMID
- 38341897
- ISSN
- 0169-2607
- eISSN
- 1872-7565
- Grant note
- DOI: 10.13039/501100021520, name: National Supercomputing Center, award: KSC-2022-CRE-0495; DOI: 10.13039/501100003725, name: National Research Foundation of Korea; DOI: 10.13039/501100003621, name: Ministry of Science, ICT and Future Planning, award: NRF-2023R1A2C2003781
- Language
- English
- Electronic publication date
- 02/06/2024
- Date published
- 04/2024
- Academic Unit
- Roy J. Carver Department of Biomedical Engineering; Radiology; Mechanical Engineering
- Record Identifier
- 9984557837902771
Metrics
18 Record Views