Virtual nasal cavity populations for flow prediction with distributed graph convolutional neural networks

Physics of Fluids, vol. 38

Abstract

Nasal air resistance is a key indicator of respiratory health and is essential for understanding nasal physiology and functions. Accurately measuring this quantity, however, remains challenging both experimentally and computationally. Data-driven methods, particularly deep learning models, offer a promising avenue for the rapid and reliable prediction of flow features, but they require large and diverse training datasets to generalize effectively to unseen cases. This study has two primary objectives: first, to develop machine learning models for respiratory flow simulations capable of accurately predicting the air resistance; and second, to introduce a data augmentation strategy for generating large virtual populations from a limited number of real patient geometries. Due to the complex and unstructured nature of nasal cavity geometries, training samples are represented as graphs, allowing direct use of computational fluid dynamic simulations as model inputs. The model is implemented as a distributed graph convolutional neural network to efficiently handle large-scale datasets, demonstrated here with 8000 graphs and scalable to even larger populations. Results show that the model achieves an R2 score of 0.999 in predicting the pressure drop, and that the prediction error on unseen cases decreases substantially as the virtual population is expanded from a limited set of real geometries.

Authors 8

1. Hadrien Calmet

   Barcelona Supercomputing Center

   Affiliation as printed

   Department of Computer Applications in Science and Engineering, Barcelona Super-Computing Centre (BSC-CNS) 1 , Barcelona,

2. Joan Calafell

   Barcelona Supercomputing Center · Universitat Politècnica de Catalunya

   Affiliation as printed

   Department of Computer Applications in Science and Engineering, Barcelona Super-Computing Centre (BSC-CNS) 1 , Barcelona,

   Department of Fluid Mechanics, Universitat Politècnica de Catalunya, BarcelonaTech (UPC) 2 , 08034 Barcelona,

3. Puri Rishabh

   Karlsruhe Institute of Technology

   Affiliation as printed

   KIT-Karlsruhe Institute of Technology, Engler-Bunte-Institute, Simulation of Reacting Thermo-Fluid Systems 3 , Engler-Bunte Ring 7, D-76131 Karlsruhe,

4. Benedikt Harald Johanning-Meiners Aachen Institute of Aerodynamics Chair of Fluid Mechanics

   RWTH Aachen University

   Affiliation as printed

   Institute of Aerodynamics and Chair of Fluid Mechanics (AIA), RWTH Aachen University 4 , 52062 Aachen,

5. Abel Gargallo‐Peiró

   Barcelona Supercomputing Center

   Affiliation as printed

   Department of Computer Applications in Science and Engineering, Barcelona Super-Computing Centre (BSC-CNS) 1 , Barcelona,

6. Rakesh Sarma

   Forschungszentrum Jülich · Jülich Supercomputing Centre

   Affiliation as printed

   Jülich Supercomputing Centre (JSC), Forschungszentrum Jülich GmbH (FZ) 5 , 52425 Jülich,

7. Mario Rüttgers

   Inha University

   Affiliation as printed

   Data-Driven Fluid Engineering Laboratory, Department of Mechanical Engineering, Inha University 6 , Incheon,

8. Guillaume Houzeaux

   Barcelona Supercomputing Center

   Affiliation as printed

   Department of Computer Applications in Science and Engineering, Barcelona Super-Computing Centre (BSC-CNS) 1 , Barcelona,