Abstract
This research investigates application of physically regularized deep learning for the estimation of blood flow parameters in bifurcations of human arteries. The study presents a comprehensive methodology that combines synthetic data generation and advanced neural network architectures. The initial step involves construction of 3D meshes for bifurcations achieved through the developed mesh generator based on the GMSH library. A diverse dataset is then generated on the basis of 3D blood flow simulations in bifurcations in a physiological range of parameters such as vessel radii, bifurcation angles and inlet and outlet pressures. The generic database for neural network training and testing contains approximately \(1.4\cdot 10^{5}\) data samples. The research focuses on training a feed-forward neural network, which provides the minimum possible error and physically relevant output. Two types of loss functions are explored for assessing the network’s performance, Huber loss function (HLF) and a physically regularized loss function (PRLF). We achieve the relative error of \(12\%\) for HLF. The study demonstrates that PRLF enhances the model’s capability to adhere to physical laws which provides faster convergence of training phase and decrease of the relative error to approximately \(2.5\%\) on the test dataset. We also study sensitivity of PRLF-based neural network convergence to different physical components. We found that adding symmetry property to the mass conservation condition accelerates the training process, decreases fluctuations of PRLF and reduces the relative error.
REFERENCES
T. Dobroserdova, M. Olshanskii, and S. Simakov, ‘‘Multiscale coupling of compliant and rigid walls blood flow models,’’ Int. J. Numer. Methods Fluids 82, 799–817 (2016). https://doi.org/10.1002/fld.4241
T. Dobroserdova, F. Liang, G. Panasenko, and Yu. Vassilevski, ‘‘Multiscale models of blood flow in the compliant aortic bifurcation,’’ Appl. Math. Lett. 93, 98–104 (2019). https://doi.org/10.1016/j.aml.2019.01.037
Yu. Vassilevski, M. Olshanskii, S. Simakov, A. Kolobov, and A. Danilov, Personalized Computational Hemodynamics. Models, Methods, and Applications for Vascular Surgery and Antitumor Therapy (Academic, New York, 2020).
A. Quarteroni and L. Formaggia, Mathematical Modelling and Numerical Simulation of the Cardiovascular System (Elsevier, Amsterdam, 2004). https://doi.org/10.1016/S1570-8659(03)12001-7
T. Dobroserdova, Yu. Vassilevski, S. Simakov, T. Gamilov, A. Svobodov, and L. Yurpolskaya, ‘‘Two-scale haemodynamic modelling for patients with Fontan circulation,’’ Russ. J. Numer. Anal. Math. Model. 36, 267–278 (2021). https://doi.org/10.1515/rnam-2021-0022
J. Kutz, ‘‘Deep learning in fluid dynamics,’’ J. Fluid Mech. 814, 1–4 (2017). https://doi.org/10.1017/jfm.2016.803
M. Du, C. Zhang, S. Xie, F. Pu, D. Zhang, and D. Li, ‘‘Investigation on aortic hemodynamics based on physics-informed neural network,’’ Math. Biosci. Eng. 20, 11545–11567 (2023). https://doi.org/10.3934/mbe.2023512
M. Sarabian, H. Babaee, and K. Laksari, ‘‘Physics-informed neural networks for brain Hemodynamic predictions using medical imaging,’’ IEEE Trans. Med. Imaging 41, 2285–2303 (2022). https://doi.org/10.1109/TMI.2022.3161653
C. Geuzaine and J.-F. Remacle, ‘‘Gmsh: A 3-D finite element mesh generator with built-in pre- and post-processing facilities,’’ Int. J. Numer. Methods Eng. 79, 1309–1331 (2009). https://doi.org/10.1002/nme.2579
D. Kingma and J. Ba, ‘‘Adam: A method for stochastic optimization,’’ arXiv: 1412.6980 (2014). https://api.semanticscholar.org/CorpusID:6628106
Yu. Vassilevski and K. Lipnikov, ‘‘An adaptive algorithm for quasioptimal mesh generation,’’ Comput. Math. Math. Phys. 39, 1468–1486 (1999). https://www.mathnet.ru/eng/zvmmf/v39/i9/p1532
P. R. Amestoy, I. S. Duff, J.-E. L’Excellent, and J. Koster, ‘‘A fully asynchronous multifrontal solver using distributed dynamic scheduling,’’ SIAM J. Matrix Anal. Appl. 23, 15–41 (2001). https://doi.org/10.1137/S0895479899358194
G. Bebis and M. Georgiopoulos, ‘‘Feed-forward neural networks,’’ IEEE Potentials 13 (4), 27–31 (1994). https://doi.org/10.1109/45.329294
N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, ‘‘Dropout: A simple way to prevent neural networks from overfitting,’’ J. Mach. Learn. Res. 15, 1929–1958 (2014). https://api.semanticscholar.org/CorpusID:6844431
N. Sridhar, ‘‘The generalized sigmoid activation function: Competitive supervised learning,’’ Inform. Sci. 99, 69–82 (1997). https://doi.org/10.1016/S0020-0255(96)00200-9
V. Nair and G. Hinton, ‘‘Rectified linear units improve restricted boltzmann machines,’’ in Proceedings of the 27th International Conference on Machine Learning ICML-10, June 21–24, 2010, Haifa, Israel (2010). https://api.semanticscholar.org/CorpusID:15539264
A. Pretorius, E. Barnard, and M. Davel, ‘‘ReLU and sigmoidal activation functions,’’ in Fundamentals of Artificial Intelligence Research (Springer, 2019). https://api.semanticscholar.org/CorpusID:211073632
C. Nwankpa, W. Ijomah, A. Gachagan, and S. Marshall, ‘‘Activation functions: Comparison of trends in practice and research for deep learning,’’ in Proceedings of the 2nd International Conference on Computational Sciences and Technologies, arXiv: 1811.03378 (2018). https://api.semanticscholar.org/CorpusID:53208763
A. Krizhevsky, I. Sutskever, and G. Hinton, ‘‘ImageNet classification with deep convolutional neural networks,’’ Commun. ACM 60, 84–90 (2012). https://api.semanticscholar.org/CorpusID:195908774
I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning (MIT Press, Boston, 2016).
P. Huber, ‘‘Robust estimation of a location parameter,’’ Ann. Math. Stat. 35, 492–518 (1964). https://api.semanticscholar.org/CorpusID:121252793
M. N. Amin and H. Meidani, ‘‘Physics-informed regularization of deep neural networks,’’ J. Comput. Inf. Sci. Eng. 20 (2018). https://api.semanticscholar.org/CorpusID:203047559
R. Rojas, ‘‘The backpropagation algorithm,’’ in Neural Networks: A Systematic Introduction (Springer, Heidelberg, 1996), pp. 149–182. https://doi.org/10.1007/978-3-642-61068-4_7
M. Claesen and B. De Moor, ‘‘Hyperparameter search in machine learning,’’ arXiv: 1502.02127 (2015). https://api.semanticscholar.org/CorpusID:17147092
L. O. Müller, S. M. Watanabe, E. F. Toro, R. A. Feijóo, and P. J. Blanco, ‘‘An anatomically detailed arterial-venous network model. Cerebral and coronary circulation,’’ Front. Physiol. 14, 1162391 (2023). https://doi.org/10.3389/fphys.2023.1162391
S. S. Simakov, T. M. Gamilov, F. Liang, D. G. Gognieva, D. G. M. K. Gappoeva, and P. Y. Kopylov, ‘‘Numerical evaluation of the effectiveness of coronary revascularization,’’ Russ. J. Numer. Anal. Math. Model. 36, 303–312 (2021). https://doi.org/10.1515/rnam-2021-0025
D. Gognieva, Yu. Mitina, T. Gamilov, R. Pryamonosov, Yu. Vasilevsky, S. Simakov, F. Liang, S. Ternovoy, N. Serova, E. Tebenkova, V. Sinitsyn, E. Pershina, S. Abugov, G. Mardanian, N. Zakarian, V. Kirakosuan, V. Betelin, D. Schekochikhin, A. Syrkin, and F. Kopylov, ‘‘Noninvasive assessment of the fractional flow reserve with the CT FFRc 1D method: Final results of a pilot study,’’ Gobal Heart 16 (1), 1 (2021). https://doi.org/10.5334/gh.837
D. G. Gognieva, E. S. Pershina, Yu. O. Mitina, T. M. Gamilov, R. A. Pryamonosov, N. A. Gogiberidze, A. N. Rozhkov, Yu. V. Vasilevsky, S. S. Simakov, F. Liang, V. E. Sinitsyn, V. B. Betelin, D. Yu. Schekochikhin, A. L. Syrkin, and F. Yu. Kopylov, ‘‘Non-invasive fractional flow reserve: A comparison of one-dimensional and three-dimensional mathematical modeling effectiveness,’’ Cardiovasc. Ther. Prevent. 19, 2303 (2020). https://doi.org/10.15829/1728-8800-2020-2303
Funding
This work has been supported by the Russian Science Foundation, grant number 21-71-30023.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
The authors of this work declare that they have no conflicts of interest.
Additional information
Publisher’s Note.
Pleiades Publishing remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
(Submitted by Yu. V. Vassilevski)
Rights and permissions
About this article
Cite this article
Isaev, A., Dobroserdova, T., Danilov, A. et al. Physically Informed Deep Learning Technique for Estimating Blood Flow Parameters in Arterial Bifurcations. Lobachevskii J Math 45, 239–250 (2024). https://doi.org/10.1134/S1995080224010219
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1995080224010219