×

A Helmholtz equation solver using unsupervised learning: application to transcranial ultrasound. (English) Zbl 07513817

Summary: Transcranial ultrasound therapy is increasingly used for the non-invasive treatment of brain disorders. However, conventional numerical wave solvers are currently too computationally expensive to be used online during treatments to predict the acoustic field passing through the skull (e.g., to account for subject-specific dose and targeting variations). As a step towards real-time predictions, in the current work, a fast iterative solver for the heterogeneous Helmholtz equation in 2D is developed using a fully-learned optimizer. The lightweight network architecture is based on a modified UNet that includes a learned hidden state. The network is trained using a physics-based loss function and a set of idealized sound speed distributions with fully unsupervised training (no knowledge of the true solution is required). The learned optimizer shows excellent performance on the test set, and is capable of generalization well outside the training examples, including to much larger computational domains, and more complex source and sound speed distributions, for example, those derived from x-ray computed tomography images of the skull.

MSC:

65Nxx Numerical methods for partial differential equations, boundary value problems
35Jxx Elliptic equations and elliptic systems
65Fxx Numerical linear algebra
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] McDannold, N.; Clement, G. T.; Black, P.; Jolesz, F.; Hynynen, K., Transcranial magnetic resonance imaging-guided focused ultrasound surgery of brain tumors: initial findings in 3 patients, Neurosurgery, 66, 2, 323-332 (2010)
[2] Legon, W.; Sato, T. F.; Opitz, A.; Mueller, J.; Barbour, A.; Williams, A.; Tyler, W. J., Transcranial focused ultrasound modulates the activity of primary somatosensory cortex in humans, Nat. Neurosci., 17, 2, 322-329 (2014)
[3] Abrahao, A.; Meng, Y.; Llinas, M.; Huang, Y.; Hamani, C.; Mainprize, T.; Aubert, I.; Heyn, C.; Black, S. E.; Hynynen, K.; Lipsman, N.; Zinman, L., First-in-human trial of blood-brain barrier opening in amyotrophic lateral sclerosis using MR-guided focused ultrasound, Nat. Commun., 10, 1, 1-9 (2019)
[4] Hynynen, K.; Jolesz, F. A., Demonstration of potential noninvasive ultrasound brain therapy through an intact skull, Ultrasound Med. Biol., 24, 2, 275-283 (1998)
[5] Chang, W. S.; Jung, H. H.; Zadicario, E.; Rachmilevitch, I.; Tlusty, T.; Vitek, S.; Chang, J. W., Factors associated with successful magnetic resonance-guided focused ultrasound treatment: efficiency of acoustic energy delivery through the skull, J. Neurosurg., 124, 2, 411-416 (2016)
[6] Albelda Gimeno, L.; Martin, E.; Wright, O.; Treeby, B. E., Experimental assessment of skull aberration and transmission loss at 270 kHz for focused ultrasound stimulation of the primary visual cortex, (2019 IEEE International Ultrasonics Symposium (IUS) (2019), IEEE)
[7] Sun, J.; Hynynen, K., Focusing of therapeutic ultrasound through a human skull: a numerical study, J. Acoust. Soc. Am., 104, 3, 1705-1715 (1998)
[8] Aubry, J.-F.; Tanter, M.; Pernot, M.; Thomas, J.-L.; Fink, M., Experimental demonstration of noninvasive transskull adaptive focusing based on prior computed tomography scans, J. Acoust. Soc. Am., 113, 1, 84-93 (2003)
[9] Lee, W.; Kim, H.-C.; Jung, Y.; Chung, Y. A.; Song, I.-U.; Lee, J.-H.; Yoo, S.-S., Transcranial focused ultrasound stimulation of human primary visual cortex, Sci. Rep., 6, 1, 1-12 (2016)
[10] Pinton, G.; Aubry, J.-F.; Fink, M.; Tanter, M., Effects of nonlinear ultrasound propagation on high intensity brain therapy, Med. Phys., 38, 3, 1207-1216 (2011)
[11] Pulkkinen, A.; Werner, B.; Martin, E.; Hynynen, K., Numerical simulations of clinical focused ultrasound functional neurosurgery, Phys. Med. Biol., 59, 7, 1679 (2014)
[12] Almquist, S.; Parker, D. L.; Christensen, D. A., Rapid full-wave phase aberration correction method for transcranial high-intensity focused ultrasound therapies, J. Ther. Ultrasound, 4, 1 (2016)
[13] Robertson, J. L.; Cox, B. T.; Jaros, J.; Treeby, B. E., Accurate simulation of transcranial ultrasound propagation for ultrasonic neuromodulation and stimulation, J. Acoust. Soc. Am., 141, 3, 1726-1738 (2017)
[14] McDannold, N.; White, P. J.; Cosgrove, R., Element-wise approach for simulating transcranial MRI-guided focused ultrasound thermal ablation, Phys. Rev. Res., 1, 3, Article 033205 pp. (2019)
[15] Clement, G. T.; Hynynen, K., A non-invasive method for focusing ultrasound through the human skull, Phys. Med. Biol., 47, 8, 1219 (2002)
[16] Kyriakou, A.; Neufeld, E.; Werner, B.; Székely, G.; Kuster, N., Full-wave acoustic and thermal modeling of transcranial ultrasound propagation and investigation of skull-induced aberration correction techniques: a feasibility study, Journal of Therapeutic Ultrasound, 3, 1, 11 (2015)
[17] Jin, K. H.; McCann, M. T.; Froustey, E.; Unser, M., Deep convolutional neural network for inverse problems in imaging, IEEE Trans. Image Process., 26, 9, 4509-4520 (2017) · Zbl 1409.94275
[18] Hammernik, K.; Schlemper, J.; Qin, C.; Duan, J.; Summers, R. M.; Rueckert, D., ∑-net: systematic evaluation of iterative deep neural networks for fast parallel MR image reconstruction (2019), arXiv preprint
[19] White, P. J.; Clement, G. T.; Hynynen, K., Longitudinal and shear mode ultrasound propagation in human skull bone, Ultrasound Med. Biol., 32, 7, 1085-1096 (2006)
[20] Robertson, J.; Martin, E.; Cox, B.; Treeby, B. E., Sensitivity of simulated transcranial ultrasound fields to acoustic medium property maps, Phys. Med. Biol., 62, 7, 2559 (2017)
[21] Rosnitskiy, P. B.; Yuldashev, P. V.; Sapozhnikov, O. A.; Gavrilov, L. R.; Khokhlova, V. A., Simulation of nonlinear trans-skull focusing and formation of shocks in brain using a fully populated ultrasound array with aberration correction, J. Acoust. Soc. Am., 146, 3, 1786-1798 (2019)
[22] Treeby, B. E.; Jaros, J.; Rendell, A. P.; Cox, B., Modeling nonlinear ultrasound propagation in heterogeneous media with power law absorption using a k-space pseudospectral method, J. Acoust. Soc. Am., 131, 6, 4324-4336 (2012)
[23] Wang, S.; Maarten, V.; Xia, J., Acoustic inverse scattering via Helmholtz operator factorization and optimization, J. Comput. Phys., 229, 22, 8445-8462 (2010) · Zbl 1201.65193
[24] Gumerov, N. A.; Duraiswami, R., A broadband fast multipole accelerated boundary element method for the three dimensional Helmholtz equation, J. Acoust. Soc. Am., 125, 1, 191-205 (2009)
[25] Bermúdez, A.; Hervella-Nieto, L.; Prieto, A.; Rodrı, R., An optimal perfectly matched layer with unbounded absorbing function for time-harmonic acoustic scattering problems, J. Comput. Phys., 223, 2, 469-488 (2007) · Zbl 1115.76041
[26] Raissi, M.; Perdikaris, P.; Karniadakis, G. E., Physics informed deep learning (part I): data-driven solutions of nonlinear partial differential equations (2017), arXiv preprint
[27] Weinan, E.; Yu, B., The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems, Commun. Math. Stat., 6, 1, 1-12 (2018) · Zbl 1392.35306
[28] Bar, L.; Sochen, N., Unsupervised deep learning algorithm for PDE-based forward and inverse problems (2019), arXiv preprint
[29] Zhu, Y.; Zabaras, N.; Koutsourelakis, P.-S.; Perdikaris, P., Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data, J. Comput. Phys., 394, 56-81 (2019) · Zbl 1452.68172
[30] Saad, Y.; Schultz, M. H., GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems, SIAM J. Sci. Stat. Comput., 7, 856-869 (July 1986) · Zbl 0599.65018
[31] Ernst, O. G.; Gander, M. J., Why it is difficult to solve Helmholtz problems with classical iterative methods, (Numerical Analysis of Multiscale Problems (2012), Springer), 325-363 · Zbl 1248.65128
[32] Erlangga, Y. A.; Oosterlee, C. W.; Vuik, C., A novel multigrid based preconditioner for heterogeneous Helmholtz problems, SIAM J. Sci. Comput., 27, 1471-1492 (Jan. 2006) · Zbl 1095.65109
[33] Oosterlee, C. W.; Vuik, C.; Mulder, W. A.; Plessix, R.-E., Shifted-Laplacian preconditioners for heterogeneous Helmholtz problems, (Koren, B.; Vuik, K., Advanced Computational Methods in Science and Engineering. Advanced Computational Methods in Science and Engineering, Lecture Notes in Computational Science and Engineering, vol. 71 (2009), Springer Berlin Heidelberg: Springer Berlin Heidelberg Berlin, Heidelberg), 21-46 · Zbl 1190.65183
[34] Engquist, B.; Ying, L., Sweeping preconditioner for the Helmholtz equation: hierarchical matrix representation, Commun. Pure Appl. Math., 64, 5, 697-735 (2011) · Zbl 1229.35037
[35] Elman, H. C.; Ernst, O. G.; O’leary, D. P., A multigrid method enhanced by Krylov subspace iteration for discrete Helmholtz equations, SIAM J. Sci. Comput., 23, 4, 1291-1315 (2001) · Zbl 1004.65134
[36] Chen, Y.; Dong, B.; Xu, J., Meta-mgnet: meta multigrid networks for solving parameterized partial differential equations (2020), arXiv preprint
[37] Rizzuti, G.; Siahkoohi, A.; Herrmann, F. J., Learned Iterative Solvers for the Helmholtz Equation, 81st EAGE Conference and Exhibition 2019, vol. 2019, 1-5 (2019), European Association of Geoscientists & Engineers
[38] Ronneberger, O.; Fischer, P.; Brox, T., U-Net: convolutional networks for biomedical image segmentation, (International Conference on Medical Image Computing and Computer-Assisted Intervention (2015), Springer), 234-241
[39] Andrychowicz, M.; Denil, M.; Gomez, S.; Hoffman, M. W.; Pfau, D.; Schaul, T.; Shillingford, B.; De Freitas, N., Learning to learn by gradient descent by gradient descent, (Advances in Neural Information Processing Systems (2016)), 3981-3989
[40] Putzky, P.; Welling, M., Recurrent inference machines for solving inverse problems (2017), arXiv preprint
[41] Zhu, B.; Liu, J. Z.; Cauley, S. F.; Rosen, B. R.; Rosen, M. S., Image reconstruction by domain-transform manifold learning, Nature, 555, 487-492 (Mar. 2018)
[42] Li, K.; Malik, J., Learning to optimize (June 2016)
[43] Hausknecht, M.; Stone, P., Deep recurrent Q-learning for partially observable MDPs (Jan. 2017)
[44] Lu, Y.; Zhong, A.; Li, Q.; Dong, B., Beyond finite layer neural networks: bridging deep architectures and numerical differential equations, (International Conference on Machine Learning, PMLR (2018)), 3276-3285
[45] Dupont, E.; Doucet, A.; Teh, Y. W., Augmented neural ODEs, (Advances in Neural Information Processing Systems (2019)), 3140-3150
[46] Adler, J.; Öktem, O., Solving ill-posed inverse problems using iterative deep neural networks, Inverse Probl., 33, Article 124007 pp. (nov 2017) · Zbl 1394.92070
[47] He, K.; Zhang, X.; Ren, S.; Sun, J., Delving deep into rectifiers: surpassing human-level performance on imagenet classification, (Proceedings of the IEEE International Conference on Computer Vision (2015)), 1026-1034
[48] He, J.; Xu, J., MgNet: a unified framework of multigrid and convolutional neural network, Sci. China Math., 62, 7, 1331-1354 (2019) · Zbl 1476.65026
[49] Fry, F. J.; Barger, J. E., Acoustical properties of the human skull, J. Acoust. Soc. Am., 63, 5, 1576-1590 (1978)
[50] Jaeger, H., Tutorial on training recurrent neural networks, covering BPPT, RTRL, EKF and the “echo state network” approach, vol. 5 (2002), German National Research Center for Information Technology, GMD Report 159
[51] Tallec, C.; Ollivier, Y., Unbiasing truncated backpropagation through time (May 2017)
[52] Kapturowski, S.; Ostrovski, G.; Quan, J.; Munos, R.; Dabney, W., Recurrent experience replay in distributed reinforcement learning, (International Conference on Learning Representations (2018))
[53] Falcon, W., PyTorch lightning (2019), vol. 3
[54] Treeby, B. E.; Cox, B. T., k-Wave: MATLAB toolbox for the simulation and reconstruction of photoacoustic wave fields, J. Biomed. Opt., 15, 2, Article 021314 pp. (2010)
[55] Chilamkurthy, S.; Ghosh, R.; Tanamala, S.; Biviji, M.; Campeau, N. G.; Venugopal, V. K.; Mahajan, V.; Rao, P.; Warier, P., Development and validation of deep learning algorithms for detection of critical findings in head CT scans (2018), arXiv preprint
[56] Martin, E.; Ling, Y. T.; Treeby, B. E., Simulating focused ultrasound transducers using discrete sources on regular Cartesian grids, IEEE Trans. Ultrason. Ferroelectr. Freq. Control, 63, 10, 1535-1542 (2016)
[57] Kawaguchi, K., A Simple Deep Equilibrium Model Converges to Global Optima with Weight Tying (2021)
[58] Sutton, R. S.; Barto, A. G., Reinforcement learning: an introduction, (Adaptive Computation and Machine Learning (1998), MIT Press: MIT Press Cambridge, Mass)
[59] Long, Z.; Lu, Y.; Ma, X.; Dong, B., PDE-Net: learning PDEs from data, (International Conference on Machine Learning (2018)), 3208-3216
[60] Boyd, J. P., Chebyshev and Fourier Spectral Methods (2001), Dover: Dover Mineola · Zbl 0994.65128
[61] Glorot, X.; Bengio, Y., Understanding the difficulty of training deep feedforward neural networks, (Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, JMLR Workshop and Conference Proceedings (2010)), 249-256
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.