Nonlinear optimization-driven deep learning framework for medical image reconstruction via partial differential equations
Abstract views: 40
/ 
PDF downloads: 7
Keywords:
Medical image reconstruction, Nonlinear optimization, Deep learning, Partial differential equations (PDEs)Abstract
High quality medical imaging is essential to accurate clinical decision-making, but reconstruction of sparse or noisy images especially under CT and MRI is still a major challenge, with traditional reconstruction algorithms vulnerable to artifacts and noise, and unwanted inference typically lacking interpretability. We introduce a new modality of addressing the problem of reconstruction with non-linear optimization, partial differential equation (PDE) constraints and deep neural networks, where the priors on physical properties should be presented as the network loss function and the architecture of the network so as to build more robust and accurate reconstruction. Having a clear formulation of a nonlinear optimization problem and by using the principles of variational approaches, we are
also able to integrate a hardware friendly circuit into our solution that could be used to acquire data in real time. Experiments on benchmark CT and MRI, indicate an increase in peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), more effective noise suppression, and convergence times than state-of-the art baselines. The need of nonlinear optimization and employment of PDEs regularization in work on edges preservation and reduction of artifacts can also be seen regarding ablation results. Taken together, it is the first piece connecting the fields of model-based regularization and modern deep learning, thereby providing a clinical pipeline toward interpretable, high-fidelity, and deployable medical image reconstruction.
References
Maier, A., Syben, C., Lasser, T., & Riess, C. (2019). A gentle introduction to deep learning in medical image processing. Zeitschrift für Medizinische Physik, 29(2), 86–101.
Kak, A. C., & Slaney, M. (1988). Principles of computerized tomographic imaging. IEEE Press.
Lustig, M., Donoho, D., & Pauly, J. M. (2007). Sparse MRI: The application of compressed sensing for rapid MR imaging. Magnetic Resonance in Medicine, 58(6), 1182–1195.
Ronneberger, O., Fischer, P., & Brox, T. (2015). U-Net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 (pp. 234–241). Springer.
Ledig, C., Theis, L., Huszár, F., et al. (2017). Photo-Realistic single image super-resolution using a generative adversarial network. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR).
Rudin, L. I., Osher, S., & Fatemi, E. (1992). Nonlinear total variation based noise removal algorithms. Physica D: Nonlinear Phenomena, 60(1–4), 259–268.
Weickert, J. (1998). Anisotropic diffusion in image processing. Teubner.
Gilton, D., Ongie, G., & Willett, R. (2021). Neural proximal gradient descent for compressive imaging. IEEE Transactions on Computational Imaging, 7, 1123–1138.
Chambolle, A., & Pock, T. (2016). An introduction to continuous optimization for imaging. Acta Numerica, 25, 161–319.
Monga, V., Li, Y., & Eldar, Y. C. (2021). Algorithm unrolling: Interpretable, efficient deep learning for signal and image processing. IEEE Signal Processing Magazine, 38(2), 18–44.
Liu, H., Zhang, T., & Yang, X. (2023). Robust dynamic MRI reconstruction with PDE-constrained deep networks. Medical Image Analysis, 87, 102766.
Zhang, Y., Wang, Z., & Luo, S. (2023). GAN-based compressive sensing for low-dose CT. IEEE Transactions on Medical Imaging, 42(3), 591–602.
Park, J., Kim, K., & Lee, S. (2024). Residual-unrolled networks with PDE-inspired regularization for MRI. IEEE Access, 12, 17322–17334.
Chen, B., Fan, J., & Zhou, Y. (2024). Deep bilevel optimization for medical imaging: A variational perspective. Neurocomputing, 553, 125556.
Wang, Q., Xu, F., & Zhou, H. (2025). Transformer-based medical image reconstruction with learned PDE flows. Pattern Recognition, 150, 110230.
Barhoumi, E. M., Charabi, Y., & Farhani, S. (2024). Detailed guide to machine learning techniques in signal processing. Progress in Electronics and Communication Engineering, 2(1), 39–47. https://doi.org/10.31838/PECE/02.01.04
Ramchurn, R. (2025). Advancing autonomous vehicle technology: Embedded systems prototyping and validation. SCCTS Journal of Embedded Systems Design and Applications, 2(2), 56–64.
Frincke, G., & Wang, X. (2025). Hardware/software co-design advances for optimizing resource allocation in reconfigurable systems. SCCTS Transactions on Reconfigurable Computing, 2(2), 15–24. https://doi.org/10.31838/RCC/02.02.03
Rucker, P., Menick, J., & Brock, A. (2025). Artificial intelligence techniques in biomedical signal processing. Innovative Reviews in Engineering and Science, 3(1), 32–40. https://doi.org/10.31838/INES/03.01.05
Rasanjani, C., Madugalla, A. K., & Perera, M. (2023). Fundamental Digital Module Realization Using RTL Design for Quantum Mechanics. Journal of VLSI Circuits and Systems, 5(2), 1–7. https://doi.org/10.31838/jvcs/05.02.01
