Complex Valued U Net with Pupil Self Calibration for Fourier Ptychographic Reconstruction
Authors
Szűcs Csanád, Horváth ZsuzsannaFiles
Abstract
Pupil aberrations introduce coupled amplitude and phase errors into Fourier ptychographic reconstruction, limiting the reliability of learned estimators when acquisition conditions depart from their training distribution. This study proposes a complex-valued U-Net that integrates pupil self-calibration with a differentiable coherent imaging operator. The object branch processes real and imaginary features through coupled convolutions, while a constrained pupil branch estimates low order aberration coefficients from measurement residuals. Alternating updates, pupil support constraints, and explicit phase gauge fixing separate optical correction from object refinement. The evaluation addresses reconstruction accuracy, aberration tolerance, measurement efficiency, and computational cost within a unified comparison protocol. The proposed configuration reaches an amplitude peak signal to noise ratio of 34.8 dB, a structural similarity index of 0.961, and an aligned phase root mean square error of 0.118 rad. The fixed pupil complex network reaches 31.2 dB and 0.187 rad under the same comparison settings. Removing pupil calibration increases pupil phase error from 0.072 to 0.241 rad. Joint analysis of object error, coefficient trajectories, and held out illumination residuals characterizes the interaction between the learned specimen representation and acquisition specific optical correction. The formulation connects complex field estimation with a constrained calibration mechanism and defines the accuracy, consistency, and resource criteria needed to assess its operating range.
References
Eadie, F., Copeland, L., Di Caprio, G., McConnell, G., & Kallepalli, A. (2025). Fourier ptychography microscopy for digital pathology. Journal of Microscopy, 300(2), 260–285. https://doi.org/10.1111/jmi.70001
Shu, Y., Sun, J., Fan, Y., Jin, Y., Chen, Q., & Zuo, C. (2024). Diagonal illumination scheme for Fourier ptychographic microscopy: Resolution doubling and aliasing minimization. Journal of the Optical Society of America A, 41(11), C62–C71. https://doi.org/10.1364/josaa.532252
Zhang, S., Wang, A., Xu, J., Feng, T., Zhou, J., & Pan, A. (2024). FPM-WSI: Fourier ptychographic whole slide imaging via feature-domain backdiffraction. Optica, 11(5), 634–646. https://doi.org/10.1364/optica.517277
Liu, H., Xu, T., Chen, Y., Wang, Y., & Li, J. (2024). Transformed pupil-function misalignment calibration strategy for Fourier ptychographic microscopy. Optics Express, 32(7), 11429–11446. https://doi.org/10.1364/oe.515196
Ma, Q., Zhao, J., & Cui, G. (2024). LED-based temporal variant noise model for Fourier ptychographic microscopy. Optics Express, 32(8), 14620–14644. https://doi.org/10.1364/oe.518908
Zhang, S., & Cao, L. (2025). Whole-field, high-resolution Fourier ptychography with neural pupil engineering. Optica, 12(10), 1615–1624. https://doi.org/10.1364/optica.575065
Wu, K., Pan, A., & Gao, W. (2024). Fourier ptychographic reconstruction with denoising diffusion probabilistic models. Optics & Laser Technology, 176, Article 111016. https://doi.org/10.1016/j.optlastec.2024.111016
Costanzo, S., & Flores, A. M. (2025). From iterative methods to neural networks: Complex-valued approaches in medical image reconstruction. Electronics, 14(10), Article 1959. https://doi.org/10.3390/electronics14101959
Wengrowicz, O., Bronstein, A. M., & Cohen, O. (2024). Unsupervised physics-informed deep learning-based reconstruction for time-resolved imaging by multiplexed ptychography. Optics Express, 32(6), 8791–8803. https://doi.org/10.1364/oe.515445
Yin, T., Song, G., Lee, D., Oh, J., & Jang, M. (2025). Impact of system-model mismatch in Fourier ptychographic microscopy. Advanced Photonics Research, 6(12), Article e202500180. https://doi.org/10.1002/adpr.202500180
Lee, K. C., Chae, H., Xu, S., Lee, K., Horstmeyer, R., Lee, S. A., & Hong, B.-W. (2024). Anisotropic regularization for sparsely sampled and noise-robust Fourier ptychography. Optics Express, 32(14), 25343–25361. https://doi.org/10.1364/oe.529023
Kang, I., Zhang, Q., Yu, S. X., & Ji, N. (2024). Coordinate-based neural representations for computational adaptive optics in widefield microscopy. Nature Machine Intelligence, 6(6), 714–725. https://doi.org/10.1038/s42256-024-00853-3
Arcab, P., Rogalski, M., Trusiak, M., & Zdańkowski, P. (2025). Two-step automated method for robust Fourier ptychographic microscopy LED misalignment calibration. Optics and Lasers in Engineering, 195, Article 109340. https://doi.org/10.1016/j.optlaseng.2025.109340
Chen, Q., Cai, C.-T., He, X.-T., & Chen, R. (2025). Fourier ptychographic microscopy with a two-stage physics-enhanced neural network. Optics & Laser Technology, 181, Article 112016. https://doi.org/10.1016/j.optlastec.2024.112016
Dou, J., An, Q., Liu, X., Mai, Y., Zhong, L., Di, J., & Qin, Y. (2025). Enhanced phase recovery in in-line holography with self-supervised complex-valued neural networks. Optics and Lasers in Engineering, 184, Article 108685. https://doi.org/10.1016/j.optlaseng.2024.108685
Bardozzo, F., Fiore, P., Valentino, M., Bianco, V., Memmolo, P., Miccio, L., Brancato, V., Smaldone, G., Gambacorta, M., Salvatore, M., Ferraro, P., & Tagliaferri, R. (2024). Enhanced tissue slide imaging in the complex domain via cross-explainable GAN for Fourier ptychographic microscopy. Computers in Biology and Medicine, 179, Article 108861. https://doi.org/10.1016/j.compbiomed.2024.108861
Zhou, X., Tong, H., Ouyang, E., Zhao, L., & Fang, H. (2025). Fourier ptychographic microscopy with optical aberration correction and phase unwrapping based on semi-supervised learning. Applied Sciences, 15(1), Article 423. https://doi.org/10.3390/app15010423
Vong, A., Henke, S., Hoidn, O., Ruth, H., Deng, J., Hexemer, A., Shapiro, D., Mehta, A., Gleason, A., Hancock, L., & Schwarz, N. (2026). Towards generalizable deep ptychography neural networks. npj Computational Materials, 12, Article 368. https://doi.org/10.1038/s41524-026-02198-4
Liu, M., Wu, R., Luo, Z., Zhen, J., Zhang, H., Luo, J., Yan, L., & Wu, Y. (2024). Fast digital refocusing Fourier ptychographic microscopy method based on convolutional neural network. Optics Express, 32(1), 339–354. https://doi.org/10.1364/oe.512330
Allain, M., Aslan, S., Coene, W., Dirksen, S., Dong, J., Flamant, J., Iwen, M., Krahmer, F., van Leeuwen, T., Melnyk, O., Menzel, A., Mosk, A. P., Nikitin, V., Salanevich, P., Plonka, G., & Wellershoff, M. (2025). Phasebook: A survey of selected open problems in phase retrieval. Sampling Theory, Signal Processing, and Data Analysis, 23(2), Article 23. https://doi.org/10.1007/s43670-025-00117-0
Breger, A., Biguri, A., Landman, M. S., Selby, I., Amberg, N., Brunner, E., Gröhl, J., Hatamikia, S., Karner, C., Ning, L., Dittmer, S., Roberts, M., AIX-COVNET Collaboration, & Schönlieb, C.-B. (2025). A study of why we need to reassess full reference image quality assessment with medical images. Journal of Imaging Informatics in Medicine, 38(6), 3444–3469. https://doi.org/10.1007/s10278-025-01462-1
Pekkilä, J., Lappi, O., Robertsén, F., & Korpi-Lagg, M. J. (2025). Stencil computations on AMD and Nvidia graphics processors: Performance and tuning strategies. Concurrency and Computation: Practice and Experience, 37(12–14), Article e70129. https://doi.org/10.1002/cpe.70129
Zhou, Y., Xu, J., Feng, T., Ma, X., Mu, J., & Pan, A. (2026). A theory of digital refocusing in Fourier ptychographic microscopy. APL Photonics, 11(5), Article 056110. https://doi.org/10.1063/5.0324237
Kesgin, K., & Usta, E. (2026). Leakage-aware benchmarking and controlled ablation of attention models for student performance prediction across multiple tabular datasets. Discover Artificial Intelligence, 6, Article 1426. https://doi.org/10.1007/s44163-026-02398-3
Fatoretto, M. B., Demétrio, C. G. B., Brophy, C., & de Andrade Moral, R. (2026). Bootstrap confidence intervals for generalized linear mixed models in dose-response analysis. Journal of Statistical Computation and Simulation, 96(14), 3630–3653. https://doi.org/10.1080/00949655.2026.2677023
Ni, Y.-H., Fan, S.-Y., Zhang, S.-Y., & Sun, M.-J. (2024). Hyperuniform illumination subsampling method for Fourier ptychographic microscopy. Optics and Lasers in Engineering, 176, Article 108106. https://doi.org/10.1016/j.optlaseng.2024.108106
Wu, K., Pan, A., Sun, Z., Shi, Y., & Gao, W. (2024). Blind deep-learning based preprocessing method for Fourier ptychographic microscopy. Optics & Laser Technology, 169, Article 110140. https://doi.org/10.1016/j.optlastec.2023.110140
Espinosa-Momox, A., Norton, B., Serrano-García, D. I., & Porras-Aguilar, R. (2024). Dynamic quantitative phase microscopy: A single-shot approach using geometric phase interferometry. Communications Physics, 7, Article 256. https://doi.org/10.1038/s42005-024-01750-2