eFIN: Enhanced Fourier Imager Network for Generalizable Autofocusing and Pixel Super-Resolution in Holographic Imaging

被引:16
作者
Chen, Hanlong [1 ]
Huang, Luzhe [1 ]
Liu, Tairan [1 ]
Ozcan, Aydogan [1 ]
机构
[1] Univ Calif Los Angeles, Calif Nano Syst Inst CNSI, Elect & Comp Engn Dept, Bioengn Dept, Los Angeles, CA 90095 USA
关键词
Digital holography; deep learning; phase retrieval; autofocusing; pixel super-resolution; physics-informed learning; computational imaging; hologram reconstruction; QUANTITATIVE PHASE; RETRIEVAL; FIELD; MICROSCOPY; RECONSTRUCTION; ALGORITHMS;
D O I
10.1109/JSTQE.2023.3248684
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
The application of deep learning techniques has greatly enhanced holographic imaging capabilities, leading to improved phase recovery and image reconstruction. Here, we introduce a deep neural network termed enhanced Fourier Imager Network (eFIN) as a highly generalizable and robust framework for hologram reconstruction with pixel super-resolution and image autofocusing. Through holographic microscopy experiments involving lung, prostate and salivary gland tissue sections and Papanicolau (Pap) smears, we demonstrate that eFIN has a superior image reconstruction quality and exhibits external generalization to new types of samples never seen during the training phase. This network achieves a wide autofocusing axial range of Delta z similar to 350 mu m, with the capability to accurately predict the hologram axial distances by physics-informed learning. eFIN enables 3x pixel super-resolution imaging and increases the space-bandwidth product of the reconstructed images by 9-fold with almost no performance loss, which allows for significant time savings in holographic imaging and data processing steps. Our results showcase the advancements of eFIN in pushing the boundaries of holographic imaging for various applications in e.g., quantitative phase imaging and label-free microscopy.
引用
收藏
页数:10
相关论文
共 55 条
[1]   Phase retrieval from series of images obtained by defocus variation [J].
Allen, LJ ;
Oxley, MP .
OPTICS COMMUNICATIONS, 2001, 199 (1-4) :65-75
[2]   On the use of deep learning for computational imaging [J].
Barbastathis, George ;
Ozcan, Aydogan ;
Situ, Guohai .
OPTICA, 2019, 6 (08) :921-943
[3]   Deep learning-based digital in-line holographic microscopy for high resolution with extended field of view [J].
Byeon, Hyeokjun ;
Go, Taesik ;
Lee, Sang Joon .
OPTICS AND LASER TECHNOLOGY, 2019, 113 :77-86
[4]   Fourier Imager Network (FIN): A deep neural network for hologram reconstruction with superior external generalization [J].
Chen, Hanlong ;
Huang, Luzhe ;
Liu, Tairan ;
Ozcan, Aydogan .
LIGHT-SCIENCE & APPLICATIONS, 2022, 11 (01)
[5]   Quantitative phase-contrast imaging with compact digital holographic microscope employing Lloyd's mirror [J].
Chhaniwal, Vani ;
Singh, Amardeep S. G. ;
Leitgeb, Rainer A. ;
Javidi, Bahram ;
Anand, Arun .
OPTICS LETTERS, 2012, 37 (24) :5127-5129
[6]   Learning to synthesize: robust phase retrieval at low photon counts [J].
Deng, Mo ;
Li, Shuai ;
Goy, Alexandre ;
Kang, Iksung ;
Barbastathis, George .
LIGHT-SCIENCE & APPLICATIONS, 2020, 9 (01)
[7]   PHASE RETRIEVAL ALGORITHMS - A COMPARISON [J].
FIENUP, JR .
APPLIED OPTICS, 1982, 21 (15) :2758-2769
[8]   Low Photon Count Phase Retrieval Using Deep Learning [J].
Goy, Alexandre ;
Arthur, Kwabena ;
Li, Shuai ;
Barbastathis, George .
PHYSICAL REVIEW LETTERS, 2018, 121 (24)
[9]   Wide-field computational imaging of pathology slides using lens-free on-chip microscopy [J].
Greenbaum, Alon ;
Zhang, Yibo ;
Feizi, Alborz ;
Chung, Ping-Luen ;
Luo, Wei ;
Kandukuri, Shivani R. ;
Ozcan, Aydogan .
SCIENCE TRANSLATIONAL MEDICINE, 2014, 6 (267)
[10]   Wide-field computational color imaging using pixel super-resolved on-chip microscopy [J].
Greenbaum, Alon ;
Feizi, Alborz ;
Akbari, Najva ;
Ozcan, Aydogan .
OPTICS EXPRESS, 2013, 21 (10) :12469-12483