Deep-learning cell imaging through Anderson localizing optical fiber

Zhao, Jian; Sun, Yangyang; Zhu, Hao; Zhu, Zheyuan; Antonio-López, Jose Enrique; Correa, Rodrigo Amezcua; Pang, Sean; Schülzgen, Axel

doi:10.1117/1.ap.1.6.066001

Cited by 23 publications

(20 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The learning-based approach has attracted lots of attention recently, mainly focusing on the DCNNs to tackle the inverse imaging problem of the fiber-optic system [18,19,27,[50][51][52]105]. DCNN is a specific class of machine learning techniques.…”

Section: Fundamentals Of Cnnmentioning

confidence: 99%

“…For clinical practices, a handheld FOIS can go deep into human organs or tissues with minimized penetration damages, which significantly benefits clinical diagnostics and surgical procedures [7][8][9]. Different types of optical fibers have been explored to develop the FOIS utilizing various mechanisms of image transport and recovery [1,4,[10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29]. Among FOIS solutions, multicore optical fibers (MCFs) and multimode optical fibers (MMFs) are two widely deployed fiber types.…”

Section: Introductionmentioning

confidence: 99%

“…While the novel waveguiding physics can relieve the device restrictions, it still requires a well-designed imaging reconstruction algorithm to make the utmost of the fiber. Due to their outstanding performance on image classifications, segmentations, and reconstructions, deep-learning (DL) methods have motivated researchers to deploy these algorithms in the fiber optical imaging area [18,19,27,[50][51][52][53][54]. DL-based research in optics and photonics is fast-growing.…”

Section: Introductionmentioning

confidence: 99%

“…On the other hand, DLbased solutions boosted by big data stand for a universal approach without the need for a handcrafting forward model [58][59][60]. They are able to directly "learn" the underlying physics of a complex waveguiding system merely relying on a set of training data without any prior knowledge [19]. Boosted by the newgeneration graphics processing units (GPUs), many DL-based tasks can be processed with a personal computer and reach milliseconds per frame imaging speed for a trained DL neural network.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Learning-Based Image Transport Through Disordered Optical Fibers With Transverse Anderson Localization

Zhao

Gausmann

et al. 2021

Front. Phys.

Self Cite

View full text Add to dashboard Cite

Fiber-optic imaging systems play a unique role in biomedical imaging and clinical practice due to their flexibilities of performing imaging deep into tissues and organs with minimized penetration damage. Their imaging performance is often limited by the waveguide mode properties of conventional optical fibers and the image reconstruction method, which restrains the enhancement of imaging quality, transport robustness, system size, and illumination compatibility. The emerging disordered Anderson localizing optical fibers circumvent these difficulties by their intriguing properties of the transverse Anderson localization of light, such as single-mode-like behavior, wavelength independence, and high mode density. To go beyond the performance limit of conventional system, there is a growing interest in integrating the disordered Anderson localizing optical fiber with deep learning algorithms. Novel imaging platforms based on this concept have been explored recently to make the best of Anderson localization fibers. Here, we review recent developments of Anderson localizing optical fibers and focus on the latest progress in deep-learning-based imaging applications using these fibers.

show abstract

Section: Fundamentals Of Cnnmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Learning-Based Image Transport Through Disordered Optical Fibers With Transverse Anderson Localization

Zhao

Gausmann

et al. 2021

Front. Phys.

Self Cite

View full text Add to dashboard Cite

show abstract

“…In recent years, deep learning technology has demonstrated its powerful capability in a wide range of complex data analysis, particularly in bio-imaging [21][22][23][24][25][26][27] . By learning intrinsic features of input images using conventional neural networks (CNN), a multi-layer mapping function can be built between measured information and input imaging data.…”

Section: Introductionmentioning

confidence: 99%

Single Cell Biological Microlasers Powered by Deep Learning

Qiao

Zhang

Jie

et al. 2021

Preprint

View full text Add to dashboard Cite

Cellular lasers are cutting-edge technologies for biomedical applications. Due to the enhanced interactions between light and cells in microcavities, cellular properties and subtle changes of cells can be significantly reflected by the laser emission characteristics. In particular, transverse laser modes from single-cell lasers which utilize Fabry–Pérot cavities are highly correlated to the spatial biophysical properties of cells. However, the high chaotic and complex variation of laser modes limits their practical applications for cell detections. Deep learning technique has demonstrated its powerful capability in solving complex imaging problems, which is expected to be applied for cell detections based on laser mode imaging. In this study, deep learning technique was applied to analyze laser modes generated from single-cell lasers, in which a correlation between laser modes and physical properties of cells was built. As a proof-of-concept, we demonstrated the predictions of cell sizes using deep learning based on laser mode imaging. In the first part, bioinspired cell models were fabricated to systematically study how cell sizes affect the characteristics of laser modes. By training a convolutional neuron network (CNN) model with laser mode images, predictions of cell model diameters with a sub-wavelength accuracy were achieved. In the second part, deep learning was employed to study laser modes generated from biological cells. By training a CNN model with laser mode images acquired from astrocyte cells, predictions of cell sizes with a sub-wavelength accuracy were also achieved. The results show the great potential of laser mode imaging integrated with deep learning for cell analysis and biophysical studies.

show abstract

Brain Cell Laser Powered by Deep‐Learning‐Enhanced Laser Modes

Qiao

Zhang

Ang

et al. 2021

Advanced Optical Materials

View full text Add to dashboard Cite

offers new ways of using laser light in biological and biomedical applications. [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] Thanks to the strong light-matter interactions provided by an optical cavity, subtle changes in biological gain media could be amplified, leading to distinctive output lasing characteristics, such as narrow linewidth, strong intensity, and high contrast. Recent advances in cellular lasers have attracted extensive interest in cell tracking, intracellular detection, and phenotyping by exploiting various types of cavities. [5][6][7][8][9] In particular, single-cell lasers that utilize Fabry-Pérot (FP) microcavities possess the advantages of whole-body interaction between the electromagnetic field and the gain medium, providing a sensitive approach to study the physical structures and interactions in cells. [4,5,20] One of the most peculiar features of FP single-cell lasers is the "lens effect" of cells introduced in FP cavities, in which the laser patterns are formed by the superposition of a series of transverse laser modes. [21] Due to the higher refractive index of cells compared to its surrounding environment, a cell performs as a lens for laser convergence in the cavity, forming a stable cavity structure. In such circumstances, transverse modes with various orders are eigen solutions of the cavity, which are confined within the cell area. Similar concepts of bio-lenses have also been demonstrated in optical imaging and manipulation previously. [22,23] As typical transverse modes, Hermite Gaussian (HG), Laguerre Gaussian (LG), and Ince Gaussian (IG) modes are mostly observed in single-cell lasers. [4,21] Most studies mainly focused on the spectral information of biological lasers, without being able to utilize the spatial information of transverse laser modes due to the huge complexity. In principle, the attributes of laser modes (e.g., mode types and orders) are highly correlated to cells' biophysical properties. For example, the morphology and the refractive index distribution of a cell illustrate the boundary conditions of laser oscillations in an FP cavity, which determines the electromagnetic field distribution of a transverse mode. Moreover, the internal structure-induced scattering losses and the fluorophore (gain) distribution in a cell strongly influence the output intensity of a laser mode. The output laser modes can be viewed as 3D integrated information of the single-cell confined in the cavity. As such, laser mode imaging of single cells possesses a great potential for studying cellular properties and changes in cells. However, the Single cellular lasers have recently attracted tremendous research due to their outstanding lasing characteristics for cell sensing and tracking. Thanks to enhanced light−cell interactions in Fabry-Pérot microcavities, transverse laser modes from cellular lasers are highly correlated to the spatial biophysical properties of cells. However, the huge complexity and randomness of laser modes set a critical challenge towards pract...

show abstract

Deep-learning cell imaging through Anderson localizing optical fiber

Cited by 23 publications

References 53 publications

Learning-Based Image Transport Through Disordered Optical Fibers With Transverse Anderson Localization

Learning-Based Image Transport Through Disordered Optical Fibers With Transverse Anderson Localization

Single Cell Biological Microlasers Powered by Deep Learning

Brain Cell Laser Powered by Deep‐Learning‐Enhanced Laser Modes

Contact Info

Product

Resources

About