Search

Supervised deep-learning models have enabled super-resolution imaging in several microscopic imaging modalities, increasing the spatial lateral bandwidth of the original input images beyond the diffraction limit. Despite their success, their practical application poses several challenges in terms of the amount of training data and its quality, requiring the experimental acquisition of large, paired databases to generate an accurate generalized model whose performance remains invariant to unseen data. Cycle-consistent generative adversarial networks (cycleGANs) are unsupervised models for image-to-image translation tasks that are trained on unpaired datasets. This paper introduces a cycleGAN framework specifically designed to increase the lateral resolution limit in confocal microscopy by training a cycleGAN model using low- and high-resolution unpaired confocal images of human glioblastoma cells. Training and testing performances of the cycleGAN model have been assessed by measuring specific metrics such as background standard deviation, peak-to-noise ratio, and a customized frequency content measure. Our cycleGAN model has been evaluated in terms of image fidelity and resolution improvement using a paired dataset, showing superior performance than other reported methods. This work highlights the efficacy and promise of cycleGAN models in tackling super-resolution microscopic imaging without paired training, paving the path for turning home-built low-resolution microscopic systems into low-cost super-resolution instruments by means of unsupervised deep learning.

Among modern optical microscopy techniques, digital lensless holographic microscopy (DLHM) is one of the simplest label-free coherent imaging approaches. However, the hardware simplicity provided by the lensless configuration is often offset by the demanding computational postprocessing required to match the retrieved sample information to the user's expectations. A promising avenue to simplify this stage is the integration of artificial intelligence and machine learning (ML) solutions into the DLHM workflow. The biggest challenge to do so is the preparation of an extensive and high-quality experimental dataset of curated DLHM recordings to train ML models. In this work, a diverse, open-access dataset of DLHM recordings is presented as support for future research, contributing to the data needs of the applied research community. The database comprises 11,760 experimental DLHM holograms of bio and non-bio samples with diversity on the main recording parameters of the DLHM architecture. The database is divided into two datasets of 10 independent imaged samples. The first group, named multi-wavelength dataset, includes 8160 holograms and was recorded using laser diodes emitting at 654 nm, 510 nm, and 405 nm; the second group, named single-wavelength dataset, is composed of 3600 recordings and was acquired using a 633 nm He-Ne laser. All the experimental parameters related to the dataset acquisition, preparation, and calibration are described in this paper. The advantages of this large dataset are validated by re-training an existing autofocusing model for DLHM and as the training set for a simpler architecture that achieves comparable performance, proving its feasibility for improving existing ML-based models and the development of new ones.

Digital holographic microscopy (DHM) enables the three-dimensional (3D) reconstruction of quantitative phase distributions from a defocused hologram. Traditional computational algorithms follow a sequential approach in which one first reconstructs the complex amplitude distribution and later applies focusing algorithms to provide an in-focus phase map. In this work, we have developed a synergistic computational framework to compensate for the linear tilt introduced in off-axis DHM systems and autofocus the defocused holograms by minimizing a cost function, providing in-focus reconstructed phase images without phase distortions. The proposed computational tool has been validated in defocused holograms of human red blood cells and three-dimensional images of dynamic sperm cells.

A holographic point source (HPS) developed for digital lensless holographic microscopy (HPS-DLHM) is presented. The HPS is an off-axis phase transmission hologram of an experimental micrometer pinhole recorded on a photopolymer holographic film. An amplitude division interferometer, adjusted to operate at maximum diffraction efficiency, has been employed to record the hologram. The results of HPS-DLHM have been contrasted with the results obtained via conventional DLHM, and the two techniques were found to give similar measurements. Compared with conventional pinhole-based DLHM illumination, our cost-effective proposal provides increased mechanical stability, the possibility of wider spherical illumination cones, and shorter reconstruction distances. These superior features pave the way to applying this quantitative phase imaging (QPI) technique in biomedical and telemedicine applications. The imaging capabilities of our HPS-DLHM proposal have been tested by using an intricate sample of a honeybee leg, a low-absorption sample of epithelial cheek cells, a 1951 USAF test target, and smeared human erythrocytes.

A digital lensless holographic microscope (DLHM) sensitive to the linear diattenuation produced by biological samples is reported. The insertion of a linear polarization-states generator and a linear polarization-states analyzer in a typical DLHM setup allows the proper linear diattenuation imaging of microscopic samples. The proposal has been validated for simulated and experimental biological samples containing calcium oxalate crystals extracted from agave leaves and potato starch grains. The performance of the proposed method is similar to that of a traditional polarimetric microscope to obtain linear diattenuation images of microscopic samples but with the advantages of DLHM, such as numerical refocusing, cost effectiveness, and the possibility of field-portable implementation.

The implementation of a digital holographic interferometry setup for high-speed micro-deformation measurement is presented. This proposal uses a dual-wavelength recording strategy to reconstruct micro-deformations up to 4.85 µm with no phase wrapping. The numerical processing required to recover the phase maps containing the information of micro-deformations is carried out in a general-purpose computing on graphics processing unit environment to boost its performance. The method completely processes recorded holograms of 1024×1024 p i x e l s in 48 ms, i.e., 21 frames per second (FPS) for a single-wavelength acquisition and 96 ms or 11 FPS for dual-wavelength recordings. The method is experimentally evaluated measuring deformations ranging from 0.033 µm to 4.85 µm with no need for phase unwrapping algorithms for an 8 cm diameter aluminum plate in a 110 c m 2 field of view.

In this work, a numerical method to enhance the contrast of intensity hologram reconstructions of digital lensless holographic microscopy (DLHM) is presented. The method manipulates the in-line hologram and reference images through mathematical operations between them; additionally, a sharpening operation, functionalized in terms of the parameters of the recording setup, is applied to the said images. The preprocessing of the recorded images produces a modified in-line hologram and a reference wave image from which an intensity reconstruction with a 25% improvement of its contrast, with respect to the conventional reconstruction procedure, is achieved. The method is illustrated with intensity reconstructions of a hologram of a monolayer of polystyrene spheres 1.09 µm in diameter. Finally, the preprocessing method is validated with a modeled hologram, successfully applied to holograms of the section of the head a Drosophila melanogaster fly and its results are contrasted with those obtained via bright-field microscopy.

The description and validation of an ImageJ open-source plugin to numerically simulate and reconstruct digital lensless holographic microscopy (DLHM) holograms are presented. Two modules compose the presented plugin: the simulation module implements a discrete version of the Rayleigh-Somerfield diffraction formula, which allows the user to directly build a simulated hologram from a known phase and/or amplitude object by just introducing the geometry parameters of the simulated setup; the plugin's reconstruction module implements a discrete version of the Kirchhoff-Helmholtz diffraction integral, thus allowing the user to reconstruct DLHM holograms by introducing the parameters of the acquisition setup and the desired reconstruction distance. The plugin offers the two said modules within the robust environment provided by a complete set of built-in tools for image processing available in ImageJ. While the simulation module has been validated through the evaluation of the forecasted lateral resolution of a DLHM setup in terms of the numerical aperture, the reconstruction module is tested by means of reconstructing experimental DLHM holograms of biological samples.

The implementation of a numerical dark field illumination to obtain reconstructions with enhanced contrast in digital lensless holographic (DLHM) microscopy is presented. In this proposal, an opaque circular stop is numerically inserted in the light path between the sample and hologram plane. The contrast of the reconstructed holograms is increased at least 46% in comparison with that obtained by the regular reconstruction methods. This proposal has been validated using experimental holograms of a paramecium in water and a section of the head of a Drosophila melanogaster fly recorded in DLHM.

In this Letter, the use of two-coupled Mach-Zehnder interferometers for four π/2-phase shifting interferometry is introduced. A multi-camera arrangement using no more than beam splitters and mirrors is utilized to obtain in a single shot the needed phase-shifted interferograms in the different output channels of the setup. The simplicity of the setup makes it ideal for high-speed interferometry applications. This proposal is validated in digital holographic microscopy to visualize a biological sample of epidermal onion cells.

The cooperative execution of two metrics to automatically determine the best focal plane in digital lensless holographic microscopy (DLHM) is presented. This proposal is comprised of two stages: first, a quick coarse search over the whole reconstruction range by using Dubois's metric allows the finding of a range in which the best focal plane can be found. In a second stage, the modified enclosed energy (MEE) metric is used within the found range in the former stage to finely determine the best focal plane. While this cooperative implementation keeps the proven effectiveness of the MEE in DLHM, it reduces by at least 11 times the total computational complexity of the auto-focusing method with respect to the MEE method only. This proposal has been validated experimentally with DLHM holograms of a paramecium specimen, polystyrene beads, and the section of the head of a Drosophila melanogaster fly.

An automatic method that fully compensates the quantitative phase measurements in off-axis digital holographic microscopy (DHM) is presented. The two main perturbations of the quantitative phase measurements in off-axis DHM are automatically removed. While the curvature phase flaw introduced by the microscope objective is avoided by the use of an optimized telecentric imaging system for the recording of the holograms, the remaining phase perturbation due to the tilt of the reference wave is removed by the automatic computation of a digital compensating reference wave. The method has been tested on both nonbiological and biological samples with and improving on the quality of the recovered phase maps.

The use of an electronically tunable lens (ETL) to produce controlled phase shifts in interferometric arrangements is shown. The performance of the ETL as a phase-shifting device is experimentally validated in phase-shifting digital holographic microscopy. Quantitative phase maps of a section of the thorax of a Drosophila melanogaster fly and of human red blood cells have been obtained using our proposal. The experimental results validate the possibility of using the ETL as a reliable phase-shifter device.

A comparative analysis of the performance of the modified enclosed energy (MEE) method for self-focusing holograms recorded with digital lensless holographic microscopy is presented. Notwithstanding the MEE analysis previously published, no extended analysis of its performance has been reported. We have tested the MEE in terms of the minimum axial distance allowed between the set of reconstructed holograms to search for the focal plane and the elapsed time to obtain the focused image. These parameters have been compared with those for some of the already reported methods in the literature. The MEE achieves better results in terms of self-focusing quality but at a higher computational cost. Despite its longer processing time, the method remains within a time frame to be technologically attractive. Modeled and experimental holograms have been utilized in this work to perform the comparative study.