Quantitative Phase Imaging XI

This talk details a new microscopic imaging approach for extremely high-throughput live cell video recording. Due to their limited field-of-view, today’s microscopes are extremely inefficient at live-cell recording across large areas (e.g., entire multi-well plates). This is particularly true for 3D recording of in vitro cultures such as organoids. This work presents a multi-camera array microscope (MCAM) that contains dozens of tightly packed imaging systems to simultaneously record video of live-cell dynamics. Associated phase contrast imaging methods that are optimized to operate within a microscope array format produce high-quality label free measurements for automated software analysis.

The Holographic Incoherent-light-source Quantitative Phase Imaging (hiQPI) technique provides optical sectioning capability similar to confocal microscopy due to the coherence-gating effect. Utilizing this capability for 3D imaging, hiQPI becomes a full-fledged alternative to Holographic Tomography (HT). We used the 1st Born approximation of scattering theory to compare the spatial-frequency transfer properties of the two 3D imaging techniques for weakly scattering specimens. We derived the 3D coherent transfer functions, which turned out to be identical for HT and hiQPI approaches. The reconstruction of the phantom’s 3D refractive index distribution from simulated hiQPI and HT data confirmed this theoretical prediction, while the reconstruction performed on experimental red-blood-cell data verified the feasibility of the hiQPI approach.

Computational microscopy integrates optical hardware and reconstruction algorithms to realize extended functionalities, higher resolution, etc., over conventional microscopes. However, sequential measurements are typically required to obtain high-dimensional data due to the limited information throughput of commercial image sensors, resulting in an inherent tradeoff between temporal resolution and imaging fidelity. Here, we propose a general computational framework, termed spatiotemporally regularized inversion (STRIVER), that exploits the spatiotemporal priors of dynamic scenes to jointly eliminate reconstruction artifacts and motion blurs. We experimentally demonstrate STRIVER on near-field ptychography, where holographic imaging of freely-moving organisms is performed at a framerate-limited speed of over 100 Hz over a large field of view. The framework can also be potentially extended to other imaging modalities, pushing the temporal resolution toward higher limits.

Lensless digital holographic microscopy (LDHM), as one of key computational microscopy techniques, performs high throughput in silico imaging. Numerical propagation of digitally recorded in-line Gabor holograms allows for accessing both amplitude (absorption) and phase (refraction) contrast, in a manner free from microscope objective limitations regarding, e.g., depth of field and field of view. The in-line coherent holographic framework induces inherent twin image errors and various coherent artifacts, however. The signal-to-noise ratio of reconstructed holograms additionally deteriorates due to low photon budget environment, favorable in terms of time-lapse photostimulation-free bioimaging of live cells. In this contribution, we discuss several techniques for minimization of LDHM reconstruction errors, with the emphasis on simultaneous validation of phase measurement fidelity via calibration target testing. We also present bio-applications of enhanced LDHM in dynamic (e.g., motile sperm cells and migrating neural cells) and static (brain tissue slices) scenarios for bioimaging of cell and tissue samples.

Phase imaging is a technique allowing any biological object (or not) to be observed without sample alteration. I will show the benefit of using phase imaging by quadrilateral shift interferometry (QLSI) for the characterization of biological objects and discuss how to apply it for nano-object in a super-resolution regime. I will discuss how and why using super-resolved label-free methods.

Label-free microscopy exploits light scattering to obtain a three-dimensional image of biological tissues. However, light propagation is affected by aberrations and multiple scattering, which drastically degrade the image quality and limit the penetration depth. Multi-conjugate adaptive optics and time-gated matrix approaches have been developed to compensate for aberrations but the associated frame rate is extremely limited for 3D imaging. Here, we develop a multi-spectral matrix approach to solve these fundamental problems. Based on a sparse illumination scheme and an interferometric measurement of the reflected wave-field at multiple wavelengths, the focusing process can be optimized in post-processing for any voxel by addressing independently each frequency component of the reflection matrix. A proof-of-concept experiment demonstrates the three-dimensional image of an opaque human cornea over a 0.1 mm^3-field-of-view at a 290 nm-resolution and a 1 Hz-frame rate. This work paves the way towards a fully-digital microscope allowing real-time, in-vivo, quantitative and deep inspection of tissues.

Optical aberrations can severely impair the clarity of microscopic images, particularly in cell biology and histopathology. Traditional methods like wavefront shaping and guide star techniques often struggle with biological tissues. We introduce a novel computational adaptive optics method for thick optical samples, utilizing tilt-tilt correlation from the optical memory effect to detect phase discrepancies caused by slight tilts in incident waves. Our technique significantly improves the imaging quality of thick tissues, even under challenging aberration conditions. Additionally, it is highly resistant to sample movement, enhancing imaging accuracy in biomedical applications.

Conventional phase microscopy techniques require transmission illumination. We present two microscopy designs that work in a back-illumination configuration, producing high-speed, high-resolution contrast of phase objects in unlabeled thick specimens. These techniques encode lateral or axial phase gradients as intensity differences at the image sensor plane, enabling high-contrast imaging of transparent features. Results from in-vivo blood cell and ex-vivo pathology specimens will be presented.

Oblique back-illumination microscopy provides phase contrast in thick samples by relying on multiple scattering inside the sample to convert incoherent epi-illumination into oblique trans-illumination. Here, we combine this geometry with coherent illumination and off-axis holography to enable post hoc digital refocusing. Incoherent summation of refocused holograms with independent speckle realizations from the multiply scattered coherent illumination mimics a partially coherent image, revealing phase gradients. Through quantitative reconstruction of phase objects in each focal plane, we envision to compensate for phase aberrations of superficial sample layers, potentially extending the imaging depth compared to conventional incoherent methods which rely on physical focusing.

Quantitative phase imaging (QPI) is a powerful method for examining label-free biomedical samples. Metasurfaces, which use subwavelength planar structures to precisely manipulate the amplitude, phase, and polarization of light, open up possibilities for novel optical functionalities. By replacing traditional bulky components with metasurfaces, QPI techniques can become simpler and more reliable, thereby enhancing system portability and expanding their use in in vivo imaging. Here, we introduce a single-shot QPI approach by incorporating all-dielectric geometric phase metasurfaces into a conventional microscope. The metasurfaces create two laterally displaced replicas with orthogonal circular polarization states of the input object beam. The interference of the replicas produces a retardance image with bias retardation controlled by an analyzer. The complex amplitude of the object is reconstructed from four retardance images captured simultaneously by a polarization camera. The metasurfaces can be positioned near any conjugate plane of the object, offering a flexible and robust setup for QPI, evincing its broad applicability in live-cell imaging.

Digital holographic microscopy (DHM) has emerged as a powerful quantitative phase imaging technique offering label-free, non-invasive visualization of cell structures and dynamics. Specifically, DHM provides a quantitative phase signal (QPS) that is highly sensitive, particularly to dry mass, which has led to the development of attractive applications in cell biology. QPS contains, in an intricate way, a large amount of information about the cell content and morphology. Its interferometric detection gives it a high degree of sensitivity, but at the same time it is contaminated with a coherent noise that makes a precise analysis of cell information it contains difficult. I’ll present a series of technical developments that have enabled us to obtain a quasi-coherent noise-free QPS from which we can extract a set of cellular parameters. This paves the way for high-content, label-free screening to identify disease-specific cell phenotypes. Some applications related to neuropsychiatric diseases will be presented. Finally, it will be shown how AI can take advantage of these technical developments to enable label-free cell phenotyping without the need for cumbersome instrumentation.

Non-invasive imaging with high resolution deep within biological materials without the use of harmful ionizing radiation is of great interest in the field of medical imaging. Nonlinear optical microscopy using second harmonic generation (SHG), and third harmonic generation (THG) microscopy provide a powerful contrast mechanism for probing biological systems. Normally, such imaging methods use laser-scanning microscopy to map the scattered SHG or THG light intensity. We introduce the use of quantitative nonlinear optical phase microscopy that is enabled by computational adaptive optics nonlinear holographic microscopy.

Microscopic specimens often exhibit color due to absorption within the visible spectrum. However, accurately measuring the volumetric structure of absorptive objects using conventional brightfield microscopy is challenging due to out-of-focus light. We introduce a method for three-dimensional (3D) imaging of absorption coefficients at submicrometer resolution in RGB wavelengths by deconvolving intensity measurements under controlled partially coherent illuminations. Absorption reconstruction accuracy was enhanced through deconvolution with regularization and non-negativity constraints, logarithmic intensity transformation and aberration correction using multiple illuminations. Validated through simulations and experiments, our method significantly improves volumetric imaging of absorptive samples, including printer toner particles, live melanocytes, in vivo plant petal cells, and stained tissues. This technique opens new possibilities for imaging 3D histology and live cell studies, capturing chemical insights.

In this contribution, we will explain how cross-grating wavefront microscopy can be upgraded to deliver images with greater informational content by using more sophisticated camera sensors, such as color cameras, polarized cameras, and high-definition cameras. These three modalities have enabled us to achieve single-shot fluorescence-phase correlation imaging in biology, electromagnetic field imaging in nanophotonics, and diffraction-limited Mpx phase images of live cells in culture.

Sickle cell disease (SCD) is a widespread genetic disorder where afflicted patients experience painful vaso-occlusive crisis events and significantly increased morbidity due to the presence of malformed hemoglobin in the patient. Sickle red blood cells (sRBCs) containing this hemoglobin exhibit unique morphological, biophysical, and mechanical properties that could relate to the disease’s clinical outcomes, but specific connections between these micro-scale cellular changes and macro-scale events are still elusive. To bridge this gap, we have developed interferometric phase and amplitude microscopy (iPAM), a high-throughput diffraction phase microscope coupled with efficient linear computational models, that extracts single-cell morphological and biophysical information from snapshot sRBC images. Across 67 patients, we show iPAM’s unique recovery of morphological and morphometric sRBC metrics at single-cell resolution matches clinically recovered global cellular volume and hemoglobin information and correlates well in predicting the patient’s hemoglobin chemical composition just from the sRBC morphological information.

I will discuss machine learning methods for imaging 3D objects from 2D projections. I will highlight MaxwellNet, a method that uses directly the satisfiabilty of Maxwell's equations to train the system.

We introduce a novel paradigm for full-field reflective geometry measurements. Contrary to full-field optical coherence tomography, which requires careful matching of the reference beam to the sample beam, our approach employs high temporal coherence and low spatial coherence illumination. This allows the reference plane to be shifted to the camera, eliminating the need for precise alignment. We further add 3D vibration stabilization, automatic calibration of the illumination incoherence and aberration correction to achieve in-vivo high-resolution imaging of the mouse brain.

We present deep learning-designed all-optical processors that can perform multiplane quantitative phase imaging (QPI). By leveraging diffractive processing and wavelength multiplexing, our approach allows the direct capture of quantitative phase distributions of input objects located at different axial planes using an intensity-only image sensor, eliminating the need for digital phase recovery algorithms. A proof-of-concept experiment validated the approach, showcasing successful quantitative phase imaging of distinct phase objects at different axial positions in the terahertz spectrum. By providing a faster, more efficient method for all-optical QPI, this technology can be applied to various applications in biomedical imaging/sensing, material science and environmental monitoring.

We describe an innovative phase retrieval method based on a deep neural network for single-shot quantitative phase gradient microscopy. Our method exploits an untrained neural network (UNN), inspired by the deep image prior (DIP), presenting promise in integrating the underlying physical prior with a DNN. We demonstrate our phase retrieval method and its fidelity by comparing the result with other methods, such as Tikhonov regularization. The enhanced imaging performance and phase accuracy are demonstrated with phase images of a calibration target and unstained cellular specimens, MTF-7 cells and human adipose-derived stem cells.

We propose a label-free deep-imaging method, known as scattering-enabled epi quantitative phase imaging (SEEQPI), for studying 3D organoids and in vivo mouse brains. SEEQPI is a custom-built laser-scanning, common-path, phase-shifting interferometor using long-wavelength illumination. SEEQPI has superior stability and sensitivity in phase measurement and reduces photodamage or heating to living systems. With the proposed method, we will demonstrate the working principle on standard samples, and the phase imaging of cancer spheroids and in vivo mouse brains, compared to colocolized multiphoton imaging.

In this talk I willl introduce AI-driven approaches to spatial multimodal 3D and 4D molecular tumor modeling, which integrates Holotomography, spatial transcriptomics, proteomics, and in situ molecular imaging. This innovative methodology allows for the comprehensive characterization of the tumor microenvironment (TME) at single-cell and subcellular levels. By leveraging AI and machine learning algorithms, we analyze petabyte-scale molecular data to identify novel biomarkers and therapeutic targets, enhance patient stratification, and develop new therapeutic modalities including CAR-T cell, ADC, and immunotherapy and combinatory therapies. These advancements have resulted in significant progress in understanding the interactions of cellular and non-cellular components within the TME and overcoming its barriers. This presentation will discuss the methodologies employed, key findings, and the potential implications for cancer treatment.

Quantitative oblique back-illumination microscopy (qOBM) is a label-free imaging technique that enables tomographic phase imaging of thick scattering samples with epi-illumination. Here, we propose the use of qOBM to study three patient-derived GBM cells exposed to radiation, chemotherapy, and immunotherapy. We compare the quantitative images, fast dynamics, and continuous imaging over the course of 72 hours of treatment. We can estimate spheroid viability non-invasively, continuously during culture and visualize the targeted attack of CAR-T cells, T-cells transversing through the spheroid, cellular mitosis, cellular apoptosis, and other cellular responses of the GBM cells to the stress of treatment to ultimately provide real-time insight into treatment efficacy.

Using a label-free and high-resolution imaging system, holotomography (HT), we observed early differentiation of human pluripotent stem cells (hPSCs) and generated a morphology-based pluripotency level prediction model by integrating artificial intelligence (AI). We applied non-invasive pluripotency prediction model for quality control of reprogrammed patient-derived induced pluripotent stem cells (iPSCs) selection, which can be further used in stem cell-based research and therapeutics. Moreover, we identified previously undetectable morphological properties depending on variable pluripotency and newly suggested standardized colonial and intracellular structure of hPSCs.

Clear cell renal cell carcinoma (ccRCC) is classified by the four-tiered World Health Organization/International Society of Urological Pathology (WHO/ISUP) grading system based on nucleolar prominence (grades 1-3) and nuclear anaplasia (grade 4) after Hematoxylin and Eosin staining. However, manual evaluation often leads to poor inter- and intra-observer reproducibility. We applied holotomography, a label-free imaging technique for measuring the refractive index (RI) of biological specimens, to quantify the RI differences in segmented nuclei by grade. The RI distribution within all nuclei and RI standard deviations in local patches significantly differed by grade (Student’s t-test, p<0.001). Grades 1 through 3 exhibited a decreasing trend, whereas grade 4 deviated from this pattern, reflecting the WHO/ISUP grading criteria. Morphology, intensity, and texture-related RI features also varied significantly among the grades (p<0.001). This quantitative approach to ccRCC grading can reduce reliance on subjective evaluation, thereby improving diagnostic accuracy.

Currently, holographic imaging flow cytometry (HIFC) is one of the most powerful label-free and high throughput imaging modality for single-cell analysis. The combination of HIFC with artificial intelligence can unlock unprecedent challenges, thus opening the way to advanced analysis of biological specimens. Here, several studies will be discusses with the focus of improving target cells identification and classification by exploiting a priori clinical information about the samples. In particular, applications regarding the differential diagnosis of anemia, liquid biopsy, cell phenotyping and drug resistance will be described.

Cilia are hair-like cellular projections from lung epithelia that facilitate mucociliary clearance. Cilia play a critical role in various pulmonary conditions, but their function cannot be studied in-vivo noninvasively. Airway organoids are three-dimensional cellular cultures that mimic the development of the human lung and provide the ability to study developing cilia. We propose quantitative oblique back-illumination microscopy, a label-free quantitative phase imaging modality that provides subcellular resolution of thick-scattering samples, as a tool for studying the growth of cilia throughout airway organoid development. Here we compare ciliary function in developing healthy and diseased airway organoids, such as cystic fibrosis.

Measurements of spatially-resolved intracellular water content are notoriously difficult, but highly desirable for studies spanning biology and biophysics. In this work, we employ the emerging technique of refractive index tomography (alternatively known as holotomography) to compute the three-dimensional, spatially-resolved distribution of water within living cells. We treat each voxel in a refractive index tomogram as a two-component mixture of dry and aqueous components, and using the average mass density of proteins, we show visualization of the 3D distribution of intracellular water in a variety of scenarios, including mitosis, growing neurons, osmotic shock, and cells responding to pulsed electric fields.

Doppler light scattering from intracellular motions generates phase drifts that can be measured quantitatively using interferometry and digital holography. This talk reviews the current state of digital holographic optical coherence tomography used for quantitative phase analysis of intracellular dynamics. Applications include drug development, in vitro fertilization, developmental biology and cancer therapy selection. In addition, the phase sensitivity of holography enables the quantitative measurement of phase excursions related to anomalous transport inside cells, introducing “Levy-alpha spectroscopy” that may help to diagnose infections and other diseases.

Acral melanoma often presents at advanced stages and responds poorly to standard therapies, necessitating targeted treatments. Using quantitative phase imaging (QPI), we evaluated the effects of Lenvatinib, Sunitinib, and Apatinib on six primary acral melanoma cultures. Lenvatinib showed minimal impact except in HCI-ASM020 and HCI-AM087 at 20 µM. Apatinib and Sunitinib were cytotoxic at 40 and 50 µM, respectively. AM088 cells had delayed responses to Sunitinib (15 hours) and Apatinib (23 hours). EC50 values ranged from 0.01 µM to 136 µM. These results underscore QPI's potential in screening and personalizing acral melanoma therapies.

Ferroptosis, an iron-dependent form of cell death, particularly affects red blood cells due to their high hemoglobin content. We present a high-throughput holographic cytometry (HC) system to evaluate ferroptosis in sickle cell disease (SCD) samples. HC's ability to analyze millions of cells allows for detailed single-cell phenotypical profiling to identify treatment effects, making it a valuable tool for studying ferroptosis interactions in SCD cytological samples.

Lens-free imaging (LFI) is a fast, compact and cost-effective computational imaging technique that is capable of volumetric imaging of sparse samples. Though LFI systems can achieve high lateral resolution (~1 µm) over a large field-of-view, they conventionally assume weak scattering and lack depth resolution. Optical diffraction tomography (ODT), on the other hand, can provide 3D refractive index information with greater axial resolution under relaxed scattering assumptions. However, traditional ODT systems image relatively small sample volumes (<0.05 mm3) using elaborate lens-based configurations. Combining the benefits of these two approaches, we developed a compact lens-free optical diffraction tomography (LF-ODT) technology for high resolution 3D imaging of sparse biological samples including red blood cells, white blood cells and bacteria in a thick volumetric sample (~0.5 - 1 mm3). This research demonstrates the promise of a catheter-inline LFI-ODT technology for continuous real-time monitoring of biological fluids in healthcare settings.

In nano-toxicity testing the analysis of particle-cell interactions is of particular importance as a variety of effects are caused by cell membrane binding and internalization. In a proof-of-concept study, utilizing a commercial tomographic phase microscopy (TPM) system with fluorescence imaging capabilities, we investigated cellular association and uptake of labelled (NR668-loaded poly (alkyl cyanoacrylate) (PACA)) nanoparticles and lipophilic dye-loaded chitosan nano capsules, in human lung epithelial cells and mouse macrophages. Cellular association of particles was analysed by locating the fluorescence signals of the labelled nanoparticles within the TPM refractive index tomograms of the cells. Our results show that correlative imaging with TPM and fluorescence microscopy represents a promising method to detect and localize labelled polymeric particles within single cells and to study cell-particle interactions.

Digital holographic microscopy (DHM) provides quantitative phase images, offering a powerful non-invasive method for studying living cells. By integrating DHM with machine learning and connectomics-inspired graph measures, we study network properties in rat neuronal cultures across maturation stages in vitro. Our framework uses two U-Net models for cell-body and neurite segmentation, achieving accuracies of 0.99 and 0.97, respectively. Besides extracting morphological properties, each segmented image is used to infer a graph, highlighting critical features like density and modularity for maturation stage classification. These results provide a basis for future studies to understand the principles of neuronal network organization in neuronal cultures at the microscopic level.

Grating-based dark-field and phase x-ray imaging aims to extend conventional attenuation-based x-ray imaging by revealing additional microstructural information. A big challenge for translating grating-based dark-field computed tomography to medical applications lies in minimizing the data acquisition time. While a continuously moving detector is ideal, it prohibits conventional phase retrieval and stepping techniques that require multiple projections under the same angle with different grating positions. In this work, we introduce a new algorithm that improves phase retrieval for continuously acquired data and mitigates movement-based cross-talk artifacts.

QPI and virtual staining of live cells and tissues address the challenges of traditional multispectral fluorescent imaging, which requires labor-intensive labeling, limits throughput, and compromises cell health. We report VSNeuromast, a 3D virtual staining model for translating phase images into landmark organelles (nuclei and plasma membrane) in zebrafish neuromasts throughout development. Combined with segmentation and tracking, this approach enables extended imaging of developmental dynamics with minimal photodamage. Our results show that virtual staining rescues dim structures, corrects photobleaching, and provides detailed, prolonged observations of the developmental process, offering a powerful tool for in-vivo developmental biology imaging.

Sickle cell disease (SCD), an inherited red blood cell disorder characterized by the cells' sickle shape, affects nearly 100 million people worldwide. Microscopy of red blood cells is crucial for disease detection. However, conventional methods often rely on bulky, expensive devices and require skilled human analysts to interpret RBC images. To address these challenges, we replaced the lens of the microscope with a phase modulator to capture the opto-biological signature (OBS) of blood samples. Our approach bypasses the need for image reconstruction, enabling direct classification using Machine Learning (ML) algorithms based on statistical features extracted from OBS data. In this project, we focus on optimizing the experimental setup using ray tracing software simulations to enhance the efficiency and accuracy of our classification system, rather than optimizing the ML algorithm itself. In this paper, we have shown how optical design can effectively help enhance the performance of lensless imaging systems.

We developed a microfluidic chip with stop-and-go technology, enabling precise cell halting for quantitative phase imaging (QPI) and classification using machine learning models. This system can image and classify various cell types label-free, with potential lab-on-chip applications like RNA sequencing. We captured QPI images of mouse NIH 3T3, human HEK 293, and HEK 293 cells with lipid deposits, extracting amplitude and phase information. Using convolutional neural networks (CNN), we achieved over 87% accuracy and an F1 score above 95% for HEK 293 with lipids. Phase imaging provided up to 15% higher accuracy than amplitude-only images, demonstrating QPI's superiority.

In vitro mammary epithelial cells (MEC), when 3D cultured in artificial extracellular matrix (aECM), tend to form spheroids that recapitulate many features of the human mammary gland. These 3D MEC cultures can be used as a testbed for understanding the effects of cancer-fighting drugs and potentially cancer-inducing toxicants. Optical coherence tomography (OCT) is a favorable method to capture label-free high-speed intracellular motility within these energetic cells by merit of their unique speckle fluctuation patterns. This talk will cover what we have learned about developing dynamic OCT-based metrics that elucidate MEC responses to drugs and toxicants, discuss a novel non-uniform sampling method for high throughput studies, and demonstrate an application in longitudinal imaging of MEC spheroids under exposure to polyfluoryl alkyl substances (PFAS), a class of chemicals currently of concern as environmental toxicants.

Migrasome plays crucial roles in intercellular communication, during which internal substances, like lipid, are absorbed by another cell, leading to the demand for analyzing substance content inside it. Existing methods for analyzing lipid content need isolation or staining procedures, which hinders observation of original statuses of migrasomes. We propose an analyzation method by using high-sensitivity quantitative phase imaging. Through establishing correlation between retrieved phase and lipid content, we can achieve quantitative lipid analysis based on individual migrasome. We envision our study will pave the way for label-free characterization of intercellular communications through quantitative analysis of lipid content in migrasomes.

In this work, we introduce qtOCT (quantitative transmission Optical Coherence Tomography) for processing data from full-field swept-source OCT in transmission mode. qtOCT provides the possibility to directly and easily recover quantitative, integrated 2D phase information and features temporal gating ability, facilitating the analysis of multiple-scattering biological objects. Experimental results validate qtOCT’s consistency in providing phase information compared to digital holographic microscopy and demonstrate its unique capability to gate time-separated signals in transmission mode. qtOCT advances transmission OCT, especially for applications requiring precise phase information and temporal signal gating.

Recent advancements in optical diffraction tomography (ODT) have introduced inverse-scattering models like the Multi-Slice Beam Propagation (MSBP) method, which enable 3D label-free refractive-index imaging of moderately scattering samples. This method can use both field and intensity data, but comprehensive comparisons of these approaches are lacking. Our study fills this gap by comparing MSBP approaches using field data versus intensity data, using scattering phantoms and biological samples. Additionally, we introduce a low-resolution initial guess from field data to speed up reconstruction and a data augmentation strategy for improved reconstruction fidelity using numerical propagation of field data to create refocused datasets. These works aim to optimize 3D refractive index microscopy using MSBP, enhancing its utility across diverse specimen types.

Digital in-line holographic microscopy (DIHM) provides a low-cost, temporally phase-stable, and non-invasive method for quantitative phase microscopy (QPM) by capturing complete wavefront information. However, its spatial resolution is fundamentally limited by the diffraction limit, as with any optical microscopy system. To address this limitation, we propose integrating microsphere-assisted microscopy (MAM) with DIHM. To experimentally demonstrate the concept, we first use a silica microsphere to assist a 100x objective lens in resolving the grating of a Blu-ray disk. Furthermore, to showcase the effectiveness of the approach with lower magnification objectives commonly used in holographic microscopy, we demonstrate the capability of a 10x objective, assisted by a larger 250 µm microsphere, to resolve the phase grating of a CD. This approach introduces a simple yet efficient method for QPM with a relatively large field of view.

Low-coherence holotomography, a promising label-free imaging technique in biomedicine, reconstructs the volumetric refractive index distribution using partially coherent light. However, various aberrations caused by the imaging system or sample-induced effects, often degrade the volumetric imaging quality. We introduce an aberration correction method for low-coherence holotomography that significantly improves the resolution and accuracy of refractive index deconvolution. This method iteratively retrieves Fourier space aberrations using overlapping information across various intensity measurements under a controlled set of illumination patterns. We validated and demonstrated the effectiveness of this method on various volumetric specimens, including pancreatic tissue and live MCF7 cells. This method holds great promise for various three-dimensional microscopy applications.

We introduce a quantitative phase microscopy (QPM) system for cell analysis, addressing traditional method limitations. This system integrates cell segmentation, confluence determination, and phase mapping into a single, non-invasive, label-free apparatus. It captures detailed data, enhances cellular structure definition, and is ideal for long-term observations, simplifying setups and reducing costs.

Fourier spectrum analysis plays a crucial role in various applications including microscopy, quantitative phase imaging, telescopy, etc. We present an experimental technique to retrieve complex Fourier spectrum in presence of random scattering media by employing structured illumination with single pixel analysis of intensity correlation. Our technique does not rely on interferometry or any phase retrieval algorithm. To avoid huge measurement for temporal samplings, our method leverages the spatial averaging rather than temporal in estimation of the intensity correlation.

The lateral resolution is a quantitative measure of the smallest resolvable structure distinguishable through a microscope, including a Digital Holographic Microscope (DHM). The diffraction limit in microscopes can be surpassed by illuminating the sample with a structured pattern, transferring high spatial frequencies of the object within the transfer function of the system. In this work, we have two novel optical configurations to improve the resolution limit of DHM systems operating in both transmission and reflection modes using common optical elements. Both DHM systems integrate an additional interferometer to generate a structured pattern with tunable modulation frequency and large field of view.

Quantitative phase imaging (QPI) offers 2D label-free live cell observations with minimal toxicity. In recent years, Holographic Tomography (HT) has elevated QPI from 2D imaging to 3D refractive index distribution (RID) reconstruction. We propose a novel, alternative way to 3D RID reconstruction, utilizing z-stacks of phase images obtained by Holographic Incoherent-light-source QPI (hiQPI). To recover high-resolution 3D RID from hiQPI z-stack, we introduce a deep-learning-based algorithm, leveraging a U-net-based convolutional neural network trained on a diverse set of simulated red blood cells and corresponding simulated hiQPI z-stacks.

Bacterial spores are dormant forms of bacteria that have a larger size with a hard outer coat, formed under adverse environmental conditions. Rapid identifying bacterial spores are crucial for food safety, biosecurity and space exploration. However, the conventional methods such as plate count method and microscopic examination after staining are time-consuming and labor-intensive. Herein, we employed NHQLive Live Cell Imaging Analyzer to identify bacterial spore from coccus. Moreover, a microfluidic chip was integrated into the system for high throughput analysis. Due to the progressiveness nature of our approach, we will achieve rapid detection of bacterial spore in a label-free manner.

Bright-field (BF) microscopy is a standard technique employed by cytologists, but it has a number of drawbacks including limited contrast and staining artifacts. To overcome these limitations and improve diagnostic accuracy, have created a refractive index-correlated pseudocoloring scheme for BF images and applied to them in human thyroid cytology as a form of an augmented reality (AR) microscope. We employed correlative holotomography to quantify the Papanicolaou staining BF image and the three-dimensional refractive index (RI) distribution of thyroid fine-needle aspiration biopsy (FNAB) samples in our methodology. The generated pseudocolor image provides a distinct visualization on both the cytological features in the BF image and the subcellular structure in the RI image, with a particular emphasis on the nuclear morphology, which is essential for a precise diagnosis. Application of this algorithm by overlaying the pseudocolor image over BF image as a form of AR microscope highlights the substantial potential of this technology in a variety of fields, such as cancer diagnosis and beyond.

Birefringence is an optical property of materials that have a polarization-dependent refractive index caused by anisotropic structures. For this reason, three-dimensional (3D) birefringence has been widely studied in various industrial and biological applications to reveal the internal alignment configuration of a target sample. Recently, dielectric tensor tomography (DTT) has emerged, enabling a 3D birefringence imaging of anisotropic structures. However, the DTT setup employs an interferometric system vulnerable to mechanical and speckle noise. Here, we present an incoherent DTT method using a non-interferometric system with modulated illumination.

High-accuracy calibration of a deformable mirror is important for the wide range of applications where adaptive optics is used in the feedforward mode, or for introducing a precise phase diversity term. While phase shifting interferometry (PSI) can be used for such calibration, it requires expensive equipment and/or cannot be performed on-tool due to space limitations. In this presentation, we demonstrate our novel method of phase retrieval from several interferograms with introduced arbitrary tilts (phase tilting interferometry) applied to the calibration of a membrane deformable mirror. As the tilts can be introduced manually by simply misaligning the reference mirror, our method represents an inexpensive and easy-to-use alternative to PSI. The method first exploits the global information to establish the phase tilt parameters using the Zoom FFT for determining the maxima location in the Fourier spectra of the interferogram differences and then retrieves the phase locally using the least-squares approach. To decrease the noise print-through in the estimated mirror response functions, linear regression is used on the phases retrieved for several values of the actuator voltages.

Digital holographic microscopy (DHM) provides label-free, non-invasive imaging of cellular structures but is challenged by coherent noise. Polychromatic-DHM (P-DHM) reduces this noise, delivering high-quality images, but is costly and slows dynamic acquisition. A significant dilemma remains: large fields of view (FOV) in DHM capture many cells but result in poor resolution. We have developed a deep learning-based denoising and single-image super-resolution AI (SISR-AI) using P-DHM images. This enhances the low-resolution of large FOV DHM images without additional equipment. Our approach reduces noise and improves spatial resolution while preserving large FOV, a prerequisite for resolving the connectivity of a neural network in culture.

Virtual staining is a technique where images of cells and cellular structures captured by one imaging modality are transformed into images that resemble the appearance of chemically stained cells by means of artificial intelligence. While virtual staining is well established in histology, its applications in cytology, and in particular for the Papanicolaou stain, are limited. In this study we have successfully trained a Generative Adversarial Network (GAN) to virtually stain label-free holotomography images of urine cytology into Papanicolaou-stained images.

Evaluating mechanical stress and strain is crucial for optimizing mechanical components and preventing catastrophic failures. To study stress or strain, optical birefringence can be utilized because it is sensitive to the molecular alignment caused by mechanical stress in materials. A new technique called dielectric tensor tomography (DTT) has recently been developed, allowing for the reconstruction of 3D optical anisotropic structures. In this research, we visualize the 3D strain tensor of micro-indented glass samples using DTT. Utilizing the stress-optic law, we derive the principal strains and their orientations, revealing changes in density and cracks along with their orientations. Our results highlight the potential of DTT for non-destructive strain analysis and its usefulness in characterizing complex strain distributions within materials.

Fabrication of OCT fibre probes in cardiology involves sequentially splicing and cleaving multiple fibres to form a lens. During this process, the splice location, which is invisible under normal illumination, must be identified with micron-accuracy. This paper presents, an approach to splice detection using digital lensless microscopy. A series of digital holograms are captured by moving the grating perpendicular to the beam. This grating shift introduces a certain phase difference between the reference and object beams. Processing the hologram series using the continuous phase shifting method allows for the computation of the object's complex field in the sensor plane. The Fresnel transformation is then used to compute the complex field in the object plane, constructing the amplitude-phase image.

Osteoarthritis (OA) is a leading cause of joint pain and physical disability. Analysis of chondrocytes, resident cells in articular cartilage, is an effective means to study OA pathophysiology as cartilage degeneration is a defining feature of OA. Most current methods for chondrocyte analysis are destructive and incompatible with real-time and continuous characterization. Particularly, fluorescence imaging, which is widely used to characterize chondrocytes, is limited by photobleaching and phototoxicity. Herein, we employed NHQLive Live Cell Imaging Analyzer to characterize chondrocyte responses to inflammatory cytokines in an in vitro OA model. We will also correlate our results with existing methods for chondrocyte analysis.

The molecular ordering of liquid crystals is crucial in materials science and biophysics, necessitating precise understanding through tomographic reconstruction of the dielectric permittivity tensor. This reveals local directors and alignment, providing a comprehensive view of orientational order. We introduce a novel method applying the Rytov approximation to vector waves, reformulated via a modified scattering matrix and matrix logarithm for series expansion, enabling full 3D permittivity tensor reconstruction. Our method successfully reconstructs the 3D permittivity tensor in birefringent liquid crystals and quantifies 3D topological defects, advancing the study of liquid crystal molecular ordering.

Stem cells (SCs) are crucial in regenerative medicine for their ability to differentiate into various cells and tissues needed for disease treatment. Non-invasive monitoring of SC status is an open challenge, as most imaging techniques require labeling that can affect SC integrity. Our study demonstrates improved monitoring of 3D subcellular structures in SC colonies using holotomography (HT) augmented with artificial intelligence (AI). HT, a label-free imaging technique, when integrated with AI, spatially maps subcellular structures including nuclei and mitochondria over a large SC colony. The proposed method reduces cell disturbance, increases acquisition rates, and lowers signal variability, supporting further exploration of SCs' subcellular structures and dynamics.

In this research a novel characterization method for borosilicate transparent wafers is introduced using Wave Front Phase Imaging (WFPI). This innovative wafer geometry measurement technique acquires 16.3 million data points in 12 seconds on a 300mm wafer, providing a lateral resolution of 65µm with the wafer held vertically. Applying this method to measure 300mm transparent Borofloat®33 wafers using our WFPI lab tool setup, which was previously utilized for silicon wafers. This study presents the methodology, results, and analysis of a fully characterized transparent wafer, demonstrating the efficacy, precision, and potential of the WFPI technique in advancing wafer characterization technology.

This work shows how to design and fabricate artificial cells and cell cultures based on phase images of real specimens. New quantitative phase imaging (QPI) systems are becoming tightly integrated and optimized for specific imaging tasks under certain conditions. This integration means that the choice of an imaging target can significantly affect the QPI system's performance. In biomedical applications, phase images undergo automated analysis, traditionally involving segmentation of cells and subcellular features and quantification of biophysical parameters. Commonly used test targets such as bar charts, steps, or microspheres cannot fully characterize a QPI system and its processing pipeline being optimized for biomedical applications. The proposed methodology for designing and fabricating artificial cell phantoms enable benchmarking and end-to-end validation of the QPI systems in real-world applications.

In this research, we utilized the Pseudospectral Time-Domain (PSTD) method to simulate electromagnetic wave propagation and we also combined by other methods such as Nyquist theory and perfect absorption boundary condition, establishing a two dimensional simulation environment. Compared to Finite-Difference Time-Domain (FDTD) method, PSTD method requires significantly fewer grid points per wavelength, enhancing the efficiency of our macroscopic simulations. Moreover, we combined Optical Phase Conjugation (OPC) techniques with different angular spans of the phase conjugate mirror to focus on the effect of light refocusing and identify the optimal angle for light refocusing. The findings of this research provide a good foundation for future researchers, and we are looking forward to applying this numerical simulation method in non-invasive therapeutic applications.

Quantitative phase microscopy (QPM) is crucial for label-free imaging of biological specimens and materials. Due to the diffraction limit, the lateral resolution of QPM cannot satisfy the growing demand to resolve subcellular structures around 100 nm in living cells. Vortex beams offer promising capabilities in optical imaging by enabling tight focusing and non-diffracting propagation. We propose the use of vortex beam for illumination in an off-axis QPM system and developed the imaging theory from vector diffraction. Our preliminary results show the spatial resolution and the image contrast can be improved on both biological and material samples.

Organoids provide a unique intermediate model between conventional 2D cell culture and costly animal models, neither of which closely represent human tissue structure or function. On scales amenable to 96 well plates, organ biology, physiology, and pathophysiology can be studied. These optically thick, 3D models are not well suited to traditional transmission quantitative phase imaging (QPI); however, epi-mode QPI has been shown to be effective at imaging different depths within the tissue. A custom QPI system is presented for this purpose, including custom illumination optics, novel optical modeling, and advanced reconstruction algorithms. This system is used to evaluate refractive index changes in organoid models that can occur during microbial infection.

The dysfunctiong of the hematopoeisis process in the bone marrow induces alteration of its cell behavior. This leads to myelofibrosis and osteosclerosis. Myelofibrosis is an abnormal generation of collagen fibers inside the marrow, whereas osteosclerosis creates herein small parts of dense trabecular bone. We study these two diseases, which often are precursors of blood cancer, by means of label-free quantitative phase and retardance imaging. For this purpose, we use quadriwave lateral shearing interferometry and a process developed earlier that measures phase dependency upon polarization to detect anisotropic objects. This reveals collagen fibers involved in myelofibrosis.

During the last years, digital holographic microscopy (DHM), an interferometric variant of quantitative phase imaging (QPI), has been demonstrated as a promising minimally invasive tool in risk assessment of nanomaterials and cancer cell analysis, as well as for monitoring of blood cell alterations associated with bacterial challenge and perioperative inflammation. However, to establish DHM as a routine method in biomedical laboratories, robust imaging with enhanced throughput is essential. This can be achieved by DHM-based QPI approaches that utilize automated optical microscopes in combination with multi-well plates. In an overview, by representative examples, multi-well plate utilization in automated high throughput QPI with DHM is discussed. Our results demonstrate well plate-based DHM as a promising method for label-free high-content imaging.

Telecentric digital holographic microscopy (T-DHM) allows efficient physical correction of quadratic phase aberrations. However, accurate alignment of T-DHM with polychromatic DHM (P-DHM) is challenging. In fact, there is no mutual afocal alignment in P-DHM due to the different focal lengths of a microscope objective at different wavelengths, i.e., due to axial chromatic aberration. In this work, we demonstrate three numerical approaches (iterative base-surface determination, modified self-reference conjugated hologram, and loss of telecentricity) to correct the phase aberration errors in P-DHM and prove their usefulness for visualizing the fine structure and dynamics of living cells.

We present an image-formation model that decomposes the 3D permittivity tensor into Gell-Mann matrices, providing a unified description of transmission-mode image formation from all linear elastic light-matter interactions. We use the model to develop an efficient inverse algorithm, and we demonstrate 3D imaging of absorption, phase, retardance, and orientation in test targets, cells, and tissues. We demonstrate that our model improves imaging metrics (contrast, resolution, sectioning) and downstream processing tasks (segmentation, tracking, estimation). We make these advances available via an open-source computational imaging library, waveOrder, which we utilize to deploy quantitative label-free imaging of phase and birefringence via a napari plugin, recOrder.

Quantitative Phase Imaging (QPI) maps amplitude- and wavefront-distortions in microscopy samples, enhancing contrast and enabling label-free biophysical studies. High-definition Wavefront Sensing (HD-WFS), a robust QPI technique, operates without a reference arm and uses incoherent light. We introduce DiPSI (Diffuser Phase Sensing and Imaging), a new HD-WFS method using a thin diffuser near a standard camera. This cost-effective setup achieves high phase sensitivity (<λ/500) and real-time imaging (>50Hz). We developed ImageJ and Micro-Manager plugins facilitate DiPSI adoption towards biology labs. DiPSI's potential is showcased for label-free cell classification, particularly to identify retinal organoids-derived cell types for vision restoration therapies, using Deep Learning.

In this work, we introduce a label-free, three-dimensional phase imaging technique that leverages intensity measurements under diffused illumination from light-emitting diodes (LEDs). Our approach achieving diffraction-limited lateral resolution and axial sectioning performance (~500 nm) using a high-numerical-aperture (NA) objective lens. Our method facilitates ultra-fast quantitative characterization of cell morphology, making it highly suitable for diverse biomedical applications.

This research studied how the depth of field of an imaging system affects the reconstruction quality in optical diffraction tomography. With numerical models we simulate the scattered field of a Mie scattering object. Subsequently, we spatially filter the field and propagate it to account for a defocus aberration. The field is used for reconstruction of the refractive index using the filtered backpropagation algorithm. The effect of defocus is quantified in the tomographic reconstruction domain. Results showed that the tomographic depth of field is much larger than the classical incoherent depth of field.

Optical anisotropy revealed by birefringence provides valuable information regarding molecular orientation and tissue structure. The noninvasive measurement of birefringence has attracted much attention with the recent development of quantitative phase imaging (QPI). In this work, we introduce three-dimensional (3D) tensorial polarization differential interference contrast (PDIC) microscopy. It simultaneously reconstructs the 3D phase and birefringence for biological samples from a stack of through-focus PDIC images via an alternating direction method of multipliers (ADMM) approach. We demonstrate accurate 3D phase and birefringence tomographic imaging, outperforming bright-field counterparts, on silica spheres, muscle tissue sections, and various other biological samples. Its potential for imaging thick samples across depths is further evaluated.

Telepresence allows users to perceive distant objects or people as if they were physically present, with potential applications in military, industrial, and medical fields. While commercialization remains distant, telepresence is well-known, often featured in science fiction. Current systems depend on complex hardware and rendering algorithms. In this work, we present a fully holographic telepresence system utilizing speckle-based holographic imaging, eliminating the need for time-intensive computer-generated holography. By applying gradient descent algorithms for real-time field reconstruction, we achieved successful live 3D streaming of dynamic objects, providing a framework to enhance streaming speed and image quality.

In this presentation, I show how we used cross-grating wavefront microscopy (CGM), a wavefront imaging technique based on quadriwave lateral shearing interferometry (QLSI), as a temperature microscopy technique for a large variety of applications. Microscale temperature gradients are generated in the field of view of a microscope by illuminating gold nanoparticles with a laser light at their plasmonic resonance. Using this association of microscale laser heating and CGM temperature imaging, were have been able to investigate the physics of micro bubble generation, water superheating, to introduce the concept of microscale hydrothermal chemistry, and to study life at high temperature. After an introduction of QLSI and CGM, all these applications will be presented.

The resolving power of imaging systems, particularly optical microscopy, is primarily constrained by diffraction. Digital holographic microscopy (DHM) offers diffraction-limited quantitative phase imaging (QPI) with high precision by retrieving complex fields via interferometric recording. Although numerous superresolution (SR) techniques have been developed, alternative methods, like the transport of intensity equation (TIE), offer advantages over interferometric recording. We demonstrate how time and angular multiplexing can enhance resolution limits in low/medium numerical aperture (NA) microscope objectives while providing QPI through the TIE algorithm. This approach, validated experimentally using a commercial microscope, achieves a resolution gain factor of 2.

The open-source software Holovibes integrates AMETEK S711 streaming cameras and EURESYS Coaxlink QSFP+ framegrabbers for high-performance real-time Doppler holography and holographic OCT. Operating at 33,000 frames per second with 384x384px resolution, it supports seamless recording, rendering, and data saving on standard hardware. We present a novel method for estimating total retinal arterial blood flow using Doppler holography. This non-invasive technique segments primary retinal arteries and measures differential Doppler frequency broadening. A forward scattering model is applied to calculate blood flow velocity based on optical properties, and the cross-sectional area of the arteries is used to estimate blood volume flow. The recorded rate in a volunteer was 35 μL/min, though underestimation may occur due to current frame rate limitations. Additionally, 3D OCT data rendering via ultrahigh-speed digital holography accelerates volumetric detection, eliminating mechanical scanning. These advances in real-time retinal imaging open doors to new quantitative techniques for amplitude and phase fluctuation imaging.

Optical diffractive imaging presents a useful framework for the quantitative estimation of refractive index distribution from the measured diffracted field data of transparent objects. Reconstructing the complex refractive index distribution from noisy diffracted field data is often ill-posed. Traditional reconstruction approaches involve assumptions about the optical system or objects, restricting their application to real-world scenarios. This paper presents an iterative blind deconvolution with TV regularization for optical diffractive imaging reconstruction. Our method estimates the refractive index distribution and point spread function (PSF) simultaneously, compensating for unknown aberrations and system imperfections. Extensive simulations and experimental results demonstrate the superiority of the proposed method over extensive algorithms, which are based on iterative reconstruction based on state-of-the-art methods like the FISTA forward-backward splitting method governed by the exact solution of the Lippmann-Schwinger equation. equation.

We introduce a novel phase imaging microscopy technique employing a single multimode fiber through frequency domain modulation. Our method demonstrated exceptional capability in high-contrast imaging of diverse samples, including microspheres, semiconductor chips, and oral epithelial cells.

Embryo imaging and analysis are essential for selecting viable embryos in clinical in vitro fertilization (IVF). Traditional methods depend on 2D morpho-kinetic profiling and manual selection, limited by the absence of advanced 3D imaging tools. In previous work, we evaluated 3D refractive index tomography of early mouse embryos using low-coherence holotomography, but manual segmentation was required. Building on this, we now propose a scalable, deep learning-based framework for automated 3D analysis of early embryogenesis. Our method’s accuracy is validated against manually annotated subcellular masks and applied across different mouse strains.

To address the challenges for accurate trans-illumination phase imaging in thick or semi-transmissive surfaces, we demonstrate a novel epi-illumination phase imaging system, based on a compact multi-camera array microscope (epi-MCAM). By simultaneously acquiring data from the 24 high-resolution epi-illumination microscope units, this system effectively mitigates the standard traditional trade-off between field of view (FOV) and resolution by capturing image data from a proportionally larger area. Half of the incident illumination is blocked in its Fourier plane to utilize illumination-based differential phase contrast (DPC), and A quantitative phase reconstruction based on the phase transfer function is applied to recover the phase information of the sample. The high-quality phase information of semi-reflective specimens over hundreds of square centimeters can be rapidly obtained, which may unlock speed-ups within industrial inspection and machine vision, as well as an effective method to study dynamic biological specimens across macroscopic fields-of-view.

We propose the idea of a partially spatially coherent light-based off-axis quantitative phase imaging system combined with MUSICAL for tracing the nanoscale motion of human sperm cells. Here, we quantified different optical parameters based on the motion trajectory, such as linear velocity, helix period, and radii, and correlated the relation between different parameters. The result indicates no significant direct correlation occurs between the helix period and radius with respect to the linear velocity of the spermatozoa. Further, we computed the Pearson correlation coefficient for the whole dataset, and k-means clustering (with 2 and 3 clusters) was used to analyze parameter correlations. Results show that spermatozoa with linear velocities <63 μm/sec exhibit helical paths with higher helix periods and smaller radii. The proposed concept can further be applied to the motion tracing of biological cells, offering the possibility to, e.g., cell monitoring, cell tracking, manipulation, select and classify different healthy and unhealthy cells based on their motion trace and quantified parameters.

Chemical composition and charge state is central to new battery development. X-ray spectral imaging can provide this information. X-ray absorption near edge structure (Xanes) further provides charge state. The measurement is usually carried out by measuring the intensity of the transmitted light. Here we devise a new setup based on differential phase contrast to measure the spectral phase and light intensity. Providing at the same time chemical composition and high-resolution structural information.

Cancer is a major global cause of death, responsible for nearly one-sixth of all fatalities. By 2040, new cancer cases are expected to reach 27.5 million, with 16.2 million deaths annually. Blood cancers, such as Diffuse Large B-cell Lymphoma (DLBCL), are particularly deadly, impacting blood cell production and immune function. Early detection is crucial for effective treatment. Histopathology is the standard for diagnosing blood cancers, but rising case numbers make manual assessment challenging. This study presents a unified framework for diagnosing DLBCL using histopathology images with a convolutional neural network (CNN) and an attention module. The optimized model significantly reduces inter-class dispersion and within-class variance, achieving 97.33% accuracy with recall, precision, and specificity, all above 97%. This approach enhances automated histopathology analysis, offering faster and more accurate diagnostics for blood cancer.

The limited option for Xray optical elements makes lens-less imaging particularly appealing. A popular method for phase imaging in x-ray is to use a speckle illumination and analyses the speckle pattern shift when passing through the object using correlation. Different algorithms have been developed for speckle based Xray imaging however most are based on acquisition of multiple shot of transmitted image under different speckle illuminations. Here we develop a preconditioned optimization algorithm for single shot phase and absorption imaging. Our method both speeds up image acquisition, reduces artifacts and improves resolution by taking illumination incoherence into account. The obtained holograms are used to realize tomographic imaging of biological samples.

Shadow Moiré demodulation is a widely recognized profilometry technique for dynamic three-dimensional measurement. Conventional applications rely on phase-shifting processing, which presents challenges when high resolution and accuracy are required. To address this, our group introduces an improvement through second-harmonic information processing, making the technique more precise and accurate compared to traditional methods. Also, a single-shot Shadow Moiré technique with a fringe pattern was employed for phase retrieval. The technique revealed enhanced recovery of high-definition information and micrometer detail detection, providing the possibility of reconstructions with high detail. These findings highlight the technique’s potential to open new opportunities for future applications.

We are proposing a high-resolution blind-phase retrieval technique using color fringe patterns. In the proposed method, we use a single channel processing using a phase-shifting algorithm for phase retrieval. This technique adequately decouples the color information and corrects the detuning error because of the crosstalk and color imbalance. A calibration process is not required since we combined the information of the three channels to obtain a low-noise-phase retrieval. We achieve high spatial resolution, and the sharp edges are preserved for complex surfaces. As well, this technique can be used even in abrupt changes in background or in the fringe modulation.

The dry mass and the orientation of biomolecules are sensitive readouts of the architecture of organelles, cells, and tissues. These properties can be imaged without a label by measuring their permittivity tensor (PT), which describes how biomolecules affect the phase and polarization of light. Three-dimensional (3D) imaging of PT has been challenging. I'll discuss a recent label-free computational microscopy technique, PT imaging (PTI), for the 3D measurement of PT. PTI encodes the invisible PT into images using oblique illumination, polarization-sensitive detection, and volumetric sampling. PTI decodes the PT using an inverse algorithm based on an accurate vectorial image formation model. In this talk, I'll summarize forward models and strategies for reconstruction for robust 3D imaging at the spatial scales of organelles, cells, and tissues.