arXiv subjects

Julianna Winnik

Publications and source records attributed to Julianna Winnik.

9 recordsLinked to original sources

arXiv2026

Lateral shearing optical diffraction tomography of brain organoid with reduced spatial coherence

Optical diffraction tomography (ODT) is a powerful technique for quantitative, label-free reconstruction of the three-dimensional refractive index (RI) distribution of biological samples. While ODT is well established for imaging thin, weakly scattering samples, it encounters significant challenges when applied to heterogeneous, strongly scattering thick samples such as tissues and organoids. In this work, a novel common-path interferometric approach to ODT is presented, specifically designed for the RI reconstruction of heterogeneous and highly scattering samples at high temporal stability. The proposed technique, termed lateral shearing (LS)-ODT, incorporates partial lateral shearing off-axis interferometry to suppress the effects of multiple scattering, similar to the mechanism in differential interference contrast (DIC) microscopy, which is widely used for imaging thick specimens. Additionally, the LS-ODT system uses dynamic speckle illumination to enhance both spatial phase and RI sensitivity compared to laser-based ODT systems. The effectiveness of this method is demonstrated through experiments on a cell phantom. Its robustness and accuracy are further validated across a wide range of samples, including mouse kidney tissue sections and brain organoids derived from human induced pluripotent stem cells (iPSCs), in both thin and thick sections. Furthermore, correlative fluorescence and RI tomography of the organoids highlights the potential of LS-ODT to enhance and support a broad spectrum of biomedical studies, particularly in the field of histology.

physics.optics↗

arXiv2026

YOSO: single-frame Gerchberg-Saxton phase retrieval with AI-based data augmentation for in-line holography

We present YOSO (You Only Shot Once), a single-frame phase retrieval framework for digital in-line holographic microscopy (DIHM) in which supervised deep learning is used to numerically generate an additional hologram corresponding to different defocus distance, creating a so-called multi-height dataset, which is then conventionally processed with a well-established Gerchberg-Saxton (GS) algorithm. YOSO is trained on computer-generated data derived from natural images, enabling strong generalization. The selected multi-scale ResNet architecture enables rapid training in under two hours on a mid-range workstation, which is done only once, enabling efficient inference thereafter. We further show that YOSO network can process inputs of varying spatial dimensions, allowing training on small inputs and direct inference on full-sized holograms while bypassing patch-and-stitch procedure. A further advantage of YOSO is its physics-consistent hologram padding, which replaces conventional zero or edge-value padding with a physically grounded approach compatible with the GS framework. The YOSO framework is tested on various systems (lens-based and lensless DIHM) and diverse samples: a resolution test target, adherent and suspended biological cells, and a mouse brain slice. The results show that YOSO is compatible with 3D objects and correctly recovers defocused object wave features, enabling holographic postprocessing such as numerical refocusing. The results of this work are available publicly as software for end-to-end implementation.

physics.optics↗

arXiv2026

Near-infrared lensless holographic microscopy on a visible sensor enables label-free high-throughput imaging in strong scattering

Lensless digital holographic microscopy (LDHM) relies on interference between an unscattered reference wave and a weakly scattered object wave - an assumption that rapidly fails in turbid samples under multiple scattering. To overcome this limitation, we present near-infrared LDHM (NIR-LDHM), a in-line holographic platform that operates up to the silicon cutoff (~1100 nm) using a conventional board-level CMOS sensor designed for visible (VIS) imaging. Using tissue-mimicking milk scattering layers and calibrated resolution targets, we quantify reconstruction performance versus wavelength, scattering strength, and sample-sensor distance. NIR-LDHM maintains resolvable features through scattering layers up to ~1.4 mm, whereas visible regime fails to resolve features below ~350 um. Importantly, despite a detector quantum efficiency of only 0.19% at 1100 nm, robust reconstructions are obtained under low-photon-budget conditions. We further identify advantageous mechanism by which increasing the sample-sensor distance from ~3 to 12 mm improves lateral resolution by twofold under strong scattering. Finally, we demonstrate wide-field, label-free amplitude-phase imaging of uncleared mouse tissues, resolving internal structure in brain and liver slices up to ~250 um and ~60 um, respectively. By extending lensless complex-field microscopy into strongly scattering regimes with minimal hardware changes, this work has relevance to computational imaging through complex media and biophotonics.

physics.optics↗

arXiv2025

Low-dose Chemically Specific Bioimaging via Deep-UV Lensless Holographic Microscopy on a Standard Camera

Deep-ultraviolet (DUV) microscopy can provide label-free biochemical contrast by exploiting the intrinsic absorption of nucleic acids, proteins and lipids, offering chemically specific morphological information that complements structural optical thickness contrast from phase-sensitive imaging. However, existing DUV microscopes typically rely on specialized optics and DUV-sensitive cameras, which restrict field of view, increase system complexity and cost, and often require high illumination doses that risk photodamage. Here, we report a low-dose deep-UV lensless holographic microscopy platform that uses standard board-level CMOS sensors designed for visible light, eliminating all imaging optics and dedicated DUV detectors. Our system achieves large field-of-view (up to 116 mm2) DUV imaging with low illumination and label-free phase and chemically specific amplitude contrast. A specialized defocus/wavelength diverse pixel super-resolution reconstruction with total-variation regularization and robust autofocusing halves the effective sensor pixel pitch and yields down to 870 nm lateral resolution. We demonstrate chemically specific, label-free bioimaging on challenging specimens, including Saccharomyces cerevisiae, extracellular vesicles and unstained mouse liver tissue. In liver sections, imaging at λ = 330 nm reveals lipid- and retinoid-rich accumulations that co-localize with Oil Red O staining, enabling label-free identification of hepatic stellate (Ito) cells. This combination of low-dose operation, chemically specific contrast and standard CMOS hardware establishes DUV lensless holographic microscopy as a practical and scalable route to high-content submicron-resolution whole-slide preparation-free bioimaging without exogenous labels.

physics.optics↗

arXiv2025

Bayesian inference for precise and uncertainty-quantified single-shot widefield interferometric geometrical nanometrology

Advanced geometrical nanometrology is critical for process control in semiconductor manufacturing, supporting applications in, e.g., photonic integrated circuits, nanoelectronics, and emerging quantum and optoelectronic technologies. Widefield interferometric approach provide a cost-effective, non-destructive solution for characterizing semiconductor optical waveguides, which are fundamental to nanophotonic devices. This work presents a Bayesian inference framework, implemented using Dynamic Nested Sampling, for estimating geometric parameters - such as width and height - of a semiconductor optical waveguide from a single widefield interferogram. The proposed framework reduces the need of leveraging near field scanning microscopy methods for measurements. The notable advantage is that Bayesian statistics not only provide the estimated parameter values but also quantify the uncertainty of the inference results and the fitness of the used model. The proposed full-field, single-shot interferometric approach, supported by Bayesian-based data analysis, achieves high accuracy and sensitivity - down to successful measurement of 8 nm rib waveguide - while remaining resilient to noise. Thus, the demonstrated methodology provides a cost-effective, robust, and scalable tool for semiconductor fabrication monitoring and process verification, as confirmed by both numerical simulations and experimental validation on optical waveguides. This method contributes to high-precision nanometrology by integrating advanced statistical modeling and inference techniques.

physics.optics↗

arXiv2025

OSI-flex: Optimization-Based Shearing Interferometry for Joint Phase and Shear Estimation Using a Flexible Open-Source Framework

Shearing interferometry is a common-path quantitative phase imaging technique in which an object beam interferes with a laterally shifted replica of itself, providing high temporal stability, reduced sensitivity to environmental noise, compact design, and compatibility with partially coherent illumination that suppresses coherence-related artifacts. Its principal limitation, however, is that it yields only sheared phase-difference measurements rather than the absolute phase, thereby requiring additional reconstruction step. In this work, we introduce OSI-flex, a flexible, open-source computational framework for quantitative phase reconstruction from sheared phase-difference measurements. The method leverages modern machine learning tools, namely automatic differentiation and the advanced ADAM (Adaptive Moment Estimation) optimizer. The method simultaneously estimates the phase and shear values, enabling it to adapt to experimental conditions where the shear cannot be precisely determined. Because defining shear value is inherently difficult in most systems, yet crucial for effective phase reconstruction, this joint optimization leads to robust and reliable phase retrieval. OSI-flex is highly versatile, supporting arbitrary numbers, magnitudes, and orientations of shear vectors. While optimal reconstruction is achieved with two orthogonal shears, the inclusion of regularization - specifically total variation minimization and sign constraint - enables OSI-flex to remain effective with nonorthogonal or even single-shear measurements. Moreover, OSI-flex accommodates a wide range of shear magnitudes, from subpixel (differential configuration) to several dozen pixels (semi-total shear configuration). Validation with simulations and experimental data confirms quantitative accuracy on calibrated phase objects and demonstrates robustness with 3D-printed cell phantom and follicular thyroid cells.

physics.optics↗

arXiv2025

Gigavoxel-Scale Multiple-Scattering-Aware Lensless Holotomography

Holotomography (HT) has revolutionized quantitative label-free 3D imaging, yet conventional lens-based implementations are fundamentally constrained in field-of-view (FOV) and imaging depth, limiting their utility for critical high-throughput applications in material and life sciences. Lensless HT (LHT) offers a promising alternative for large-volume examination, however existing approaches fail to accurately reconstruct highly scattering samples over extended depths, which remains a critical challenge in optical imaging field. Here, we introduce a gigavoxel-scale, multiple-scattering-aware LHT with a large FOV (surpassing 0.6 cm2), millimeter-scale axial range and pixel level (~2.4 micron) resolution. Our approach leverages a multi-wavelength, oblique-illumination hologram reconstruction and a robust, automatic illumination angle calibration, which are necessary for precise large-volume 3D holographic reconstruction. Moreover, we propose optimization-driven multi-slice tomographic framework to accurately capture multiple-scattering effects outperforming first order Born/Rytov-based inversions. To rigorously validate our method, we reconstruct bespoke multi-layer two-photon polymerized test structure over a 1.7 mm imaging depth and 25 mm2 FOV, yielding an unprecedented 3D space-bandwidth product exceeding a gigavoxel level. Furthermore, we demonstrate for the first time on-chip label-free imaging of entire 500-micron-thick tissue slice of optically-cleared mouse brain. With the proposed method, we aim to unlock powerful new capabilities for large-scale, quantitative, label-free 3D imaging across biomedicine, neuroscience, material sciences and beyond.

physics.optics↗

arXiv2024

Gradient Optical Diffraction Tomography

Optical diffraction tomography (ODT) enables non-invasive information-rich 3D refractive index (RI) reconstruction of unimpaired transparent biological and technical samples, crucial in biomedical research, optical metrology, materials sciences, and other fields. ODT bypasses the inherent limitations of 2D integrated quantitative phase imaging methods. To increase the signal-to-noise ratio easy-to-implement common-path shearing interferometry setups are successfully combined with low spatiotemporal coherence of illumination. The need for self-interference generated holograms, with small shear values, critically impedes the analysis of dense and thick samples, e.g., cell cultures, tissue sections, and embryos/organoids. Phase gradient imaging techniques, deployed as a popular solution in the small shear regime, up to now were constrained to 2D integrated quasi-quantitative phase imaging and z-scanning for depth resolution. To fill this significant scientific gap, we propose a novel gradient optical diffraction tomography (GODT) method. The GODT uses coherence-tailored illumination-scanning sequence of phase gradient measurements to tomographically reconstruct, for the first time to the best of our knowledge, a derivative of the 3D RI distribution (in the shear direction) with clearly visible 3D sample structure and high sensitivity to its spatial variations. We present a mathematically rigorous theory based on the first-order Rytov approximation behind the new method, validate it using simulations deploying numerical Shepp-Logan target and corroborate experimentally via successful tomographic imaging of the calibrated nano-printed cell phantom and efficient examination of neural cells. This novel imaging modality opens new possibilities in biomedical quantitative phase imaging, advancing the field and putting forward a first of a kind contrast domain: 3D RI gradient.

physics.optics↗

arXiv2023

Versatile optimization-based speed-up method for autofocusing in digital holographic microscopy

We propose a speed-up method for the in-focus plane detection in digital holographic microscopy that can be applied to a broad class of autofocusing algorithms that involve repetitive propagation of an object wave to various axial locations to decide the in-focus position. The classical autofocusing algorithms apply a uniform search strategy, i.e., they probe multiple, uniformly distributed axial locations, which leads to heavy computational overhead. Our method substantially reduces the computational load, without sacrificing the accuracy, by skillfully selecting the next location to investigate, which results in a decreased total number of probed propagation distances. This is achieved by applying the golden selection search with parabolic interpolation, which is the gold standard for tackling single-variable optimization problems. The proposed approach is successfully applied to three diverse autofocusing cases, providing up to 136-fold speed-up.

q-bio.QM↗