Imaging Of Embryos Research Articles

Artificial intelligence: Friend or foe?

Artificial intelligence: Friend or foe?

Transmissive multifocal laser speckle contrast imaging through thick tissue

Laser speckle contrast imaging (LSCI) is a powerful tool for monitoring blood flow changes in tissue or vessels in vivo, but its applications are limited by shallow penetration depth under reflective imaging configuration. The traditional LSCI setup has been used in transmissive imaging for depth extension up to [Formula: see text]–[Formula: see text] ([Formula: see text] is the transport mean free path), but the blood flow estimation is biased due to the depth uncertainty in large depth of field (DOF) images. In this study, we propose a transmissive multifocal LSCI for depth-resolved blood flow in thick tissue, further extending the transmissive LSCI for tissue thickness up to [Formula: see text]. The limited-DOF imaging system is applied to the multifocal acquisition, and the depth of the vessel is estimated using a robust visibility parameter [Formula: see text] in the coherent domain. The accuracy and linearity of depth estimation are tested by Monte Carlo simulations. Based on the proposed method, the model of contrast analysis resolving the depth information is established and verified in a phantom experiment. We demonstrated its effectiveness in acquiring depth-resolved vessel structures and flow dynamics in in vivo imaging of chick embryos.

Surface-Acoustic-Wave (SAW)-Driven Device for Three-Dimensional Cell Manipulation.

This paper proposes a system with a surface-acoustic wave (SAWs) to perform noninvasive cell orientation. The system employed a SAW chip with a pair of interdigital transducers (IDTs) to rotate embryos around the X and Y axis using acoustic streaming force. Instead of rotating the entire embryos like other methods, the proposed system rotates the cytoplasm alone through the cell chorion. We evaluated the cellular structure recognition algorithm and the rotation control using 100 embryo images and 30 zebrafish embryos. The system successfully recognized all required cellar structures for visual feedback. Furthermore, it rotated all cells into the desired position, including 26 cases completed within 10s with an orientation angle error of less than 4°. All 30 embryos hatched after manipulation. The proposed technique can automatically rotate the cytoplasm through the cell chorion noninvasively. The system provides a starting point for noninvasive cell manipulation tasks, such as fast intracellular structure scanning and analysis, and microinjection.

A spiral microfluidic device for rapid sorting, trapping, and long-term live imaging of Caenorhabditis elegans embryos

Caenorhabditis elegans embryos have been widely used to study cellular processes and developmental regulation at early stages. However, most existing microfluidic devices focus on the studies of larval or adult worms rather than embryos. To accurately study the real-time dynamics of embryonic development under different conditions, many technical barriers must be overcome; these can include single-embryo sorting and immobilization, precise control of the experimental environment, and long-term live imaging of embryos. This paper reports a spiral microfluidic device for effective sorting, trapping, and long-term live imaging of single C. elegans embryos under precisely controlled experimental conditions. The device successfully sorts embryos from a mixed population of C. elegans at different developmental stages via Dean vortices generated inside a spiral microchannel and traps the sorted embryos at single-cell resolution through hydrodynamic traps on the sidewall of the spiral channel for long-term imaging. Through the well-controlled microenvironment inside the microfluidic device, the response of the trapped C. elegans embryos to mechanical and chemical stimulation can be quantitatively measured. The experimental results show that a gentle hydrodynamic force would induce faster growth of embryos, and embryos developmentally arrested in the high-salinity solution could be rescued by the M9 buffer. The microfluidic device provides new avenues for easy, rapid, high-content screening of C. elegans embryos.

Dynamics of gametes and embryos in the oviduct: what can in vivo imaging reveal?

In vivo imaging of gametes and embryos in the oviduct enables new studies of the native processes that lead to fertilization and pregnancy. This review article discusses recent advancements in the in vivo imaging methods and insights which contribute to understanding the oviductal function. Understanding the physiological dynamics of gametes and embryos in the fallopian tube (oviduct) has significant implications for managing reproductive disorders and improving assisted reproductive technologies. Recent advancements in imaging of the mouse oviduct in vivo uncovered fascinating dynamics of gametes and embryos in their native states. These new imaging approaches and observations are bringing exciting momentum to uncover the otherwise-hidden processes orchestrating fertilization and pregnancy. For mechanistic investigations, in vivo imaging in genetic mouse models enables dynamic phenotyping of gene functions in the reproductive process. Here, we review these imaging methods, discuss insights recently revealed by in vivo imaging, and comment on emerging directions, aiming to stimulate new in vivo studies of reproductive dynamics.

Endoderm-derived islet1-expressing cells differentiate into endothelial cells to function as the vascular HSPC niche in zebrafish.

Endothelial cells (ECs) line blood vessels and serve as a niche for hematopoietic stem and progenitor cells (HSPCs). Recent data point to tissue-specific EC specialization as well as heterogeneity; however, it remains unclear how ECs acquire these properties. Here, by combining live-imaging-based lineage-tracing and single-cell transcriptomics in zebrafish embryos, we identify an unexpected origin for part of the vascular HSPC niche. We find that islet1 (isl1)-expressing cells are the progenitors of the venous ECs that constitute the majority of the HSPC niche. These isl1-expressing cells surprisingly originate from the endoderm and differentiate into ECs in a process dependent on Bmp-Smad signaling and subsequently requiring npas4l (cloche) function. Single-cell RNA sequencing analyses show that isl1-derived ECs express a set of genes that reflect their distinct origin. This study demonstrates that endothelial specialization in the HSPC niche is determined at least in part by the origin of the ECs.

Graded mesoderm assembly governs cell fate and morphogenesis of the early mammalian heart

Using four-dimensional whole-embryo light sheet imaging with improved and accessible computational tools, we longitudinally reconstruct early murine cardiac development at single-cell resolution. Nascent mesoderm progenitors form opposing density and motility gradients, converting the temporal birth sequence of gastrulation into a spatial anterolateral-to-posteromedial arrangement. Migrating precardiac mesoderm does not strictly preserve cellular neighbor relationships, and spatial patterns only become solidified as the cardiac crescent emerges. Progenitors undergo a mesenchymal-to-epithelial transition, with a first heart field (FHF) ridge apposing a motile juxta-cardiac field (JCF). Anchored along the ridge, the FHF epithelium rotates the JCF forward to form the initial heart tube, along with push-pull morphodynamics of the second heart field. In Mesp1 mutants that fail to make a cardiac crescent, mesoderm remains highly motile but directionally incoherent, resulting in density gradient inversion. Our practicable live embryo imaging approach defines spatial origins and behaviors of cardiac progenitors and identifies their unanticipated morphological transitions.

Longitudinal in Utero Analysis of Engrailed-1 Knockout Mouse Embryonic Phenotypes Using High-Frequency Ultrasound.

Large-scale international efforts to generate and analyze loss-of-function mutations in each of the approximately 20,000 protein-encoding gene mutations are ongoing using the "knockout" mouse as a model organism. Because one-third of gene knockouts are expected to result in embryonic lethality, it is important to develop non-invasive in utero imaging methods to detect and monitor mutant phenotypes in mouse embryos. We describe the utility of 3-D high-frequency (40-MHz) ultrasound (HFU) for longitudinal in utero imaging of mouse embryos between embryonic days (E) 11.5 and E14.5, which represent critical stages of brain and organ development. Engrailed-1 knockout (En1-ko) mouse embryos and their normal control littermates were imaged with HFU in 3-D, enabling visualization of morphological phenotypes in the developing brains, limbs and heads of the En1-ko embryos. Recently developed deep learning approaches were used to automatically segment the embryonic brain ventricles and bodies from the 3-D HFU images, allowing quantitative volumetric analyses of the En1-ko brain phenotypes. Taken together, these results show great promise for the application of longitudinal 3-D HFU to analyze knockout mouse embryos in utero.

Advancements in the future of automating micromanipulation techniques in the IVF laboratory using deep convolutional neural networks.

To determine if deep learning artificial intelligence algorithms can be used to accurately identify key morphologic landmarks on oocytes and cleavage stage embryo images for micromanipulation procedures such as intracytoplasmic sperm injection (ICSI) or assisted hatching (AH). Two convolutional neural network (CNN) models were trained, validated, and tested over three replicates to identify key morphologic landmarks used to guide embryologists when performing micromanipulation procedures. The first model (CNN-ICSI) was trained (n = 13,992), validated (n = 1920), and tested (n = 3900) to identify the optimal location for ICSI through polar body identification. The second model (CNN-AH) was trained (n = 13,908), validated (n = 1908), and tested (n = 3888) to identify the optimal location for AH on the zona pellucida that maximizes distance from healthy blastomeres. The CNN-ICSI model accurately identified the polar body and corresponding optimal ICSI location with 98.9% accuracy (95% CI 98.5-99.2%) with a receiver operator characteristic (ROC) with micro and macro area under the curves (AUC) of 1. The CNN-AH model accurately identified the optimal AH location with 99.41% accuracy (95% CI 99.11-99.62%) with a ROC with micro and macro AUCs of 1. Deep CNN models demonstrate powerful potential in accurately identifying key landmarks on oocytes and cleavage stage embryos for micromanipulation. These findings are novel, essential stepping stones in the automation of micromanipulation procedures.

Biocompatible, Biodegradable, and Improved Fluorescent Silicon Quantum Dots for Zebrafish Imaging

One of the greatest benefits of nanomedicine elucidated to date includes the non-invasive tracking and monitoring of living organisms by the selective uptake of harmless metallic nanoparticles. Several nanoscale probes have been employed for biomolecular imaging. Among them, fluorescent nanoscale silicon materials have been recently established with a strong and safe potential for bioimaging and biosensing applications due to their bright fluorescence coupled with strong photostability, biocompatibility and negligible toxicity. Herein, we developed high-quality silicon nanomaterials (4–5 nm; SiNPs) as biological fluorescent probes for bioimaging of living organisms through an easy aquatic synthesis method with a quantum yield of ∼8%. In this regard, we report that the presently synthesized SiNPs-based sensors/probes are attractive materials for solvent-based fluorescence measurements and are biocompatible, non-toxic, highly photo-stable and pH stable. Most importantly, their fluorescence lifetime is much longer than that of native probes in living cells. Thus, these presently formulated SiNPs are improved fluorescent probes for in vivo biological imaging in zebra fish embryos as well as numerous other living organisms and, thus, should be further studied.

Application of Ultrashort Lasers in Developmental Biology: A Review

The evolution of laser technologies and the invention of ultrashort laser pulses have resulted in a sharp jump in laser applications in life sciences. Developmental biology is no exception. The unique ability of ultrashort laser pulses to deposit energy into a microscopic volume in the bulk of transparent material without disrupting the surrounding tissues makes ultrashort lasers a versatile tool for precise microsurgery of cells and subcellular components within structurally complex and fragile specimens like embryos as well as for high-resolution imaging of embryonic processes and developmental mechanisms. Here, we present an overview of recent applications of ultrashort lasers in developmental biology, including techniques of noncontact laser-assisted microsurgery of preimplantation mammalian embryos for oocyte/blastomere enucleation and embryonic cell fusion, as well as techniques of optical transfection and injection for targeted delivery of biomolecules into living embryos and laser-mediated microsurgery of externally developing embryos. Possible applications of ultrashort laser pulses for use in Assisted Reproductive Technologies are also highlighted. Moreover, we discuss various nonlinear optical microscopy techniques (two-photon excited fluorescence, second and third harmonic generation, and coherent Raman scattering) and their application for label-free non-invasive imaging of embryos in their unperturbed state or post-laser-induced modifications.

Embryo Grading after In Vitro Fertilization using YOLO

In vitro fertilization is an implementation of Assistive Reproductive Technology. This technology can produce embryos outside the mother's womb by manipulating gametes outside the human body. The success rate of in vitro fertilization is the selection of good-grading embryos. In this study, the authors used Yolo Version 3 to perform object detection objectively by introducing grades for each embryo image. The author uses an embryo image sourced from the Indonesian Medical Education and Research Institute with information on the quality of the embryo. In this study, the author separated the data into two schemes. The first scheme separates data into training data of 70%, 15% validation data, and 15% for testing data. The second scheme uses a Stratified K-Fold Cross-Validation with a fold value =3. In training, the writer configures the values ??of Max Batches=6000, Steps=4800,5400, Batch=64, and Subdivision=16 by doing image augmentation (saturation=1.5, exposure=1.5, hue=0.1, jitter=0.3, random=1). For each of the obtained mAP (Mean Average Precision) values ??for data separation schemes, one is 100.00% in the 6000th iteration, while for the two-data separation scheme, the highest mAP is 97.33%.% in the fold=3 and 5000th iteration. It means that both separation schemes are sufficient in terms of mAP.

In Vivo Imaging of Mammalian Embryos by NIR-I Photoacoustic Tomography and NIR-II Optical Coherence Tomography Using Gold Nanostars as Multifunctional Contrast Agents

High quality visualization of live mammalian embryo is an important requirement for studying fetal development. Photoacoustic tomography (PAT) and optical coherence tomography (OCT) are two advanced imaging modalities that have been utilized for embryonic imaging separately. However, high contrast, multiscale, and deep tissue imaging of embryos in vivo remains challenging. Here, we demonstrate multimodal contrast-enhanced imaging of live embryos by gold nanostar-meditated near-infrared I (NIR-I) PAT and near-infrared II (NIR-II) OCT. We investigate intravenous and intravaginal administration of gold nanostars, and significant enhancement of signal, image contrast, and imaging depth are achieved for both PAT and OCT. Our findings may facilitate detailed studies of the morphological and functional changes during embryonic development.

Bnip3lb Regulated Mitophagy Maintains the Embryonic Pool of Hematopoietic Stem Cells By Protecting Them from ROS Induced Apoptosis

We have previously shown that Reactive Oxygen Species (ROS) promote Hematopoietic Stem and Progenitor Cell (HSPC) specification in the dorsal aorta, however, in the adult marrow it is established that HSPCs must maintain low levels of ROS to protect the stem cell pool against excessive cycling, differentiation, loss of stemness and apoptosis. We aim to elucidate the developmental timing and the mechanism by which HSCs transition from their initial dependence on ROS, to their precise regulation of it in the adult marrow. Since adult HSPCs regulate ROS by maintaining high levels of mitophagy, a process in which damaged mitochondria, which disproportionately generate ROS, are removed from the cell and recycled, we sought to define the initiation, regulation and impact of mitophagy in embryonic HSPC development. Our recent data shows that oxidative stress limits HSPC numbers as early as their colonization of the secondary hematopoietic niche, the Caudal Hematopoietic Tissue (CHT). This is the zebrafish analog of the mammalian fetal liver in that it is the tissue to which the HSPCs migrate after emerging from the hemogenic endothelium of the dorsal aorta. Once in the CHT they mature and proliferate before migrating again to their terminal niche in the marrow. In contrast to published data showing that ROS promotes HSPC specification in the dorsal aorta, when ROS levels were increased in the embryo using a morpholino against PRDX1 we saw a decrease in runx1/c-myb expression in the CHT. Conversely, when ROS was decreased using the antioxidant N-acetyl cystine (NAC), there was an increase in HSPC marker expression. By live imaging of zebrafish embryos expressing a recently developed fluorescent transgenic mitophagy-reporter, Tg(ubi:mitoGR), we observed that mitophagy is low in the hemogenic endothelial tissue which gives rise to HSPCs, but is upregulated as HSPCs emerge and begin to migrate to the CHT, and is occurring at high levels at the developmental stage when ROS begins to become damaging to the HSPC population. Our scRNAseq analysis and in situ-hybridization experiments allowed us to examine which genes and pathways were regulating the induction of mitophagy in these cells. These data show that the increase in mitophagy induction in the HSPC population at this timepoint coincides precisely with the initialization of expression of the bnip3lb mitophagy receptor gene. Fluorescent whole-mount in situ hybridization experiments demonstrated that bnip3lb is co-expressed with runx1 (a HSPC marker), in the CHT (figure 1), and, by examining our Tg(ubi:mitoGR) mitophagy reporter line after morpholino knock down of bnip3lb we confirmed that it is this gene which is driving mitophagy in HSPCs at this developmental stage. NIX, the murine homolog of bnip3lb, is responsible for developmentally regulated mitophagy in reticulocytes, lens fibers and macrophages, but has, thus far, had no reported role in maintaining HSPCs. Functionally, mitophagy appears to be essential for the survival of the HSPC population in the CHT. Reducing mitophagy via morpholino directed knockdown of bnip3lb has no effect on the initial specification of HSPCs in the dorsal aorta, but significantly decreases HSPC marker expression in the CHT, as measured by runx1/c-myb in-situ hybridization, and quantified by CD41:GFP HSPC reporter expression. In contrast upregulating mitophagy, which can be done via a novel heat-shock driven ΔOTC transgenic line or via small molecule mitophagy induction, elevates HSPC numbers in the CHT. Mechanistically, we found that bnip3lb knockdown increases detection of CellROX, a fluorescent dye which reports ROS levels, and increases the number of acridine orange+ apoptotic runx1+ HSPCs. Further, reducing ROS with antioxidant NAC reduces apoptosis and rescues the HSPC population (figure 2). We therefore propose that developmentally programmed mitophagy directed by bnip3lb is responsible for protecting proliferative embryonic HSPCs from the harmful effects of oxidative stress while the HSPC pool is expanding to seed lifelong hematopoiesis. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Genetic algorithm–assisted machine learning for clinical pregnancy prediction in in vitro fertilization

A clinical pregnancy prediction model was developed by implementing machine learning technology that uses a combination of static images and medical data to calculate the outcome of an in vitro fertilization cycle. To provide a system that can accurately and sufficiently assist with decision making that is critical to in vitro fertilization cycles, primarily embryo selection. Historical medical data, which consist of clinical information and a complete transferred embryo image dataset, of 697 patients who underwent unique in vitro fertilization were collected. Various techniques of machine learning were used, namely decision tree, random forest, and gradient boosting; each technique used the same data configuration for performance comparison and was subsequently optimized using genetic algorithm. A prediction model with a peak accuracy of approximately 65% was achieved. Significant differences in the performances of the 3 selected algorithms were apparent. Nonetheless, additional metric measurements, such as receiver operating characteristic, area under the receiver operating characteristic curve score, accuracy, and loss, suggested that the gradient boosting model performed the best in predicting clinical pregnancy. This study served as a stepping stone toward the application of in vitro fertilization prediction models that use machine learning techniques. However, additional validation steps are required to boost the model's performance for its implementation in the clinical setting.

An explainable deep learning-based algorithm with an attention mechanism for predicting the live birth potential of mouse embryos

In assisted reproductive technology (ART), embryos produced by in vitro fertilization (IVF) are graded according to their live birth potential, and high-grade embryos are preferentially transplanted. However, rates of live birth following clinical ART remain low worldwide. Grading is based on the embryo shape at a limited number of stages and does not consider the shape of embryos and intracellular structures, e.g., nuclei, at various stages important for normal embryogenesis. Here, we developed a Normalized Multi-View Attention Network (NVAN) that directly predicts live birth potential from the nuclear structure in live-cell fluorescence images of mouse embryos from zygote to across a wide range of stages. The input is morphological features of cell nuclei, which were extracted as multivariate time-series data by using the segmentation algorithm for mouse embryos. The classification accuracy of our method (83.87%) greatly exceeded that of existing machine-learning methods and that of visual inspection by embryo culture specialists. Our method also has a new attention mechanism that allows us to determine which values of multivariate time-series data, used to describe nuclear morphology, were the basis for the prediction. By visualizing the features that contributed most to the prediction of live birth potential, we found that the size and shape of the nucleus at the morula stage and at the time of cell division were important for live birth prediction. We anticipate that our method will help ART and developmental engineering as a new basic technology for IVF embryo selection.

Two‐step nuclear centring by competing microtubule‐ and actin‐based mechanisms in 2‐cell mouse embryos

Microtubules typically promote nuclear centring during early embryonic divisions in centrosome‐containing vertebrates. In acentrosomal mouse zygotes, microtubules also centre male and female pronuclei prior to the first mitosis, this time in concert with actin. How nuclear centring is brought about in subsequent acentrosomal embryonic divisions has not been studied. Here, using time‐lapse imaging in mouse embryos, we find that although nuclei are delivered to the cell centre upon completion of the first mitotic anaphase, the majority do not remain stationary and instead travel all the way to the cortex in a microtubule‐dependent manner. High cytoplasmic viscosity in 2‐cell embryos is associated with non‐diffusive mechanisms involving actin for subsequent nuclear centring when microtubules again exert a negative influence. Thus, following the first mitotic division, pro‐centring actin‐dependent mechanisms work against microtubule‐dependent de‐centring forces. Disrupting the equilibrium of this tug‐of‐war compromises nuclear centring and symmetry of the subsequent division potentially risking embryonic development. This circuitous centring process exposes an embryonic vulnerability imposed by microtubule‐dependent de‐centring forces.

Increasing the field-of-view in oblique plane microscopy via optical tiling.

Fast volumetric imaging of large fluorescent samples with high-resolution is required for many biological applications. Oblique plane microscopy (OPM) provides high spatiotemporal resolution, but the field of view is typically limited by its optical train and the pixel number of the camera. Mechanically scanning the sample or decreasing the overall magnification of the imaging system can partially address this challenge, albeit by reducing the volumetric imaging speed or spatial resolution, respectively. Here, we introduce a novel dual-axis scan unit for OPM that facilitates rapid and high-resolution volumetric imaging throughout a volume of 800 × 500 × 200 microns. This enables us to perform volumetric imaging of cell monolayers, spheroids and zebrafish embryos with subcellular resolution. Furthermore, we combined this microscope with a multi-perspective projection imaging technique that increases the volumetric interrogation rate to more than 10 Hz. This allows us to rapidly probe a large field of view in a dimensionality reduced format, identify features of interest, and volumetrically image these regions with high spatiotemporal resolution.

ARTIFICIAL INTELLIGENCE CAN BE USED AS A GUIDE TO SELECT WHICH BLASTOCYSTS TO BIOPSY AND TEST FOR PGT-A

PGT-A testing is time consuming and expensive. Recent AI models have been used to predict ploidy status using embryo images or time-lapse videos with area under the curve (AUC) between 0.62 and 0.75, indicating reasonable but not perfect predictions. AI models thus cannot replace PGT-A entirely, but may help select which blastocysts to test in order to achieve a desired number of euploids. In this study, we investigated if AI models could be used to prioritize and reduce the number of blastocysts requiring PGT-A testing to attain at least one euploid blastocyst per treatment.

Preprint Highlight: Dodecaploid Xenopus longipes provides insight into the emergence of size scaling relationships during development.

Although the relationship between genome and cell size is well defined in developed tissues, it remains unclear when in development nuclear-cytoplasmic scaling is established. Here the authors analyzed multiple Xenopus species with varying genome and cell sizes to investigate scaling relationships during development. Imaging of cells, nuclei, and embryos shows that in early development, nuclear size correlates with genome size across species and cell size is predicted by egg size. Mitotic spindle size during cell division was determined by cell size, not genome size. These findings indicate that nuclear-cytoplasmic scaling mechanisms have evolved to support development despite dramatic variations in genome size. This work also suggests that maternal components in the egg drive cell size scaling in early development.