Search

Polaritons are light-matter quasiparticles that govern the optical response of quantum materials at the nanoscale, enabling on-chip communication and local sensing. Here, we report Landau-phonon polaritons (LPPs) in magnetized charge-neutral graphene encapsulated in hexagonal boron nitride (hBN). These quasiparticles emerge from the interaction of Dirac magnetoexciton modes in graphene with the hyperbolic phonon polariton modes in hBN. Using infrared magneto-nanoscopy, we reveal the ability to completely halt the LPP propagation in real space at quantized magnetic fields, defying the conventional optical selection rules. The LPP-based nanoscopy also tells apart two fundamental many-body phenomena: the Fermi velocity renormalization and field-dependent magnetoexciton binding energies. Our results highlight the potential of magnetically tuned Dirac heterostructures for precise nanoscale control and sensing of light-matter interaction.

Abstract Medical datasets inherently contain errors from subjective or inaccurate test results, or from confounding biological complexities. It is difficult for medical experts to detect these elusive errors manually, due to lack of contextual information, limiting data privacy regulations, and the sheer scale of data to be reviewed. Current methods for training robust artificial intelligence (AI) models on data containing mislabeled examples generally fall into one of several categories—attempting to improve the robustness of the model architecture, the regularization techniques used, the loss function used during training, or selecting a subset of data that contains cleaner labels. This last category requires the ability to efficiently detect errors either prior to or during training, either relabeling them or removing them completely. More recent progress in error detection has focused on using multi-network learning to minimize deleterious effects of errors on training, however, using many neural networks to reach a consensus on which data should be removed can be computationally intensive and inefficient. In this work, a deep-learning based algorithm was used in conjunction with a label-clustering approach to automate error detection. For dataset with synthetic label flips added, these errors were identified with an accuracy of up to 85%, while requiring up to 93% less computing resources to complete compared to a previous model consensus approach developed previously. The resulting trained AI models exhibited greater training stability and up to a 45% improvement in accuracy, from 69 to over 99% compared to the consensus approach, at least 10% improvement on using noise-robust loss functions in a binary classification problem, and a 51% improvement for multi-class classification. These results indicate that practical, automated a priori detection of errors in medical data is possible, without human oversight.

Abstract Computer-Assisted Semen Analysis (CASA) enables reliable analysis of semen images, and is designed to process large number of images with high consistency, accuracy, and repeatability. Design and testing of CASA algorithms can be accelerated greatly if reliable simulations of semen images under a variety of conditions and sample quality modes are available. Using life-like simulation of semen images can quantify the performance of existing and proposed CASA algorithms, since the parameters of the simulated image are known and controllable. We present simulation models for sperm cell image and swimming modes observed in real 2D (top-down) images of sperm cells in laboratory specimen. The models simulate human sperm using four (4) types of swimming, namely linear mean, circular, hyperactive, and immotile (or dead). The simulation models are used in studying algorithms for segmentation, localization, and tracking of sperm cells. Several segmentation and localization algorithms were tested under varying levels of noise, and then compared using precision, recall, and the optimal subpattern assignment (OSPA) metric. Images of real human semen sample were used to validate the segmentation and localization observations obtained from simulations. An example is given of sperm cell tracking on simulated semen images of cells using the different tracking algorithms (nearest neighbor (NN), global nearest neighbor (GNN), probabilistic data association filter (PDAF), and joint probabilistic data association filter (JPDAF)). Tracking performance was evaluated through multi-object tracking precision (MOTP) and multi-object tracking accuracy (MOTA). Simulation models enable objective assessments of semen image processing algorithms. We demonstrate the use of a new simulation tool to assess and compare segmentation, localization, and tracking methods. The simulation software allows testing along a large spectrum of parameter values that control the appearance and behavior of simulated semen images. Users can generate scenarios of different characteristics and assess the effectiveness of different CASA algorithms in these environments. The simulation was used to assess and compare algorithms for segmentation and tracking of sperm cells in semen images.

Abstract Background Neuronal dysfunction and degeneration linked to α-synuclein (αS) pathology is thought to be responsible for the progressive nature of Parkinson’s disease and related dementia with Lewy bodies. Studies have indicated bidirectional pathological relationships between αS pathology and tau abnormalities. We recently showed that A53T mutant human αS (HuαS) can cause post-synaptic and cognitive deficits that require microtubule-associated protein tau expression. However, the role of tau in the development of αS pathology and subsequent neuronal dysfunction has been controversial. Herein, we set out to determine the role of tau in the onset and progression of αS pathology (α-synucleinopathy) using a transgenic mouse model of α-synucleinopathy lacking mouse tau expression. Methods Transgenic mice expressing A53T mutant HuαS (TgA53T) were crossed with mTau−/− mice to generate TgA53T/mTau−/−. To achieve more uniform induction of α-synucleinopathy in mice, we used intramuscular injections of αS preformed fibrils (PFF) in non-transgenic (nTg), TgA53T, TgA53T/mTau−/−, and mTau−/− mice. Motor behavior was analyzed at 70 days post inoculation (dpi) of PFF and tissues for biochemical and neuropathological analysis were collected at 40 dpi, 70 dpi, and end stage. Results Loss of tau expression significantly delayed the onset of motor deficits in the TgA53T model and the progression of α-synucleinopathy disease, as evidenced by a significant reduction in histopathological and behavioral markers of neurodegeneration and disease, and a significant improvement in survival. In vitro application of PFF to primary mouse hippocampal neurons demonstrated no changes in PFF uptake and processing or pS129 αS aggregation as a function of tau expression. However, PFF-induced neurotoxicity, including morphological deficits in nTg neurons, was prevented with tau removal. Conclusions Collectively, our data suggest that tau is likely acting downstream of αS pathology to affect neuronal homeostasis and survival. This work further supports the investigation of tau in α-synucleinopathies to identify novel disease-modifying therapeutic strategies.

Abstract Training on multiple diverse data sources is critical to ensure unbiased and generalizable AI. In healthcare, data privacy laws prohibit data from being moved outside the country of origin, preventing global medical datasets being centralized for AI training. Data-centric, cross-silo federated learning represents a pathway forward for training on distributed medical datasets. Existing approaches typically require updates to a training model to be transferred to a central server, potentially breaching data privacy laws unless the updates are sufficiently disguised or abstracted to prevent reconstruction of the dataset. Here we present a completely decentralized federated learning approach, using knowledge distillation, ensuring data privacy and protection. Each node operates independently without needing to access external data. AI accuracy using this approach is found to be comparable to centralized training, and when nodes comprise poor-quality data, which is common in healthcare, AI accuracy can exceed the performance of traditional centralized training.

Background and Aims: Artificial intelligence (AI) is being increasingly used for non-invasive evaluation of embryo quality during IVF. Previous studies described development of AI for selecting embryos likely to be euploid (genetics AI), or likely to lead to clinical pregnancy (viability AI), based on analysis of images of blastocysts on day 5 of development. The aim of this study was to determine if a combination of these AI scores could be used to effectively evaluate both outcomes. Method: 936 embryo images with pre-implantation genetic testing for aneuploidies (PGT-A) outcomes, and 479 embryo images with clinical pregnancy outcomes, were retrospectively obtained from 12 IVF clinics in 5 countries. Performance was evaluated for each AI score alone, and the average score of the two AIs. The ability to select euploid or viable embryos was evaluated using ROC-AUC analyses, and a simulated cohort ranking method reported in the literature. Results: The average score of the two AIs was generally as effective at selecting euploid embryos as the genetics AI, and just as effective at selecting viable embryos as the viability AI. Results for both analyses are presented below. Conclusion: An AI score that can evaluate both embryo ploidy and viability simultaneously is useful for selecting preferred embryos for analysis or transfer. These results suggest that it is feasible to generate a single score for evaluating overall embryo quality using a non-invasive approach.

Background and Aims: Embryologist evaluation of embryos is critical for ensuring successful pregnancy outcomes. Standard, manual evaluation is variable, subjective, and time-consuming. The aim of this study was to evaluate whether an artificial intelligence (AI) algorithm can standardize and improve embryo evaluation during IVF. Method: 20 images of blastocyst-stage embryos on day 5 of in vitro development were selected to represent a range of morphological qualities. All embryos had been transferred and the clinical pregnancy outcome was known for each embryo based on detection of fetal heartbeat at first ultrasound scan (∼7-9 weeks gestation). 50% of embryos in the dataset resulted in pregnancy. 158 embryologists made a total of 236 attempts at providing their evaluation of the morphological quality of the 20 embryo images using the Gardner system. The embryologist-assigned grades were then used to generate their prediction of whether that embryo would lead to pregnancy or not (≥ 3BB indicated a pregnancy prediction, and

Background and Aims: Oocyte vitrification and warming is very time consuming. Oocyte vitrification protocols are different from that of embryos or blastocysts, while warming protocols allow all to be warmed the same. For blastocysts, frequently only one device is warmed per patient; however, numerous devices are required in oocyte warming. We aim to evaluate whether vitrified oocytes can successfully survive a warming protocol involving only Thawing Solution (TS) with a total elapsed warming time of a mere 90 seconds. Methods: 58 vitrified donor oocytes (Treatment Group) were prospectively warmed in January 2023. 71 vitrified donor oocytes (Control Group) warmed in 2022 were retrospectively analyzed. Both groups were vitrified in Kitazato or Irvine Scientific Vit Kit with identical protocols (elapsed time approximately 17-minutes). In both groups the TS was equilibrated to 37-degrees for a minimum of 2-hours. The Control Group was warmed in Irvine Scientific Vit Kit (1-minute TS, 3-minute DS, 8-minute WS) for a total warming time of 12-minutes. Warming for the Treatment Group was in Irvine Scientific Vit Kit Warm-NX, TS only, for 90-seconds. Results: After Chi-square analysis, [Formula: see text] value < 0.05 reflected significance. The Control Group warmed with 90.1% survival. The Treatment Group resulted in 98.3% survival, ([Formula: see text] -value = 0.056). Survival at 3-hours post warm in the Control Group was 100%. At this point for the Control Group, ICSI was performed, and the oocytes placed back into the incubator for resulting embryo culture. For the Treatment Group, oocyte survival was re-assessed at 1-hour and 24-hours, and 100% survival was maintained over the course of 24-hour incubation, with little-to-no visible cryo-damage. Conclusions: Vitrified oocytes exhibit outstanding survival when warmed with a highly modified and considerably faster protocol than current practice. The extended oocyte survival of at least 24-hours shows a very favorable outcome for the accelerated warming protocol and gives credibility to the survival assessment.

Summary: At the moment of union in fertilization, sperm and oocyte are transcriptionally silent. The ensuing onset of embryonic transcription (embryonic genome activation [EGA]) is critical for development, yet its timing and profile remain elusive in any vertebrate species. We here dissect transcription during EGA by high-resolution single-cell RNA sequencing of precisely synchronized mouse one-cell embryos. This reveals a program of embryonic gene expression (immediate EGA [iEGA]) initiating within 4 h of fertilization. Expression during iEGA produces canonically spliced transcripts, occurs substantially from the maternal genome, and is mostly downregulated at the two-cell stage. Transcribed genes predict regulation by transcription factors (TFs) associated with cancer, including c-Myc. Blocking c-Myc or other predicted regulatory TF activities disrupts iEGA and induces acute developmental arrest. These findings illuminate intracellular mechanisms that regulate the onset of mammalian development and hold promise for the study of cancer.

Abstract The detection and removal of poor-quality data in a training set is crucial to achieve high-performing AI models. In healthcare, data can be inherently poor-quality due to uncertainty or subjectivity, but as is often the case, the requirement for data privacy restricts AI practitioners from accessing raw training data, meaning manual visual verification of private patient data is not possible. Here we describe a novel method for automated identification of poor-quality data, called Untrainable Data Cleansing. This method is shown to have numerous benefits including protection of private patient data; improvement in AI generalizability; reduction in time, cost, and data needed for training; all while offering a truer reporting of AI performance itself. Additionally, results show that Untrainable Data Cleansing could be useful as a triage tool to identify difficult clinical cases that may warrant in-depth evaluation or additional testing to support a diagnosis.

A prospective, multicenter, randomized, sibling oocyte study was conducted with 86 couples to evaluate if a microfluidic sperm separation device improved ICSI sperm selection and subsequent cycle outcomes of fertilization, blastocyst utilization, ploidy, and clinical pregnancy rate when applied to a general patient population. Patients with at least 10 metaphase II oocytes were enrolled in the study and sibling oocyte groups were split in half. One half of the oocytes underwent ICSI with the control processed sperm and the other half were injected with sperm sorted by the ZyMōt microfluidic sperm separation device. Fertilization rate was recorded and resulting blastocysts were biopsied and evaluated for ploidy status with NGS. Euploid, non-mosaic embryos were randomly selected for single embryo transfer. A total of 787 oocytes were evaluated in the ZyMōt group and 777 in the control group. No statistical differences were observed between ZyMōt and control processing methods in any of the study outcomes evaluated. It is possible that the selection of normal, progressive sperm for ICSI, and the repair capacity of oocytes are sufficient to promote normal embryonic development in the general infertility population.

Abstract Purpose The aim of this study was to examine whether the translocator protein 18-kDa (TSPO) PET ligand [18F]FEPPA has the sensitivity for detecting changes in CD68-positive microglial/macrophage activation in hemiparkinsonian rhesus macaques treated with allogeneic grafts of induced pluripotent stem cell-derived midbrain dopaminergic neurons (iPSC-mDA). Methods In vivo positron emission tomography (PET) imaging with [18F]FEPPA was used in conjunction with postmortem CD68 immunostaining to evaluate neuroinflammation in the brains of hemiparkinsonian rhesus macaques (n = 6) that received allogeneic iPSC-mDA grafts in the putamen ipsilateral to MPTP administration. Results Based on assessment of radiotracer uptake and confirmed by visual inspection of the imaging data, nonhuman primates with allogeneic grafts showed increased [18F]FEPPA binding at the graft sites relative to the contralateral putamen. From PET asymmetry analysis of the images, the mean asymmetry index of the monkeys was AI = − 0.085 ± 0.018. Evaluation and scoring of CD68 immunoreactivity by an investigator blind to the treatment identified significantly more neuroinflammation in the grafted areas of the putamen compared to the contralateral putamen (p = 0.0004). [18F]FEPPA PET AI showed a positive correlation with CD68 immunoreactivity AI ratings in the monkeys (Spearman’s ρ = 0.94; p = 0.005). Conclusion These findings reveal that [18F]FEPPA PET is an effective marker for detecting increased CD68-positive microglial/macrophage activation and demonstrates sufficient sensitivity to detect changes in neuroinflammation in vivo following allogeneic cell engraftment.