Underestimation of Thermogenic Methane Emissions in New York City.

Recent studies have shown that methane emissions are underestimated by inventories in many US urban areas. This has important implications for climate change mitigation policy at the city, state, and national levels. Uncertainty in both the spatial distribution and sectoral allocation of urban emissions can limit the ability of policy makers to develop appropriately focused emission reduction strategies. Top-down emission estimates based on atmospheric greenhouse gas measurements can help to improve inventories and inform policy decisions. This study presents a new high-resolution (0.02 × 0.02°) methane emission inventory for New York City and its surrounding area, constructed using the latest activity data, emission factors, and spatial proxies. The new high-resolution inventory estimates of methane emissions for the New York-Newark urban area are 1.3 times larger than those for the gridded Environmental Protection Agency inventory. We used aircraft mole fraction measurements from nine research flights to optimize the high-resolution inventory emissions within a Bayesian inversion. These sectorally optimized emissions show that the high-resolution inventory still significantly underestimates methane emissions within the New York-Newark urban area, primarily because it underestimates emissions from thermogenic sources (by a factor of 2.3). This suggests that there remains a gap in our process-based understanding of urban methane emissions.

A model of pregnancy-associated malaria for inducing adverse pregnancy outcomes in ICR mouse.

BACKGROUND: Based on understanding of placental pathological features and safe medication in pregnancy-associated malaria (PAM), establishment of a stable pregnant mouse infection model with Plasmodium was urgently needed. METHODS: ICR mice with vaginal plugs detected were randomly divided into post-pregnancy infection (Malaria+) and uninfected pregnancy (Malaria-) cohorts. Age-matched mice that had not been mated were infected as pre-pregnancy infection group (Virgin control), which were subsequently mated with ICR males. All mice were inoculated with 1 × 106Plasmodium berghei ANKA-infected RBCs by intraperitoneal injection, and the same amount of saline was given to Malaria- group. We recorded the incidence of adverse pregnancy outcomes and the amounts of offspring in each group. RESULTS: The Virgin group mice were unable to conceive normally, and vaginal bleeding, abortion, or stillbirth appeared in the Malaria+ group. The incidence of adverse pregnancy outcomes was extremely high and statistically significant compared with the control (Malaria-) group (P < 0.05), of which placenta exhibited pathological features associated with human gestational malaria. CONCLUSIONS: The intraperitoneal injection of 1 × 106Plasmodium berghei ANKA-infected RBCs could establish a model of pregnancy-associated malaria in ICR mouse.

Erythrocyte membrane with CLIPPKF as biomimetic nanodecoy traps merozoites and attaches to infected red blood cells to prevent Plasmodium infection.

BACKGROUND: Malaria remains a serious threat to global public health. With poor efficacies of vaccines and the emergence of drug resistance, novel strategies to control malaria are urgently needed. RESULTS: We developed erythrocyte membrane-camouflaged nanoparticles loaded with artemether based on the growth characteristics of Plasmodium. The nanoparticles could capture the merozoites to inhibit them from repeatedly infecting normal erythrocytes, owing to the interactions between merozoites and heparin-like molecules on the erythrocyte membrane. Modification with a phosphatidylserine-targeting peptide (CLIPPKF) improved the drug accumulation in infected red blood cells (iRBCs) from the externalized phosphatidylserine induced by Plasmodium infection. In Plasmodium berghei ANKA strain (pbANKA)-infected C57BL/6 mice, the nanoparticles significantly attenuated Plasmodium-induced inflammation, apoptosis, and anemia. We observed reduced weight variation and prolonged survival time in pbANKA-challenged mice, and the nanoparticles showed good biocompatibility and negligible cytotoxicity. CONCLUSION: Erythrocyte membrane-camouflaged nanoparticles loaded with artemether were shown to provide safe and effective protection against Plasmodium infection.

A drug-free nanozyme for mitigating oxidative stress and inflammatory bowel disease.

Inflammatory bowel disease (IBD) is an incurable disease of the gastrointestinal tract with a lack of effective therapeutic strategies. The proinflammatory microenvironment plays a significant role in both amplifying and sustaining inflammation during IBD progression. Herein, biocompatible drug-free ceria nanoparticles (CeNP-PEG) with regenerable scavenging activities against multiple reactive oxygen species (ROS) were developed. CeNP-PEG exerted therapeutic effect in dextran sulfate sodium (DSS)-induced colitis murine model, evidenced by corrected the disease activity index, restrained colon length shortening, improved intestinal permeability and restored the colonic epithelium disruption. CeNP-PEG ameliorated the proinflammatory microenvironment by persistently scavenging ROS, down-regulating the levels of multiple proinflammatory cytokines, restraining the proinflammatory profile of macrophages and Th1/Th17 response. The underlying mechanism may involve restraining the co-activation of NF-κB and JAK2/STAT3 pathways. In summary, this work demonstrates an effective strategy for IBD treatment by ameliorating the self-perpetuating proinflammatory microenvironment, which offers a new avenue in the treatment of inflammation-related diseases.

Toward Accurate, Policy-Relevant Fossil Fuel CO2 Emission Landscapes.

The bottom-up (BU) approach has been used to develop spatiotemporally resolved, sectorally disaggregated fossil fuel CO2 (FFCO2) emission data products. These efforts are critical constraints to atmospheric assessment of anthropogenic fluxes in addition to offering the climate change policymaking community usable information to guide mitigation. In the United States, there are two high-resolution FFCO2 emission data products, Vulcan and the Anthropogenic Carbon Emissions System (ACES). As a step toward developing improved, accurate, and detailed FFCO2 emission landscapes, we perform a comparison of the two data products. We find that while agreeing on total FFCO2 emissions at the aggregate scale (relative difference = 1.7%), larger differences occur at smaller spatial scales and in individual sectors. Differences in the smaller-emitting sectors are likely errors in ACES input data or emission factors. ACES advances the approach for estimating emissions in the gas and oil sector, while Vulcan shows better geocoordinate correction in the electricity production sector. Differences in the subcounty residential and commercial building sectors are driven by different spatial proxies and suggest a task for future investigation. The gridcell absolute median relative difference, a measure of the average gridcell-scale relative difference, indicates a 53.5% difference. The recommendation for improved BU granular FFCO2 emission estimation includes review, assessment, and archive of point source geolocations, CO emission input data, CO and CO2 emission factors, and uncertainty approaches including those due to spatial errors. Finally, intensives where local utility data are publicly available could test the spatial proxies used in estimating residential and commercial building emissions. These steps toward best practices will lead to more accurate, granular emissions, enabling optimal emission mitigation policy choices.

Synthesis of Urban CO2 Emission Estimates from Multiple Methods from the Indianapolis Flux Project (INFLUX).

Urban areas contribute approximately three-quarters of fossil fuel derived CO2 emissions, and many cities have enacted emissions mitigation plans. Evaluation of the effectiveness of mitigation efforts will require measurement of both the emission rate and its change over space and time. The relative performance of different emission estimation methods is a critical requirement to support mitigation efforts. Here we compare results of CO2 emissions estimation methods including an inventory-based method and two different top-down atmospheric measurement approaches implemented for the Indianapolis, Indiana, U.S.A. urban area in winter. By accounting for differences in spatial and temporal coverage, as well as trace gas species measured, we find agreement among the wintertime whole-city fossil fuel CO2 emission rate estimates to within 7%. This finding represents a major improvement over previous comparisons of urban-scale emissions, making urban CO2 flux estimates from this study consistent with local and global emission mitigation strategy needs. The complementary application of multiple scientifically driven emissions quantification methods enables and establishes this high level of confidence and demonstrates the strength of the joint implementation of rigorous inventory and atmospheric emissions monitoring approaches.

Integrating Active Learning and Transfer Learning for Carotid Intima-Media Thickness Video Interpretation.

Cardiovascular disease (CVD) is the number one killer in the USA, yet it is largely preventable (World Health Organization 2011). To prevent CVD, carotid intima-media thickness (CIMT) imaging, a noninvasive ultrasonography method, has proven to be clinically valuable in identifying at-risk persons before adverse events. Researchers are developing systems to automate CIMT video interpretation based on deep learning, but such efforts are impeded by the lack of large annotated CIMT video datasets. CIMT video annotation is not only tedious, laborious, and time consuming, but also demanding of costly, specialty-oriented knowledge and skills, which are not easily accessible. To dramatically reduce the cost of CIMT video annotation, this paper makes three main contributions. Our first contribution is a new concept, called Annotation Unit (AU), which simplifies the entire CIMT video annotation process down to six simple mouse clicks. Our second contribution is a new algorithm, called AFT (active fine-tuning), which naturally integrates active learning and transfer learning (fine-tuning) into a single framework. AFT starts directly with a pre-trained convolutional neural network (CNN), focuses on selecting the most informative and representative AU s from the unannotated pool for annotation, and then fine-tunes the CNN by incorporating newly annotated AU s in each iteration to enhance the CNN's performance gradually. Our third contribution is a systematic evaluation, which shows that, in comparison with the state-of-the-art method (Tajbakhsh et al., IEEE Trans Med Imaging 35(5):1299-1312, 2016), our method can cut the annotation cost by >81% relative to their training from scratch and >50% relative to their random selection. This performance is attributed to the several advantages derived from the advanced active, continuous learning capability of our AFT method.

Synergistic Effects of Salvianolic Acid B and Puerarin on Cerebral Ischemia Reperfusion Injury.

Ischemic stroke (IS) is characterized by the sudden loss of blood circulation to an area of the brain, resulting in a corresponding loss of neurologic function. It has been a worldwide critical disease threatening to the health and life of human beings. Despite significant progresses achieved, effective treatment still remains a formidable challenge due to the complexity of the disease. Salvianolic acid B (Sal-B) and Puerarin (Pue) are two active neuroprotectants isolated from traditional Chinese herbs, Salvia miltiorrhiza and Kudzu root respectively, which have been used for the prevention and treatment of IS for thousands of years in China. The activities of two compounds against cerebral ischemia reperfusion injury have been confirmed via various pathways. However, the therapeutic efficacy of any of the two components is still unsatisfied. In the present study, the effect of the combination of Sal-B and Pue on IS was evaluated and validated in vitro and in vivo. The ratio of two compounds was firstly optimized based on the results of CoCl2 damaged PC12 cells model. The co-administration exhibited significantly protective effect in CoCl2 induced PC12 cells injury model by reducing ROS, inhibiting apoptosis and improving mitochondrial membrane potential in vitro. Moreover, Sal-B + Pue significantly relieved neurological deficit scores and infarct area than Sal-B or Pue alone in vivo. The results indicated that neuroprotection mechanism of Sal-B + Pue was related to TLR4/MyD88 and SIRT1 activation signaling pathway to achieve synergistic effect, due to the inhibition of NF-κB transcriptional activity and expression of pro-inflammatory cytokine (TNF-α, IL-1ß, IL-6). In conclusion, the combination of Sal-B and Pue exerted much stronger neuroprotective effect than Sal-B or Pue alone, which provides a potential new drug and has great significance for the treatment of IS.

Detection of Multiple Parallel Transmission Outbreak of Streptococcus suis Human Infection by Use of Genome Epidemiology, China, 2005.

Streptococcus suis sequence type 7 emerged and caused 2 of the largest human infection outbreaks in China in 1998 and 2005. To determine the major risk factors and source of the infections, we analyzed whole genomes of 95 outbreak-associated isolates, identified 160 single nucleotide polymorphisms, and classified them into 6 clades. Molecular clock analysis revealed that clade 1 (responsible for the 1998 outbreak) emerged in October 1997. Clades 2-6 (responsible for the 2005 outbreak) emerged separately during February 2002-August 2004. A total of 41 lineages of S. suis emerged by the end of 2004 and rapidly expanded to 68 genome types through single base mutations when the outbreak occurred in June 2005. We identified 32 identical isolates and classified them into 8 groups, which were distributed in a large geographic area with no transmission link. These findings suggest that persons were infected in parallel in respective geographic sites.

Study on the Clinical Significance of JAK2V617F Allele Burden in Philadelphia Chromosome-Negative Myeloproliferative Neoplasm.

BACKGROUND: It was discovered that the somatic mutation in JAK2 exon 14 (JAK2V617F) totally modified the understanding and diagnosis of Philadelphia-Negative myeloproliferative neoplasm (Ph-MPNs), including polycythemia vera (PV), essential thrombocythemia (ET), and primary myelofibrosis (PMF). Real-time quantitative PCR is the most widely used method for JAK2V617F detection in clinical laboratory. In this study, we aimed to evaluate the clinical significance of JAK2V617F allele burden in Ph-MPNs detected by real-time quantitative PCR. METHODS: A total of 208 bone marrow samples were collected from patients suspected to have Ph-MPNs. Real-time quantitative PCR was performed on each sample to obtain the JAK2V617F allele burden. Clinical and laboratory data from these participants were also recorded for their first visit. RESULTS: Out of 208 participants, 118 patients were confirmed with Ph-MPNs. JAK2V617F mutations were found in 59 patients in the PV group (86.8%), 31 patients in the ET group (70.5%). PV, PMF, and ET showed a significant difference in the distribution of JAK2V617F allele burden. In JAK2V617F positive patients, JAK2V617F allele burden was closely related with WBC counts, platelet counts, and hemoglobin concentration. CONCLUSIONS: JAK2V617F allele burden is a useful marker in the diagnosis, discrimination, and evaluation of PhMPNs.

Cell-penetrating apoptotic peptide/p53 DNA nanocomplex as adjuvant therapy for drug-resistant breast cancer.

Drug resistance becomes a formidable challenge against effective cancer therapy. Defective apoptosis in cancer cells is a key factor responsible for chemoresistance or radioresistance. Promoting apoptosis is an important method to sensitize the resistant cells, thereby achieving successful treatment for MDR cancer. We present a strategy of codelivery of apoptotic AVPI peptide and p53 DNA as apoptosis-induction adjuvant therapy for combating the resistant breast cancer. AVPI tetrapeptide is poorly cell-permeable, thereby with very limited value for therapeutic use. Cell-penetrating chimeric AVPI derivative was developed by modification with an octa-arginine sequence (R8). The AVPIR8 is able to not only efficiently penetrate into tumor cells but also work as a vector for gene delivery by forming nanocomplexes based on its cationic R8 moiety. The combination of AVPIR8/p53 DNA was selected for targeting apoptotic pathways, thereby sensitizing the cancer cells to chemotherapeutics. The anti-MDR effect was demonstrated both in vitro and in vivo. The synergistic use of AVPIR8/p53 significantly increased the sensitivity of the resistant tumor cells to the cytotoxic agent doxorubicin by inducing apoptosis, as demonstrated in the cellular studies. Importantly, the treatment improvement was also observed in the animal studies with resistant breast tumor model. Coadministration of AVPIR8/p53 enabled a full arrest of tumor growth combined with a reduced DOX dose, yielding a productive and safe cancer treatment.

[Cell-penetrating chimeric apoptotic peptide AVPI-LMWP/DNA co-delivery system for cancer therapy].

To develop a cell-penetrating chimeric apoptotic peptide AVPI-LMWP/DNA co-delivery system for cancer therapy, we prepared the AVPI-LMWP/pTRAIL self-assembled complexes containing a therapeutic combination of peptide drug AVPI and DNA drug TRAIL. The chimeric apoptotic peptide AVPI-LMWP was synthesized using the standard solid-phase synthesis. The cationic AVPI-LMWP could condense pTRAIL by electrostatic interaction. The physical-chemical properties of the AVPI-LMWP/pTRAIL complexes were characterized. The cellular uptake efficiency and the inhibitory activity of the AVPI-LMWP/pTRAIL complexes on tumor cell were also performed. The results showed that the AVPI-LMWP/pTRAIL complexes were successfully prepared by co-incubation. With the increase of mass ratio (AVPI-LMWP/DNA), the particle size was decreased and the zeta potential had few change. Agarose gel electrophoresis showed that AVPI-LMWP could fully bind and condense pTRAIL at a mass ratio above 15:1. Cellular uptake efficiency was improved along with the increased ratio of W(AVPI-LMWP)/WpTRAIL. The in vitro cytotoxicity experiments demonstrated that the AVPI-LMWP/pTRAIL (W:W = 20:1) complexes was significantly more effective than the pTRAIL, AVPI-LMWP alone or LMWP/pTRAIL complexes on inhibition of HeLa cell growth. Our studies indicated that the AVPI-LMWP/pTRAIL co-delivery system could deliver plasmid into HeLa cell and induce tumor cell apoptosis efficiently, which showed its potential in cancer therapy using combination of apoptoic peptide and gene drugs.

Towards Foundation Models Learned from Anatomy in Medical Imaging via Self-supervision.

Human anatomy is the foundation of medical imaging and boasts one striking characteristic: its hierarchy in nature, exhibiting two intrinsic properties: (1) locality: each anatomical structure is morphologically distinct from the others; and (2) compositionality: each anatomical structure is an integrated part of a larger whole. We envision a foundation model for medical imaging that is consciously and purposefully developed upon this foundation to gain the capability of "understanding" human anatomy and to possess the fundamental properties of medical imaging. As our first step in realizing this vision towards foundation models in medical imaging, we devise a novel self-supervised learning (SSL) strategy that exploits the hierarchical nature of human anatomy. Our extensive experiments demonstrate that the SSL pretrained model, derived from our training strategy, not only outperforms state-of-the-art (SOTA) fully/self-supervised baselines but also enhances annotation efficiency, offering potential few-shot segmentation capabilities with performance improvements ranging from 9% to 30% for segmentation tasks compared to SSL baselines. This performance is attributed to the significance of anatomy comprehension via our learning strategy, which encapsulates the intrinsic attributes of anatomical structures-locality and compositionality-within the embedding space, yet overlooked in existing SSL methods. All code and pretrained models are available at GitHub.com/JLiangLab/Eden.

Stepwise incremental pretraining for integrating discriminative, restorative, and adversarial learning.

We have developed a United framework that integrates three self-supervised learning (SSL) ingredients (discriminative, restorative, and adversarial learning), enabling collaborative learning among the three learning ingredients and yielding three transferable components: a discriminative encoder, a restorative decoder, and an adversary encoder. To leverage this collaboration, we redesigned nine prominent self-supervised methods, including Rotation, Jigsaw, Rubik's Cube, Deep Clustering, TransVW, MoCo, BYOL, PCRL, and Swin UNETR, and augmented each with its missing components in a United framework for 3D medical imaging. However, such a United framework increases model complexity, making 3D pretraining difficult. To overcome this difficulty, we propose stepwise incremental pretraining, a strategy that unifies the pretraining, in which a discriminative encoder is first trained via discriminative learning, the pretrained discriminative encoder is then attached to a restorative decoder, forming a skip-connected encoder-decoder, for further joint discriminative and restorative learning. Last, the pretrained encoder-decoder is associated with an adversarial encoder for final full discriminative, restorative, and adversarial learning. Our extensive experiments demonstrate that the stepwise incremental pretraining stabilizes United models pretraining, resulting in significant performance gains and annotation cost reduction via transfer learning in six target tasks, ranging from classification to segmentation, across diseases, organs, datasets, and modalities. This performance improvement is attributed to the synergy of the three SSL ingredients in our United framework unleashed through stepwise incremental pretraining. Our codes and pretrained models are available at GitHub.com/JLiangLab/StepwisePretraining.

Seeking an optimal approach for Computer-aided Diagnosis of Pulmonary Embolism.

Pulmonary Embolism (PE) represents a thrombus ("blood clot"), usually originating from a lower extremity vein, that travels to the blood vessels in the lung, causing vascular obstruction and in some patients death. This disorder is commonly diagnosed using Computed Tomography Pulmonary Angiography (CTPA). Deep learning holds great promise for the Computer-aided Diagnosis (CAD) of PE. However, numerous deep learning methods, such as Convolutional Neural Networks (CNN) and Transformer-based models, exist for a given task, causing great confusion regarding the development of CAD systems for PE. To address this confusion, we present a comprehensive analysis of competing deep learning methods applicable to PE diagnosis based on four datasets. First, we use the RSNA PE dataset, which includes (weak) slice-level and exam-level labels, for PE classification and diagnosis, respectively. At the slice level, we compare CNNs with the Vision Transformer (ViT) and the Swin Transformer. We also investigate the impact of self-supervised versus (fully) supervised ImageNet pre-training, and transfer learning over training models from scratch. Additionally, at the exam level, we compare sequence model learning with our proposed transformer-based architecture, Embedding-based ViT (E-ViT). For the second and third datasets, we utilize the CAD-PE Challenge Dataset and Ferdowsi University of Mashad's PE Dataset, where we convert (strong) clot-level masks into slice-level annotations to evaluate the optimal CNN model for slice-level PE classification. Finally, we use our in-house PE-CAD dataset, which contains (strong) clot-level masks. Here, we investigate the impact of our vessel-oriented image representations and self-supervised pre-training on PE false positive reduction at the clot level across image dimensions (2D, 2.5D, and 3D). Our experiments show that (1) transfer learning boosts performance despite differences between photographic images and CTPA scans; (2) self-supervised pre-training can surpass (fully) supervised pre-training; (3) transformer-based models demonstrate comparable performance but slower convergence compared with CNNs for slice-level PE classification; (4) model trained on the RSNA PE dataset demonstrates promising performance when tested on unseen datasets for slice-level PE classification; (5) our E-ViT framework excels in handling variable numbers of slices and outperforms sequence model learning for exam-level diagnosis; and (6) vessel-oriented image representation and self-supervised pre-training both enhance performance for PE false positive reduction across image dimensions. Our optimal approach surpasses state-of-the-art results on the RSNA PE dataset, enhancing AUC by 0.62% (slice-level) and 2.22% (exam-level). On our in-house PE-CAD dataset, 3D vessel-oriented images improve performance from 80.07% to 91.35%, a remarkable 11% gain. Codes are available at GitHub.com/JLiangLab/CAD_PE.

Self-supervised learning for medical image analysis: Discriminative, restorative, or adversarial?

Discriminative, restorative, and adversarial learning have proven beneficial for self-supervised learning schemes in computer vision and medical imaging. Existing efforts, however, fail to capitalize on the potentially synergistic effects these methods may offer in a ternary setup, which, we envision can significantly benefit deep semantic representation learning. Towards this end, we developed DiRA, the first framework that unites discriminative, restorative, and adversarial learning in a unified manner to collaboratively glean complementary visual information from unlabeled medical images for fine-grained semantic representation learning. Our extensive experiments demonstrate that DiRA: (1) encourages collaborative learning among three learning ingredients, resulting in more generalizable representation across organs, diseases, and modalities; (2) outperforms fully supervised ImageNet models and increases robustness in small data regimes, reducing annotation cost across multiple medical imaging applications; (3) learns fine-grained semantic representation, facilitating accurate lesion localization with only image-level annotation; (4) improves reusability of low/mid-level features; and (5) enhances restorative self-supervised approaches, revealing that DiRA is a general framework for united representation learning. Code and pretrained models are available at https://github.com/JLiangLab/DiRA.

Ginsenoside Rh2-Based Multifunctional Liposomes for Advanced Breast Cancer Therapy.

Background: Most solid tumors are not diagnosed and treated until the advanced stage, in which tumors have shaped mature self-protective power, leading to off-target drugs and nanomedicines. In the present studies, we established a more realistic large tumor model to test the antitumor activity of a multifunctional ginsenoside Rh2-based liposome system (Rh2-lipo) on advanced breast cancer. Methods: Both cholesterol and PEG were substituted by Rh2 to prepare the Rh2-lipo using ethanol-water system and characterized. The effects of Rh2-lipo on cell uptake, penetration of the tumor spheroid, cytotoxicity assay was investigated with 4T1 breast cancer cells and L929 fibroblast cells. The 4T1 orthotopic-bearing large tumor model was established to study the targeting effect of Rh2-lipo and inhibitory effect of paclitaxel loaded Rh2-lipo (PTX-Rh2-lipo) on advanced breast tumors. Results: Rh2-lipo exhibit many advantages that address the limitations of current liposome formulations against large tumors, such as enhanced uptake in TAFs and tumor cells, high targeting and penetration capacity, cytotoxicity against TAFs, normalization of the vessel network, and depletion of stromal collagen. In in vivo study, PTX-Rh2-lipo effectively inhibiting the growth of advanced breast tumors and outperformed most reported PTX formulations, including Lipusu® and Abraxane®. Conclusion: Rh2-lipo have improved drug delivery efficiency and antitumor efficacy in advanced breast cancer, which offers a novel promising platform for advanced tumor therapy.

Intrauterine insulin resistance in fetuses of overweight mothers.

AIM: To investigate the relationship between maternal overweight and fetal insulin resistance. MATERIAL AND METHODS: Nineteen overweight and 30 lean pregnant women were recruited in the present study. Maternal and fetal insulin resistance were determined by measuring sex hormone binding globulin (SHBG) concentrations in maternal venous or umbilical cord serum, respectively. Maternal age, gestational age, height, pre-gravidity weight, pre-partum weight, as well as fetal gender, birth weight, birth height, and head circumference were collected as clinical data. RESULTS: Fetuses of overweight mothers had larger birth weight (3.58±0.55kg vs 3.32±0.42, adjusted P=0.006) and lower SHBG concentrations (26.64±3.65 vs 34.36±7.84, adjusted P=0.007) than those of lean mothers after values were adjusted for potential cofactors. Fetal SHBG level was negatively correlated with pre-gravidity body mass index (R=-0.392, adjusted P=0.025) and weight gain during pregnancy (R=-0.332, adjusted P=0.026) even with adjustment for potential cofactors. Among the 29 pregnant women with gestational diabetes mellitus, the overweight mothers had higher H1AC levels than their lean counterparts (6.47±0.44 vs 5.74±0.52, adjusted P=0.004). CONCLUSION: Intrauterine insulin resistance is more prominent in fetuses of overweight mothers, an effect that is decreased by weight gain control during pregnancy.

Foundation Ark: Accruing and Reusing Knowledge for Superior and Robust Performance.

Deep learning nowadays offers expert-level and sometimes even super-expert-level performance, but achieving such performance demands massive annotated data for training (e.g., Google's proprietary CXR Foundation Model (CXR-FM) was trained on 821,544 labeled and mostly private chest X-rays (CXRs)). Numerous datasets are publicly available in medical imaging but individually small and heterogeneous in expert labels. We envision a powerful and robust foundation model that can be trained by aggregating numerous small public datasets. To realize this vision, we have developed Ark, a framework that accrues and reuses knowledge from heterogeneous expert annotations in various datasets. As a proof of concept, we have trained two Ark models on 335,484 and 704,363 CXRs, respectively, by merging several datasets including ChestX-ray14, CheXpert, MIMIC-II, and VinDr-CXR, evaluated them on a wide range of imaging tasks covering both classification and segmentation via fine-tuning, linear-probing, and gender-bias analysis, and demonstrated our Ark's superior and robust performance over the state-of-the-art (SOTA) fully/self-supervised baselines and Google's proprietary CXR-FM. This enhanced performance is attributed to our simple yet powerful observation that aggregating numerous public datasets diversifies patient populations and accrues knowledge from diverse experts, yielding unprecedented performance yet saving annotation cost. With all codes and pretrained models released at GitHub.com/JLiangLab/Ark, we hope that Ark exerts an important impact on open science, as accruing and reusing knowledge from expert annotations in public datasets can potentially surpass the performance of proprietary models trained on unusually large data, inspiring many more researchers worldwide to share codes and datasets to build open foundation models, accelerate open science, and democratize deep learning for medical imaging.

Learning Anatomically Consistent Embedding for Chest Radiography.

Self-supervised learning (SSL) approaches have recently shown substantial success in learning visual representations from unannotated images. Compared with photographic images, medical images acquired with the same imaging protocol exhibit high consistency in anatomy. To exploit this anatomical consistency, this paper introduces a novel SSL approach, called PEAC (patch embedding of anatomical consistency), for medical image analysis. Specifically, in this paper, we propose to learn global and local consistencies via stable grid-based matching, transfer pre-trained PEAC models to diverse downstream tasks, and extensively demonstrate that (1) PEAC achieves significantly better performance than the existing state-of-the-art fully/self-supervised methods, and (2) PEAC captures the anatomical structure consistency across views of the same patient and across patients of different genders, weights, and healthy statuses, which enhances the interpretability of our method for medical image analysis. All code and pretrained models are available at GitHub.com/JLiangLab/PEAC.