The implementation of clinical-decision support algorithms for medical imaging faces challenges with reliability and interpretability. Here, we establish a diagnostic tool based on a deep-learning framework for the screening of patients with common treatable blinding retinal diseases. Our framework utilizes transfer learning, which trains a neural network with a fraction of the data of conventional approaches. Applying this approach to a dataset of optical coherence tomography images, we demonstrate performance comparable to that of human experts in classifying age-related macular degeneration and diabetic macular edema. We also provide a more transparent and interpretable diagnosis by highlighting the regions recognized by the neural network. We further demonstrate the general applicability of our AI system for diagnosis of pediatric pneumonia using chest X-ray images. This tool may ultimately aid in expediting the diagnosis and referral of these treatable conditions, thereby facilitating earlier treatment, resulting in improved clinical outcomes. VIDEO ABSTRACT.
The human gut microbiome can modulate metabolic health and affect insulin resistance, and it may play an important role in the etiology of gestational diabetes mellitus (GDM). Here, we compared the gut microbial composition of 43 GDM patients and 81 healthy pregnant women via whole-metagenome shotgun sequencing of their fecal samples, collected at 21–29 weeks, to explore associations between GDM and the composition of microbial taxonomic units and functional genes. A metagenome-wide association study identified 154 837 genes, which clustered into 129 metagenome linkage groups (MLGs) for species description, with significant relative abundance differences between the 2 cohorts. Parabacteroides distasonis, Klebsiella variicola, etc., were enriched in GDM patients, whereas Methanobrevibacter smithii, Alistipes spp., Bifidobacterium spp., and Eubacterium spp. were enriched in controls. The ratios of the gross abundances of GDM-enriched MLGs to control-enriched MLGs were positively correlated with blood glucose levels. A random forest model shows that fecal MLGs have excellent discriminatory power to predict GDM status. Our study discovered novel relationships between the gut microbiome and GDM status and suggests that changes in microbial composition may potentially be used to identify individuals at risk for GDM.
Few studies have explored the relationship between dietary patterns and the risk of gestational diabetes mellitus (GDM). Evidence from non-Western areas is particularly lacking. In the present study, we aimed to examine the associations between dietary patterns and the risk of GDM in a Chinese population. A total of 3063 pregnant Chinese women from an ongoing prospective cohort study were included. Data on dietary intake were collected using a FFQ at 24-27 weeks of gestation. GDM was diagnosed using a 75 g, 2 h oral glucose tolerance test. Dietary patterns were determined by principal components factor analysis. A log-binomial regression model was used to examine the associations between dietary pattern and the risk of GDM. The analysis identified four dietary patterns: vegetable pattern; protein-rich pattern; prudent pattern; sweets and seafood pattern. Multivariate analysis showed that the highest tertile of the vegetable pattern was associated with a decreased risk of GDM (relative risk (RR) 0·79, 95 % CI 0·64, 0·97), compared with the lowest tertile, whereas the highest tertile of the sweets and seafood pattern was associated with an increased risk of GDM (RR 1·23, 95 % CI 1·02, 1·49). No significant association was found for either the protein-rich or the prudent pattern. The protective effect of a high vegetable pattern score was more evident among women who had a family history of diabetes (P for interaction¼ 0·022). These findings suggest that the vegetable pattern was associated with a decreased risk of GDM, while the sweets and seafood pattern was associated with an increased risk of GDM. These findings may be useful in dietary counselling during pregnancy.
BackgroundHerpes simplex virus type 1 strain 129 (H129) has represented a promising anterograde neuronal circuit tracing tool, which complements the existing retrograde tracers. However, the current H129 derived tracers are multisynaptic, neither bright enough to label the details of neurons nor capable of determining direct projection targets as monosynaptic tracer.MethodsBased on the bacterial artificial chromosome of H129, we have generated a serial of recombinant viruses for neuronal circuit tracing. Among them, H129-G4 was obtained by inserting binary tandemly connected GFP cassettes into the H129 genome, and H129-ΔTK-tdT was obtained by deleting the thymidine kinase (TK) gene and adding tdTomato coding gene to the H129 genome. Then the obtained viral tracers were tested in vitro and in vivo for the tracing capacity.ResultsH129-G4 is capable of transmitting through multiple synapses, labeling the neurons by green florescent protein, and visualizing the morphological details of the labeled neurons. H129-ΔTK-tdT neither replicates nor spreads in neurons alone, but transmits to and labels the postsynaptic neurons with tdTomato in the presence of complementary expressed TK from a helper virus. H129-ΔTK-tdT is also capable to map the direct projectome of the specific neuron type in the given brain regions in Cre transgenic mice. In the tested brain regions where circuits are well known, the H129-ΔTK-tdT tracing patterns are consistent with the previous results.ConclusionsWith the assistance of the helper virus complimentarily expressing TK, H129-ΔTK-tdT replicates in the initially infected neuron, transmits anterogradely through one synapse, and labeled the postsynaptic neurons with tdTomato. The H129-ΔTK-tdT anterograde monosynaptic tracing system offers a useful tool for mapping the direct output in neuronal circuitry. H129-G4 is an anterograde multisynaptic tracer with a labeling signal strong enough to display the details of neuron morphology.Electronic supplementary materialThe online version of this article (doi:10.1186/s13024-017-0179-7) contains supplementary material, which is available to authorized users.
Background:Although effects of weather changes on human health have been widely reported, there is limited information regarding effects on pregnant women in developing countries.Objective:We investigated the association between maternal exposure to ambient temperature and the risk of preterm birth (< 37 weeks of gestation) in Guangzhou, China.Methods:We used a Cox proportional hazards model to estimate associations between preterm birth and average temperature during each week of gestation, with weekly temperature modeled as a time-varying exposure during four time windows: 1 week (the last week of the pregnancy), 4 weeks (the last 4 weeks of the pregnancy), late pregnancy (gestational week 20 onward), and the entire pregnancy. Information on singleton vaginal birth between 2001 and 2011 was collected. Daily meteorological data during the same period were obtained from the Guangzhou Meteorological Bureau.Results:A total of 838,146 singleton vaginal births were included, among which 47,209 (5.6%) were preterm births. High mean temperatures during the 4 weeks, late pregnancy, and the entire pregnancy time windows were associated with an increased risk of preterm birth. Compared with the median temperature (24.4°C), weekly exposures during the last 4 weeks of the pregnancy to extreme cold (7.6°C, the 1st percentile) and extreme heat (31.9°C, the 99th percentile) were associated with 17.9% (95% CI: 10.2, 26.2%) and 10.0% (95% CI: 2.9, 17.6%) increased risks of preterm birth, respectively. The association between extreme heat and preterm birth was stronger for preterm births during weeks 20–31 and 32–34 than those during weeks 35–36.Conclusions:These findings might have important implications in preventing preterm birth in Guangzhou as well as other areas with similar weather conditions.Citation:He JR, Liu Y, Xia XY, Ma WJ, Lin HL, Kan HD, Lu JH, Feng Q, Mo WJ, Wang P, Xia HM, Qiu X, Muglia LJ. 2016. Ambient temperature and the risk of preterm birth in Guangzhou, China (2001–2011). Environ Health Perspect 124:1100–1106; http://dx.doi.org/10.1289/ehp.1509778
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.