545298 research outputs found
Sort by
Genotype Prediction from Retinal Fundus Images Using Deep Learning in Eyes with Age-Related Macular Degeneration
Get PDFPurpose: To employ deep learning models to predict high-risk genetic variants associated with age-related macular degeneration (AMD) from retinal fundus photographs of patients with this condition.
Design: Deep learning algorithm development to classify single-nucleotide polymorphism in the complement factor H (CFH) and age-related maculopathy susceptibility 2 (ARMS2) genes using retinal fundus images.
Participants: Thirty-one thousand two hundred seventy-one retinal color fundus photographs of 1754 participants from the Age-Related Eye Disease Study.
Methods: We trained deep learning models including convolution neural networks and vision transformers (ViTs) to classify patients into high-risk (homozygous high-risk alleles) or low-risk (heterozygous or homozygous low-risk alleles) genotypes for CFH or ARMS2, then evaluated algorithm performance on an independent test set. The complexity of genotype predictions was compared with AMD severity or gender classification tasks using V-usable information. Attribution mapping was performed to identify fundus regions used to predict genotype from phenotype.
Main Outcome Measures: Area under the receiver operating characteristic curve (AUROC), balanced accuracy, and average precision for predicting high-risk genotypes.
Results: Our trained ViT models predicted high-risk genotypes in CFH and ARMS2 with an AUROC of 0.719 and 0.741 across all eyes, respectively. For genotype predictions for ARMS2, model performance is improved in eyes with advanced AMD (AUROC 0.867), choroidal neovascularization (AUROC 0.833), and geographic atrophy (AUROC 0.957). Genotype predictions from fundus images appear more difficult than AMD severity or gender classification tasks, although saliency mapping supports biological plausibility by demonstrating attention to the central macula for genotype predictions.
Conclusions: Deep learning can predict high-risk genotypes in CFH and ARMS2 from retinal fundus images of patients with AMD. Our findings provide proof of principle for inferring genotype from noninvasive eye imaging and reveal insights into genotype-phenotype relationships in AMD.
Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article
Cortex
No full textMy work centers on prose and lyrical poetry surrounding diasporic Central American narratives. I explore the stories of my family and their surrounding communities. Focusing on the loss, familial separation, colonization, and displacement that have occurred under the various adiministrations from Reagan through to Trump, my work investigates the first-hand effects of US intervention, immigrant journeys, generational trauma, and community resilience. “Cortex” is part of a cycle of poems written during the election season of 2024. The poems aim to understand the connection between longing and homesickness in times of explicit racism and bias. What does the diaspora call home in times like these? 
Cluster randomised controlled trial of a household-level, group preconception nutrition awareness and norm intervention (SUMADHUR) combined with multiple micronutrient supplements (MMS) for newly married households: a protocol
No full textINTRODUCTION: Micronutrient deficiencies remain prominent drivers of adverse maternal and child health outcomes in Nepal. Gender-based inequalities and norms around women's status and access to nutrition exacerbate poor nutritional status. Many newly married, preconception women lack adequate nutrition due to delayed engagement with the health system and limited autonomy to prioritise their own health. To address this gap, the Sumadhur trial provides multiple micronutrient supplements (MMS) alongside a household-level behavioural intervention targeting newly married women, their husbands and mothers-in-law.
METHODS AND ANALYSIS: This will be a village-cluster randomised controlled trial across three districts in Nepal, enrolling 700 households, each comprising a triad of newly married woman, husband and mother-in-law. Villages will be randomised to receive either Sumadhur behavioural intervention+MMS (intervention) or standard of care (control). In intervention villages, participants will join weekly group sessions for 5 months, covering maternal and reproductive health, equitable household food allocation and nutrition information, and gender norms and household relationships. Women will receive three bottles of MMS (180 tablets each) over 18 months. Quantitative data collection at baseline, 6, 12 and 18 months will include surveys, venous blood draws (not at 12 months) and anthropometry. Primary outcomes will be anaemia prevalence and micronutrient status (iron, folate, vitamin B12). Secondary outcomes will include reproductive behaviours, birth outcomes and intrahousehold relationship dynamics. A nested qualitative component will employ longitudinal in-depth interviews with triads to understand the mechanisms of behavioural change. Impact will be measured through an intention-to-treat approach using mixed-effects logistic regression analyses.
ETHICS AND DISSEMINATION: The study is approved by institutional review boards in the Ethics Board of the Nepal Health Research Council and the University of California, San Francisco IRB. Results will be disseminated to participating communities, local stakeholders and international audiences through workshops, peer-reviewed publications and policy briefs.All data will be made publicly available (deidentified) after the publication of the main impact paper.
TRIAL REGISTRATION NUMBER: NCT06810440
Aqueous exposure to a pyrethroid pesticide results in behavioural effects in early life stage sturgeon.
No full textThe presence of chemical contaminants in freshwater systems poses a threat to many aquatic organisms, and understanding the extent and nature of this threat can facilitate conservation management actions. Sturgeon are considered threatened worldwide and they differ in many important ways from other fishes. Two sturgeon species, green sturgeon (Acipenser medirostris) and white sturgeon (A. transmontanus), are found in California and utilize anthropogenically impacted freshwater habitats of the Central Valley. This study evaluated the behavioural effects in endogenously feeding larvae (3-7 days post hatch) of both sturgeon species following an acute exposure (96 hours) to the pyrethroid pesticide bifenthrin at aqueous concentrations ranging from 10 to 2000 ng/l, with selected levels based on previous environmental monitoring. Sturgeon had high survival at all concentrations tested (~95%), yet at higher concentrations (>1000 ng/l) they displayed altered behavioural patterns, including reduced activity, increased meander of the movement path and reduced thigmotaxis. While these higher concentrations of bifenthrin have been observed within water samples from the sturgeon habitats of California, they appear uncommon. The present study suggests that sturgeon larvae are not highly sensitive to acute aqueous exposure under environmentally relevant concentrations of bifenthrin (1-10 ng/l), yet these aqueous concentrations do have behavioural effects that may be of concern for the conservation of these declining species. Additionally, impacts to these species may also occur through exposure to sediment-bound bifenthrin or dietary bioaccumulation, and more work needs to be done to understand the implications of these exposure routes
From Waste to Wellness: Exploring Bioactive Carbohydrates in Plant Byproducts for Improved Health
Get PDFThe rise in plant-based consumption and the resulting accumulation of byproducts have created a strong demand for sustainable valorization strategies. To address this need, the food industry must adopt advanced analytical technologies that optimize extraction methods to preserve the bioactive compounds and their health benefits of plant-based materials. In this context, plant-based byproducts should be rigorously evaluated for their content of bioactive compounds, particularly prebiotic glycans such as oligosaccharides and polysaccharides. Recent advances in mass spectrometry have enabled molecular-level insights into carbohydrate structures. This presentation features two case studies examining carbohydrate composition and abundance in Chardonnay and pomegranate byproducts, using integrated “omics” approaches with a focus on mass spectrometry. It also discusses the strengths and limitations of current extraction and derivatization methods used prior to analysis. Chardonnay and pomegranate pomace contain a fascinating diversity of carbohydrates, including monosaccharides, oligosaccharides, and polysaccharides, accounting for around 20-40% of total weight. A total of 178 and 80 oligosaccharides (including isomers) built from 14 monosaccharide units were characterized in Chardonnay and pomegranate pomace. The discovery of these carbohydrates' presence in side streams can promote their valorization as natural sweeteners, prebiotics, and ingredients for functional food/beverage/skincare products
Early-life Nutritional Supplementation Protects Against Home Environmental Risks in Ghanaian Children’s Social-Emotional Development
Get PDFChildren in low- and middle-income countries (LMICs) face increased risks of undernutrition and adverse home environments, contributing to social-emotional difficulties. Using a randomized controlled trial in Ghana, we examined the direct and interactive effects of prenatal lipid-based nutrient supplementation (LNS) and home environment factors on children’s social-emotional development at age 4-6 years (N = 962). Path analysis models showed that LNS was associated with fewer total difficulties and peer relationship problems, while high-quality home environments—including more responsive caregiving, fewer punitive discipline practices, and greater access to learning materials—predicted better social-emotional outcomes. Multi-group comparisons revealed that LNS moderated associations between home environment factors and children’s total difficulties (χ²(5)=13.746, p=.017) and hyperactivity/inattention (χ²(5)=16.176, p=.006). LNS was linked to fewer difficulties and less hyperactivity/inattention in safer physical environments, whereas the physical environment was not predictive for the control group. Children in the control group had higher hyperactivity/inattention when academic stimulation was lower, an association not observed in the LNS group. These findings suggest that early-life nutritional supplementation differentially interacts with distinct home environment factors, influencing Ghanaian children’s social-emotional development. Integrated interventions targeting both early-life nutritional supplementation and supportive home environments may promote children’s social-emotional development in LMICs
Machine Learning Quantification of Amyloid-β deposits in three brain regions of over 250 Hispanic and Non-Hispanic White Decedents with Alzheimer Disease
Get PDFHistorically, there have been limited autopsy-based studies on persons of Hispanic descent in the realm of Alzheimer Disease (AD). Evaluating more diverse cohorts with ML tools can aid in providing deeper phenotyping on the heterogeneity of the neuropathologic landscape in AD. Our objective was to evaluate densities of amyloid beta (Aβ) deposits in non-Hispanic White decedents (n = 185) and Hispanic decedents (n = 92). Three brain regions were evaluated: frontal, temporal, and parietal cortices. We utilized a previously published ML pipeline to evaluate grey matter and white matter densities (#/um^2) for cerebral amyloid angiopathy, cored and diffuse Aβ plaques and their relation to select pathological, clinical, and demographic variables. Log-transformed linear models of neuroanatomic specific quantification of Aβ deposits revealed 1.71-fold higher CAA density in the temporal lobe GM (95% CI 1.17, 2.51), and lower cored plaque density in the frontal and parietal lobes WM of HD compared to NHWD (0.72-fold (0.56, 0.93) and 0.69-fold (0.52, 0.91) respectively). This study further validates a prior published pipeline and reveals similarities and differences in Aβ densities in a diverse cohort, demonstrating the need for more generalizable results to aid in precision medicine approaches for AD.  
Early Brain Transcriptional Signatures of Adult Male Rats with Differential Seizure Responses to Acute Organophosphate Intoxication
Get PDFAcute organophosphate (OP) poisoning poses a significant public health risk through chemical warfare, accidental exposure, and suicide attempts. OPs inhibit acetylcholinesterase, triggering a cholinergic crisis marked by parasympathomimetic symptoms, respiratory distress, and seizures that may progress to status epilepticus (SE). Survivors often develop chronic neurological deficits despite treatment with atropine, an oxime, and midazolam (MDZ), which, while effective at controlling seizures, does not prevent long-term neurological damage. In a diisopropylfluorophosphate (DFP) rat model, approximately 13% of animals (low responders, LR) do not develop SE, whereas 87% (normal responders, NR) progress to SE. To elucidate seizure-independent mechanisms underlying chronic neurotoxicity, we performed RNA-sequencing on multiple brain regions—the hippocampus, piriform cortex, and cerebellum—in LRs, NRs, and MDZ-pretreated animals. Hierarchical clustering revealed distinct transcriptomic profiles, largely independent of sex. Notably, NR animals exhibited marked upregulation of TGFβ signaling (e.g., Tgfb1, Stat3, c-fos, Jak1) and neuroinflammatory genes (e.g., Cd11b, Gfap, Cd68), suggesting a maladaptive immune response likely driven by severe seizure activity. In contrast, MDZ pre-treatment attenuated these responses and reduced genes associated with protein stabilization and endoplasmic reticulum stress. These findings identify potential therapeutic targets for mitigating chronic OP-induced neuropathology and suggest seizure severity is a significant driver of neuroinflammation
Comparing methods for glomerular filtration rate estimation
No full textBackground: The glomerular filtration rate (GFR), estimated from serum creatinine (SCr), is widely used in clinical practice for kidney function assessment, but SCr-based equations are limited by non-GFR determinants and may introduce inaccuracies across racial groups. Few studies have evaluated whether advanced modeling techniques enhance their performance.
Methods: Using multivariable fractional polynomials (MFP), generalized additive models (GAM), random forests (RF), and gradient boosted machines (GBM), we developed four SCr-based GFR-estimating equations in a pooled data set from four cohorts (n = 4665). Their performance was compared to that of the refitted linear regression-based 2021 CKD-EPI SCr equation using bias (median difference between measured GFR [mGFR] and estimated GFR [eGFR]), precision, and accuracy metrics (e.g., P10 and P30, percentage of eGFR within 10% and 30% of mGFR, respectively) in a pooled validation data set from three additional cohorts (n = 2215).
Results: In the validation data set, the greatest bias and lowest accuracy, were observed in Black individuals for all equations across subgroups defined by race, sex, age, and eGFR. The MFP and GAM equations performed similarly to the refitted CKD-EPI SCr equation, with slight improvements in P10 and P30 in subgroups including Black individuals and females. The GBM and RF equations demonstrated smaller biases, but lower accuracy compared to other equations. Generally, differences among equations were modest overall and across subgroups.
Conclusions: Our findings suggest that advanced methods provide limited improvement in SCr-based GFR estimation. Future research should focus on integrating novel biomarkers for GFR estimation and improving the feasibility of GFR measurement