27941 research outputs found
Sort by
A systematic review and meta-analysis of GPT-based differential diagnostic accuracy in radiological cases: 2023-2025
Objective: To systematically evaluate the diagnostic accuracy of various GPT models in radiology, focusing on differential diagnosis performance across textual and visual input modalities, model versions, and clinical contexts.
Methods: A systematic review and meta-analysis were conducted using PubMed and SCOPUS databases on March 24, 2025, retrieving 639 articles. Studies were eligible if they evaluated GPT model diagnostic accuracy on radiology cases. Non-radiology applications, fine-tuned/custom models, board-style multiple-choice questions, or studies lacking accuracy data were excluded. After screening, 28 studies were included. Risk of bias was assessed using the Newcastle-Ottawa Scale (NOS). Diagnostic accuracy was assessed as top diagnosis accuracy (correct diagnosis listed first) and differential accuracy (correct diagnosis listed anywhere). Statistical analysis involved Mann-Whitney U tests using study-level median (median) accuracy with interquartile ranges (IQR), and a generalized linear mixed-effects model (GLMM) to evaluate predictors influencing model performance.
Results: Analysis included 8,852 radiological cases across multiple radiology subspecialties. Differential accuracy varied significantly among GPT models, with newer models (GPT-4T: 72.00%, median 82.32%; GPT-4o: 57.23%, median 53.75%; GPT-4: 56.46%, median 56.65%) outperforming earlier versions (GPT-3.5: 37.87%, median 36.33%). Textual inputs demonstrated higher accuracy (GPT-4: 56.46%, median 58.23%) compared to visual inputs (GPT-4V: 42.32%, median 41.41%). The provision of clinical history was associated with improved diagnostic accuracy in the GLMM (OR = 1.27, p = .001), despite unadjusted medians showing lower performance when history was provided (61.74% vs. 52.28%). Private data (86.51%, median 94.00%) yielded higher accuracy than public data (47.62%, median 46.45%). Accuracy trends indicated improvement in newer models over time, while GPT-3.5's accuracy declined. GLMM results showed higher odds of accuracy for advanced models (OR = 1.84), and lower odds for visual inputs (OR = 0.29) and public datasets (OR = 0.34), while accuracy showed no significant trend over successive study years (p = 0.57). Egger's test found no significant publication bias, though considerable methodological heterogeneity was observed.
Conclusion: This meta-analysis highlights significant variability in GPT model performance influenced by input modality, data source, and model version. High methodological heterogeneity across studies emphasizes the need for standardized protocols in future research, and readers should interpret pooled estimates and medians with this variability in mind.No embarg
ACR Appropriateness Criteria® Acute Left Upper Quadrant Pain
Acute abdominal pain is one of the most common chief complaints encountered in the emergency department and outpatient setting. Across all sites, the left upper quadrant (LUQ) is the least common location for abdominal pain with reportedly the lowest physical examination interrater agreement. The limited sensitivity of physical examination, combined with an uncommon and nonspecific clinical presentation, make assessment of LUQ pain challenging. Acute LUQ pain may arise directly from the spleen or infections in the LUQ, or represent referred pain from other organs such as stomach, bowels, pancreas, or kidney. This document aims to address acute pain with suspected splenomegaly, with fever, and not otherwise specified. CT abdomen and pelvis with contrast is usually appropriate for detection of a broad range of pathologies. In patients with suspected splenomegaly, ultrasound abdomen is also usually an appropriate alternate for initial imaging. The American College of Radiology Appropriateness Criteria are evidence-based guidelines for specific clinical conditions that are reviewed annually by a multidisciplinary expert panel. The guideline development and revision process support the systematic analysis of the medical literature from peer reviewed journals. Established methodology principles such as Grading of Recommendations Assessment, Development, and Evaluation or GRADE are adapted to evaluate the evidence. The RAND/UCLA Appropriateness Method User Manual provides the methodology to determine the appropriateness of imaging and treatment procedures for specific clinical scenarios. In those instances where peer reviewed literature is lacking or equivocal, experts may be the primary evidentiary source available to formulate a recommendation.No embarg
Prevalence of Food Insufficiency Across Subgroups of Children with Special Healthcare Needs
OBJECTIVE: To determine the association between subgroups of children with special healthcare needs (SHCN) and food insufficiency, and assess whether this association varies by income level, and to evaluate how food insufficiency trends have changed over the time.
STUDY DESIGN: This was a cross-sectional survey study using the 2016-2023 National Surveys of Children's Health. SHCN subgroups were defined as follows: no special healthcare needs, prescription medication use only, elevated use of services, and functional limitations. The association between SHCN subgroup and food insufficiency was measured using weighted multivariable logistic regression models. Effect modification by income was evaluated. In addition, linear models described significant changes in food insufficiency rates by SHCN subgroup from 2016-2023.
RESULTS: Compared with children without SHCN, children with SHCN who used medications only (aOR=1.31; 95% CI 1.21-1.41), had elevated use of services (aOR=1.54; 95% CI 1.45-1.63), or had functional limitations (aOR 1.97; 95% CI 1.82-2.13) had higher odds of food insufficiency. Effect modification by poverty was significant for children with functional limitations. From 2016-2023, the associations between SHCN and food insufficiency were similar.
CONCLUSION: Children with functional limitations are an especially high-risk group who require attention in public health efforts to reduce food insufficiency.No embarg
Worm Perturb-Seq: massively parallel whole-animal RNAi and RNA-seq [preprint]
This article is a preprint. Preprints are preliminary reports of work that have not been certified by peer review.The transcriptome provides a highly informative molecular phenotype to connect genotype to phenotype and is most frequently measured by RNA-sequencing (RNA-seq). Therefore, an ultimate goal is to perturb every gene and measure changes in the transcriptome. However, this remains challenging, especially in intact organisms due to different experimental and computational challenges. Here, we present 'Worm Perturb-Seq (WPS)', which provides high-resolution RNA-seq profiles for hundreds of replicate perturbations at a time in a living animal. WPS introduces multiple experimental advances that combine strengths of bulk and single cell RNA-seq, and that further provides an analytical framework, EmpirDE, that leverages the unique power of the large WPS datasets. EmpirDE identifies differentially expressed genes (DEGs) by using gene-specific empirical null distributions, rather than control conditions alone, thereby systematically removing technical biases and improving statistical rigor. We applied WPS to 103 nuclear hormone receptors (NHRs) to delineate a Gene Regulatory Network (GRN) and found that this GRN presents a striking 'pairwise modularity' where pairs of NHRs regulate shared target genes. We envision that the experimental and analytical advances of WPS should be useful not only for , but will be broadly applicable to other models, including human cells.No embarg
VPS13D mutations affect mitochondrial homeostasis and locomotion in Caenorhabditis elegans
Mitochondria control cellular metabolism, serve as hubs for signaling and organelle communication, and are important for the health and survival of cells. VPS13D encodes a cytoplasmic lipid transfer protein that regulates mitochondrial morphology, mitochondria and endoplasmic reticulum (ER) contact, quality control of mitochondria. VPS13D mutations have been reported in patients displaying ataxic and spastic gait disorders with variable age of onset. Here we used CRISPR/Cas9 gene editing to create VPS13D related-spinocerebellar ataxia-4 (SCAR4) missense mutations and C-terminal deletion in VPS13D's orthologue vps-13D in C. elegans. Consistent with SCAR4 patient movement disorders and mitochondrial dysfunction, vps-13D mutant worms exhibit locomotion defects and abnormal mitochondrial morphology. Importantly, animals with a vps-13D deletion or a N3017I missense mutation exhibited an increase in mitochondrial unfolded protein response (UPRmt). The cellular and behavioral changes caused by VPS13D mutations in C. elegans advance the development of animal models that are needed to study SCAR4 pathogenesis.No embarg
Stories of life during the COVID-19 pandemic: Personal growth among individuals living in the Italian speaking part of Switzerland
Multiple influences such as climate change, environmental, and social factors, have been increasing the incidence of infectious disease outbreaks. Therefore, it is imperative to understand how people make sense of such events. Our study focuses on 100 Swiss-Italian participants during the initial COVID-19 pandemic wave, investigating how they navigated this period. We collected "Stories of Life," personal narratives about their experience and analyzed them with Thematic Analysis. Results show that over a third of the participants made sense of the pandemic by framing it as a time for personal growth. Our themes illustrated the importance of relationships, resilience, and appreciation of time. These findings align with existing literature on personal growth domains and underscore the significance of relational health in adversity. We discuss personal growth during a pandemic, emphasizing the need to consider socio-demographic characteristics and supportive social contexts, which can create inequalities in circumstances promoting growth.No embarg
Usability and Implementation Considerations of Fitbit and App Intervention for Diverse Cancer Survivors: Mixed Methods Study
Background: Despite the known benefits of physical activity, cancer survivors remain insufficiently active. Prior trials have adopted digital health methods, although several have been pedometer-based and enrolled mainly female, non-Hispanic White, and more highly educated survivors of breast cancer.
Objective: The objective of this study was to test a previously developed mobile health system consisting of a Fitbit activity tracker and the MyDataHelps smartphone app for feasibility in a diverse group of cancer survivors, with the goal of refining the program and setting the stage for a larger future trial.
Methods: Participants were identified from one academic medical center's electronic health records, referred by a clinician, or self-referred to participate in the study. Participants were screened for eligibility, enrolled, provided a Fitbit activity tracker, and instructed to download the Fitbit: Health & Wellness and MyDataHelps apps. They completed usability surveys at 1 and 3 months. Interviews were conducted at the end of the 3-month intervention with participants and cancer care clinicians to assess the acceptability of the intervention and the implementation of the intervention into clinical practice, respectively. Descriptive statistics were calculated for demographics, usability surveys, and Fitbit adherence and step counts. Rapid qualitative analysis was used to identify key findings from interview transcriptions.
Results: Of the 100 patients screened for eligibility, 31 were enrolled in the trial (mean age 64.8, SD 11.1 years; female patients=17/31, 55%; Hispanic or Latino=7/31, 23%; non-White=11/31, 35%; less than a bachelor's degree=14/31, 45%; and household income <US $75,000=11/31, 35%). The mean (SD) years since diagnosis was 7.1 (8.2), and the two most frequent cancer diagnoses were prostate (9/31, 29%) and breast (4/31, 13%) cancer. Participants provided positive feedback on the MyDataHelps app usability; the overall app quality received a mean score of 3.79 (SD 0.82) on a 5-point Likert scale (1=worst, 5=best). Interviews with 10 patients yielded four themes: (1) Fitbit and app setup was easy but the research team provided assistance, when needed, which was helpful, (2) motivational messages within the app were not memorable, (3) step counts and Fitbit notifications were motivating, and (4) medical professionals viewing their data were acceptable. Interviews with 5 cancer care clinicians yielded four themes: (1) some patients used wearables but rarely discussed data with clinicians; (2) activity trackers can be helpful to motivate patients and keep them accountable; (3) objective activity measures-similar to BMI, weight, and blood pressure- that they can track over time and refer to afterward were preferred; and (4) training and systematic processes to view these data as part of active workflow were desired.
Conclusions: Implementing a remotely delivered, light-intensity physical activity program was feasible and acceptable in a sample of diverse cancer survivors. Future studies should consider registry-based methods and work with clinicians to engage hard-to-reach survivor populations who have low physical activity levels and disproportionately high adverse health outcomes.No embarg
The Use of CT Densitometry for the Assessment of Emphysema in Clinical Trials: A Position Paper from the Fleischner Society
Emphysema's significant morbidity and mortality underscore the need for reliable outcome metrics in clinical trials. However, commonly accepted COPD outcome measures do not adequately capture emphysema severity or progression. Computed tomography (CT) metrics have been validated as accurate indicators of pathological emphysema and predictors of COPD progression, exacerbations, and mortality. This Position Paper reviews the evidence supporting CT densitometry as a biomarker for emphysema, establishes implementation standards, and highlights areas for future research. A systematic literature review addressed three key questions: whether CT densitometry can be used as a diagnostic biomarker of emphysema, whether CT densitometry can be used as prognostic biomarker, and whether longitudinal change in densitometry can be used as a disease progression monitoring biomarker. Emphysema metrics, such as the percentage of low attenuation areas (LAA), are validated, highly reproducible diagnostic and prognostic biomarkers. Volume-adjusted lung density is recommended for disease monitoring. Both metrics demonstrate a scan-rescan intra-class correlation coefficient of 0.99 with proper technique. The paper also discusses relevant CT physics, techniques, and sources of variation, including technical factors, physiological changes, and software analysis. Key recommendations for clinical trials include using standardized CT techniques, proper subject selection, and longitudinal evaluation with volume-adjusted lung density.No embarg
Incorporating Career and Technical Education in Transition Planning for Students with Emotional Disturbance, Second Edition
Youth and young adults (ages 14–24) living with mental health conditions are capable of successfully engaging in school, training, and employment. Unfortunately, they often face barriers that, when compared to youth with other disabilities or youth without disabilities, result in higher rates of high school drop-out, lower rates of post-secondary education and employment, and even higher rates of involvement with law enforcement, poverty, and homelessness upon their exit from high school.
However, with the right information, tools, and resources, educators can positively impact these students. The support these individuals receive as they progress through secondary education can help them overcome the barriers and realize their potential in life after high school. Educators need resources to provide those supports, including assistance with planning and preparing for student transition from high school into education and training programs and employment in young adulthood.
This Guide provides practical and actionable steps to incorporating Career and Technical Education (CTE) into planning the transition component of an Individual Education Program (IEP). It is written for educators who conduct transition planning for students with emotional disturbance. This second edition of “Incorporating Career and Technical Education in Transition Planning for Students with Emotional Disturbance” has been updated from the 2020 version by offering special considerations for students with culturally diverse backgrounds, as well as underrepresented or marginalized populations of students who have historically been excluded from mainstream educational, social, or other aspects of life, or students living with limited resources. The second edition incorporates feedback from a focus group of former CTE students from diverse and underrepresented backgrounds and expert input from Evelyn Frankford, Abimelech Velazco, Maya Ingram, and Dr. Jonathan Delman, whom we gratefully acknowledge.
This Guide is designed for special educators who are conducting transition planning with students as part of their Individualized Education Program (IEP); and special educators who may have limited training in working with students with mental health conditions. The Guide is also useful for student support staff, guidance counselors, and CTE Instructors who serve students with mental health conditions whether they receive special education services.
Visit our website (https://www.umassmed.edu/TransitionsACR/models/test/) to learn more.The contents of this manual were developed under grants with funding from the National Institute on Disability, Independent Living, and Rehabilitation Research (NIDILRR) (grant# A-90DP0063 and 90IFRE0064). NIDILRR is a Center within the Administration for Community Living (ACL), Department of Health and Human Services (HHS). The contents of this manual do not necessarily represent the policy of NIDILRR, and you should not assume endorsement by the Federal Government.No embarg
Human Cytomegalovirus Infection of Primary Human Oral Keratinocytes Induces Intermediate Keratinocyte Differentiation and an Altered Innate Immune Response [preprint]
This article is a preprint. Preprints are preliminary reports of work that have not been certified by peer review.The oral mucosa is central for human cytomegalovirus (HCMV) acquisition. However, the initial infection events in the oral cavity, including how HCMV modulates oral cells to establish infection and spread, are supported by limited experimental data and therefore remain poorly understood. We investigated HCMV infection in cultures of primary human oral keratinocytes (HOKs) as a model to study HCMV infection in the oral mucosa using the TB40E-Gfp and MOLD strains of HCMV. Both viral strains successfully infected and replicated in HOKs as demonstrated by the appearance of cytopathic effects, full expression of viral transcripts, and the shedding of infectious virus. To investigate how HCMV modulates the transcriptome of primary HOKs, we performed single cell RNA sequencing (scRNA-seq) which revealed 11 subtypes of cells. Furthermore, infection resulted in significant changes in host gene expression. HCMV upregulated genes involved in immune responses, cell cycle regulation, cancer-related pathways, neuronal and synaptic functions, metabolism, stress responses, molecular chaperones, and vesicular trafficking which might be critical for viral protein expression and assembly. HCMV also increased the expression of genes related to the extracellular matrix, cell adhesion, microtubules, signal transduction, kinases, and transcriptional regulation. Conversely, HCMV downregulated genes associated with inflammation and immune responses, cell cycle control and apoptosis, cell adhesion and migration, as well as signaling pathways, growth factors, ion channels, transporters, and transcription factors when compared to uninfected cells. These findings suggest that HCMV modulates a wide range of host cellular pathways to create a favorable environment for its replication. HCMV also induced changes in HOK differentiation genes downregulating basal state genes and upregulating intermediate genes and select terminal differentiation genes, indicating HCMV might be driving HOK differentiation towards a dysregulated intermediate phenotype while inhibiting their terminal differentiation to maintain a cell state capable of sustaining viral replication. Using SciViewer, our recently developed scRNA-seq data visualization tool, we determined that as the infection progressed from low to high viral transcript accumulation, HCMV infection upregulated E2F targeted genes, some antiviral and immune responses as well as autophagy and cellular stress responses, while downregulating interferon-stimulated genes, immune and antiviral response genes, pyroptosis, inflammatory cell death, and membrane remodeling and viral entry genes. Overall, our study shows that primary human HOKs can be used as a model to study HCMV infection in the oral mucosa and that HCMV drives dysregulated intermediate oral keratinocyte differentiation and altered immune responses which likely facilitate viral replication in the oral mucosa and spread into the rest of the host.No embarg