Hasil untuk "Pediatrics"

S. Hyman, S. Levy, S. Myers

Autism spectrum disorder (ASD) is a common neurodevelopmental disorder with reported prevalence in the United States of 1 in 59 children (approximately 1.7%). Core deficits are identified in 2 domains: social communication/interaction and restrictive, repetitive patterns of behavior. Children and youth with ASD have service needs in behavioral, educational, health, leisure, family support, and other areas. Standardized screening for ASD at 18 and 24 months of age with ongoing developmental surveillance continues to be recommended in primary care (although it may be performed in other settings), because ASD is common, can be diagnosed as young as 18 months of age, and has evidenced-based interventions that may improve function. More accurate and culturally sensitive screening approaches are needed. Primary care providers should be familiar with the diagnostic criteria for ASD, appropriate etiologic evaluation, and co-occurring medical and behavioral conditions (such as disorders of sleep and feeding, gastrointestinal tract symptoms, obesity, seizures, attention-deficit/hyperactivity disorder, anxiety, and wandering) that affect the child’s function and quality of life. There is an increasing evidence base to support behavioral and other interventions to address specific skills and symptoms. Shared decision making calls for collaboration with families in evaluation and choice of interventions. This single clinical report updates the 2007 American Academy of Pediatrics clinical reports on the evaluation and treatment of ASD in one publication with an online table of contents and section view available through the American Academy of Pediatrics Gateway to help the reader identify topic areas within the report.

J. Kevill

Christopher P. Johnson, S. Myers

Autism spectrum disorders are not rare; many primary care pediatricians care for several children with autism spectrum disorders. Pediatricians play an important role in early recognition of autism spectrum disorders, because they usually are the first point of contact for parents. Parents are now much more aware of the early signs of autism spectrum disorders because of frequent coverage in the media; if their child demonstrates any of the published signs, they will most likely raise their concerns to their child's pediatrician. It is important that pediatricians be able to recognize the signs and symptoms of autism spectrum disorders and have a strategy for assessing them systematically. Pediatricians also must be aware of local resources that can assist in making a definitive diagnosis of, and in managing, autism spectrum disorders. The pediatrician must be familiar with developmental, educational, and community resources as well as medical subspecialty clinics. This clinical report is 1 of 2 documents that replace the original American Academy of Pediatrics policy statement and technical report published in 2001. This report addresses background information, including definition, history, epidemiology, diagnostic criteria, early signs, neuropathologic aspects, and etiologic possibilities in autism spectrum disorders. In addition, this report provides an algorithm to help the pediatrician develop a strategy for early identification of children with autism spectrum disorders. The accompanying clinical report addresses the management of children with autism spectrum disorders and follows this report on page 1162 [available at www.pediatrics.org/cgi/content/full/120/5/1162]. Both clinical reports are complemented by the toolkit titled “Autism: Caring for Children With Autism Spectrum Disorders: A Resource Toolkit for Clinicians,” which contains screening and surveillance tools, practical forms, tables, and parent handouts to assist the pediatrician in the identification, evaluation, and management of autism spectrum disorders in children.

J. Duffee, A. Kuo, B. Gitterman et al.

Almost half of young children in the United States live in poverty or near poverty. The American Academy of Pediatrics is committed to reducing and ultimately eliminating child poverty in the United States. Poverty and related social determinants of health can lead to adverse health outcomes in childhood and across the life course, negatively affecting physical health, socioemotional development, and educational achievement. The American Academy of Pediatrics advocates for programs and policies that have been shown to improve the quality of life and health outcomes for children and families living in poverty. With an awareness and understanding of the effects of poverty on children, pediatricians and other pediatric health practitioners in a family-centered medical home can assess the financial stability of families, link families to resources, and coordinate care with community partners. Further research, advocacy, and continuing education will improve the ability of pediatricians to address the social determinants of health when caring for children who live in poverty. Accompanying this policy statement is a technical report that describes current knowledge on child poverty and the mechanisms by which poverty influences the health and well-being of children.

R. Moon

Approximately 3500 infants die annually in the United States from sleep-related infant deaths, including sudden infant death syndrome (SIDS), ill-defined deaths, and accidental suffocation and strangulation in bed. After an initial decrease in the 1990s, the overall sleep-related infant death rate has not declined in more recent years. Many of the modifiable and nonmodifiable risk factors for SIDS and other sleep-related infant deaths are strikingly similar. The American Academy of Pediatrics recommends a safe sleep environment that can reduce the risk of all sleep-related infant deaths. Recommendations for a safe sleep environment include supine positioning, use of a firm sleep surface, room-sharing without bed-sharing, and avoidance of soft bedding and overheating. Additional recommendations for SIDS risk reduction include avoidance of exposure to smoke, alcohol, and illicit drugs; breastfeeding; routine immunization; and use of a pacifier. New evidence and rationale for recommendations are presented for skin-to-skin care for newborn infants, bedside and in-bed sleepers, sleeping on couches/armchairs and in sitting devices, and use of soft bedding after 4 months of age. In addition, expanded recommendations for infant sleep location are included. The recommendations and strength of evidence for each recommendation are published in the accompanying policy statement, “SIDS and Other Sleep-Related Infant Deaths: Updated 2016 Recommendations for a Safe Infant Sleeping Environment,” which is included in this issue.

Wei Hu, Qin Liu, Xuemei Fu et al.

Wei-Cheng Chao, Asaduzzaman Khan, Jui-Chi Shih et al.

Background: Chinese-Australian adolescents face unique academic and cultural challenges that may impact their lifestyle and psychological well-being. Physical activity, screen time, and sleep are known to influence well-being. However, research on the adherence to the 24-Hour Movement Guidelines among Chinese-Australian adolescents remains limited and awaits further investigation. Objective: This study hypothesized a significant positive association between adherence to the 24-Hour Movement Guidelines for physical activity, screen time, and sleep, and the psychological well-being of Chinese-Australian adolescents. Methods: A self-reported questionnaire was distributed to two language schools in Brisbane, Australia, targeting high school students from grades 7 to 12 with Chinese-Australian backgrounds. This study used multiple linear regression modelling to examine the associations between meeting or not meeting recommendations. Meeting the 24-Hour Movement Guidelines was defined as ≥60 min/day of moderate to vigorous physical activity (MVPA), ≤2 h/day of recreational screen time, and 9–11 h/night of sleep. Results: Out of 251 participants (average age: 13.31 years; 58% female), only 20.3% met two or three recommendations, while 43.3% met one, and 36.2% met none. The most common compliance was meeting only the screen time guideline alone (48%), while 9.6% met either MVPA + screen time or screen time + sleep. The regression analysis showed that meeting at least MVPA (β = 1.41, 95% CI: 0.07 to 2.74) or at least sleep (β = 1.40, 95% CI: 0.19 to 2.60) was associated with better psychological well-being. Notably, meeting MVPA and sleep guidelines was significantly associated with higher well-being (β = 3.83, 95% CI: 1.06–6.60). From the results, adherence to additional 24-Hour Movement Guidelines was associated with improved psychosocial well-being. However, a small proportion of adolescents met all the guidelines. Conclusions: Greater adherence to physical activity and sleep guidelines is linked to better psychological well-being among Chinese-Australian adolescents. These results highlight the importance of promoting healthy behaviours and implementing public health strategies to enhance education on exercise and sleep, particularly at the school and family levels, to support adolescents’ psychological well-being.

Ali ME, Hassan YO, Ahmed MA et al.

Mohamud Eyow Ali,1 Yusuf Omar Hassan,2 Mohammed AM Ahmed,3 Liban Bile Mohamud4 1Department of Pediatrics, Mogadishu Somali Turkish Training and Research Hospital, Mogadishu, Somalia; 2Department of Pediatrics and Child Health, Mogadishu University, Mogadishu, Somalia; 3Department of Pediatric and Congenital Cardiology, Mogadishu Heart Center, Mogadishu, Somalia; 4Somali National Bureau of Statistics, Mogadishu, SomaliaCorrespondence: Mohamud Eyow Ali, Email mohamud.eyow@mu.edu.soBackground: Neonatal mortality is a significant global health challenge, particularly in sub-Saharan Africa. In Somalia, there is a notable absence of comprehensive reports or data on neonatal mortality rates within tertiary-level neonatal intensive care units (NICU). This study aims to identify key factors associated with neonatal mortality in Mogadishu, Somalia.Materials and Methods: A retrospective review of medical records was conducted for neonates admitted to the Neonatal Intensive Care Unit (NICU) of Mogadishu Somali Turkish Training and Research Hospital from August 2017 to September 2019. Logistic regression analysis was employed using SPSS (version 25) to compute adjusted odds ratios (aORs) along with 95% confidence intervals (CIs).Results: Of 1043 neonates, 63.8% (n=665) were male, with a mean age of 1.48 days. Most neonates were full-term (55.3%, n=577), while 25.8% (n=269) were preterm (< 32 weeks), and 11.9% (n=124) were late preterm (33– 37 weeks). In total, 25.5% (n=266) had very low birth weight (< 1500 grams). The average length of stay in the NICU was 7.38 days, and the overall mortality rate was 18.7% (n=195). Indications for NICU admissions were prematurity 27.0% (n=282), followed by birth asphyxia (18.0%, n=188), neonatal sepsis (14.6%, n=152), and acute respiratory distress syndrome (12.2%, n=127). Preterm neonates had significantly higher mortality rates (OR=2.14, 95% CI: 1.32– 3.47, p=0.002), and those with a birth weight of < 1500 grams had an even higher risk of mortality (OR=3.85, 95% CI: 2.50– 5.92, p< 0.001). Lack of ANC visits was associated with increased mortality risk (OR=1.67, 95% CI: 1.09– 2.54, p=0.019), while cesarean delivery was also linked to higher mortality risk (OR=1.92, 95% CI: 1.29– 2.85, p=0.002).Conclusion: The study identified a Neonatal Mortality Rate that is acceptable compared to the mortality rates in other studies in Somalia and the sub-Saharan African region. These findings inform care strategies and resource allocation in prenatal and neonatal health services.Keywords: neonatal mortality, low-resource settings, tertiary-level NICU, preterm birth, neonatal outcomes, Mogadishu, Somalia

Jiacheng Xie, Yang Yu, Ziyang Zhang et al.

Traditional Chinese Medicine (TCM), as an effective alternative medicine, has been receiving increasing attention. In recent years, the rapid development of large language models (LLMs) tailored for TCM has highlighted the urgent need for an objective and comprehensive evaluation framework to assess their performance on real-world tasks. However, existing evaluation datasets are limited in scope and primarily text-based, lacking a unified and standardized multimodal question-answering (QA) benchmark. To address this issue, we introduce TCM-Ladder, the first comprehensive multimodal QA dataset specifically designed for evaluating large TCM language models. The dataset covers multiple core disciplines of TCM, including fundamental theory, diagnostics, herbal formulas, internal medicine, surgery, pharmacognosy, and pediatrics. In addition to textual content, TCM-Ladder incorporates various modalities such as images and videos. The dataset was constructed using a combination of automated and manual filtering processes and comprises over 52,000 questions. These questions include single-choice, multiple-choice, fill-in-the-blank, diagnostic dialogue, and visual comprehension tasks. We trained a reasoning model on TCM-Ladder and conducted comparative experiments against nine state-of-the-art general domain and five leading TCM-specific LLMs to evaluate their performance on the dataset. Moreover, we propose Ladder-Score, an evaluation method specifically designed for TCM question answering that effectively assesses answer quality in terms of terminology usage and semantic expression. To the best of our knowledge, this is the first work to systematically evaluate mainstream general domain and TCM-specific LLMs on a unified multimodal benchmark. The datasets and leaderboard are publicly available at https://tcmladder.com and will be continuously updated.

Daria Laslo, Efthymios Georgiou, Marius George Linguraru et al.

Predicting the spatio-temporal progression of brain tumors is essential for guiding clinical decisions in neuro-oncology. We propose a hybrid mechanistic learning framework that combines a mathematical tumor growth model with a guided denoising diffusion implicit model (DDIM) to synthesize anatomically feasible future MRIs from preceding scans. The mechanistic model, formulated as a system of ordinary differential equations, captures temporal tumor dynamics including radiotherapy effects and estimates future tumor burden. These estimates condition a gradient-guided DDIM, enabling image synthesis that aligns with both predicted growth and patient anatomy. We train our model on the BraTS adult and pediatric glioma datasets and evaluate on 60 axial slices of in-house longitudinal pediatric diffuse midline glioma (DMG) cases. Our framework generates realistic follow-up scans based on spatial similarity metrics. It also introduces tumor growth probability maps, which capture both clinically relevant extent and directionality of tumor growth as shown by 95th percentile Hausdorff Distance. The method enables biologically informed image generation in data-limited scenarios, offering generative-space-time predictions that account for mechanistic priors.

Yuanhan Wang, Yifei Chen, Shuo Jiang et al.

Reliable brain tumor segmentation in MRI is indispensable for treatment planning and outcome monitoring, yet models trained on curated benchmarks often fail under domain shifts arising from scanner and protocol variability as well as population heterogeneity. Such gaps are especially severe in low-resource and pediatric cohorts, where conventional test-time or source-free adaptation strategies often suffer from instability and structural inconsistency. We propose SmaRT, a style-modulated robust test-time adaptation framework that enables source-free cross-domain generalization. SmaRT integrates style-aware augmentation to mitigate appearance discrepancies, a dual-branch momentum strategy for stable pseudo-label refinement, and structural priors enforcing consistency, integrity, and connectivity. This synergy ensures both adaptation stability and anatomical fidelity under extreme domain shifts. Extensive evaluations on sub-Saharan Africa and pediatric glioma datasets show that SmaRT consistently outperforms state-of-the-art methods, with notable gains in Dice accuracy and boundary precision. Overall, SmaRT bridges the gap between algorithmic advances and equitable clinical applicability, supporting robust deployment of MRI-based neuro-oncology tools in diverse clinical environments. Our source code is available at https://github.com/baiyou1234/SmaRT.

Jinhee Kim, Taesung Kim, Taewoo Kim et al.

In pediatric orthodontics, accurate estimation of growth potential is essential for developing effective treatment strategies. Our research aims to predict this potential by identifying the growth peak and analyzing cervical vertebra morphology solely through lateral cephalometric radiographs. We accomplish this by comprehensively analyzing cervical vertebral maturation (CVM) features from these radiographs. This methodology provides clinicians with a reliable and efficient tool to determine the optimal timings for orthodontic interventions, ultimately enhancing patient outcomes. A crucial aspect of this approach is the meticulous annotation of keypoints on the cervical vertebrae, a task often challenged by its labor-intensive nature. To mitigate this, we introduce Attend-and-Refine Network (ARNet), a user-interactive, deep learning-based model designed to streamline the annotation process. ARNet features Interaction-guided recalibration network, which adaptively recalibrates image features in response to user feedback, coupled with a morphology-aware loss function that preserves the structural consistency of keypoints. This novel approach substantially reduces manual effort in keypoint identification, thereby enhancing the efficiency and accuracy of the process. Extensively validated across various datasets, ARNet demonstrates remarkable performance and exhibits wide-ranging applicability in medical imaging. In conclusion, our research offers an effective AI-assisted diagnostic tool for assessing growth potential in pediatric orthodontics, marking a significant advancement in the field.

Daniel Capellán-Martín, Abhijeet Parida, Zhifan Jiang et al.

Robust and generalizable segmentation of brain tumors on multi-parametric magnetic resonance imaging (MRI) remains difficult because tumor types differ widely. The BraTS 2025 Lighthouse Challenge benchmarks segmentation methods on diverse high-quality datasets of adult and pediatric tumors: multi-consortium international pediatric brain tumor segmentation (PED), preoperative meningioma tumor segmentation (MEN), meningioma radiotherapy segmentation (MEN-RT), and segmentation of pre- and post-treatment brain metastases (MET). We present a flexible, modular, and adaptable pipeline that improves segmentation performance by selecting and combining state-of-the-art models and applying tumor- and lesion-specific processing before and after training. Radiomic features extracted from MRI help detect tumor subtype, ensuring a more balanced training. Custom lesion-level performance metrics determine the influence of each model in the ensemble and optimize post-processing that further refines the predictions, enabling the workflow to tailor every step to each case. On the BraTS testing sets, our pipeline achieved performance comparable to top-ranked algorithms across multiple challenges. These findings confirm that custom lesion-aware processing and model selection yield robust segmentations yet without locking the method to a specific network architecture. Our method has the potential for quantitative tumor measurement in clinical practice, supporting diagnosis and prognosis.

Longhao Chen, Xingchen Zhou, Chao Yang et al.

Abstract Background Juvenile idiopathic arthritis (JIA) is a type of chronic childhood arthritis with complex pathogenesis. Immunological studies have shown that JIA is an acquired self-inflammatory disease, involving a variety of immune cells, and it is also affected by genetic and environmental susceptibility. However, the precise causative relationship between the phenotype of immune cells and JIA remains unclear to date. The objective of our study is to approach this inquiry from a genetic perspective, employing a method of genetic association analysis to ascertain the causal relationship between immune phenotypes and the onset of JIA. Methods In this study, a two-sample Mendelian randomization (MR) analysis was used to select single nucleotide polymorphisms (SNPs) significantly associated with immune cells as instrumental variables to analyze the bidirectional causal relationship between 731 immune cells and JIA. There were four types of immune features (median fluorescence intensity (MFI), relative cellular (RC), absolute cellular (AC), and morphological parameters (MP)). Finally, the heterogeneity and horizontal reproducibility of the results were verified by sensitivity analysis, which ensured more robust results. Results We found that CD3 on CM CD8br was causally associated with JIA at the level of 0.05 significant difference (95% CI = 0.630 ~ 0.847, P = 3.33 × 10−5, PFDR = 0.024). At the significance level of 0.20, two immunophenotypes were causally associated with JIA, namely: HLA DR on CD14+ CD16- monocyte (95% CI = 0.633 ~ 0.884, P = 6.83 × 10–4, PFDR = 0.16) and HLA DR on CD14+ monocyte (95% CI = 0.627 ~ 0.882, P = 6.9 × 10−4, PFDR = 0.16). Conclusion Our study assessed the causal effect of immune cells on JIA from a genetic perspective. These findings emphasize the complex and important role of immune cells in the pathogenesis of JIA and lay a foundation for further study of the pathogenesis of JIA.

Jiajie Jing, Yuan Ma, Ziwen Xie et al.

Acute lymphoblastic leukemia (ALL) is a prevalent malignancy affecting the hematopoietic system, encompassing both B-cell ALL (B-ALL) and T-cell ALL (T-ALL). T-ALL, characterized by the proliferation of T-cell progenitors in the bone marrow, presents significant treatment challenges, with patients often experiencing high relapse rates and poor long-term survival despite advances in chemotherapy and hematopoietic stem cell transplantation (HSCT). This review explores the pathogenesis and traditional treatment strategies of T-ALL, emphasizing the promising potential of chimeric antigen receptor (CAR) technology in overcoming current therapeutic limitations. CAR therapy, leveraging genetically modified immune cells to target leukemia-specific antigens, offers a novel and precise approach to T-ALL treatment. The review critically analyzes recent developments in CAR-T and CAR-NK cell therapies, their common targets, optimization strategies, clinical outcomes, and the associated challenges, providing a comprehensive overview of their clinical prospects in T-ALL treatment.

Rui-Yang Ju, Chun-Tse Chien, Chia-Min Lin et al.

Children often suffer wrist injuries in daily life, while fracture injuring radiologists usually need to analyze and interpret X-ray images before surgical treatment by surgeons. The development of deep learning has enabled neural network models to work as computer-assisted diagnosis (CAD) tools to help doctors and experts in diagnosis. Since the YOLOv8 models have obtained the satisfactory success in object detection tasks, it has been applied to fracture detection. The Global Context (GC) block effectively models the global context in a lightweight way, and incorporating it into YOLOv8 can greatly improve the model performance. This paper proposes the YOLOv8+GC model for fracture detection, which is an improved version of the YOLOv8 model with the GC block. Experimental results demonstrate that compared to the original YOLOv8 model, the proposed YOLOv8-GC model increases the mean average precision calculated at intersection over union threshold of 0.5 (mAP 50) from 63.58% to 66.32% on the GRAZPEDWRI-DX dataset, achieving the state-of-the-art (SOTA) level. The implementation code for this work is available on GitHub at https://github.com/RuiyangJu/YOLOv8_Global_Context_Fracture_Detection.

Petra Grahn, Tiina Remes, Reetta Kivisaari et al.

Abstract Background Childhood brain tumor (BT) survivors have an increased risk of treatment-related late effects, which can reduce health-related quality of life and increase morbidity. This study aimed to investigate lumbar disc degeneration in magnetic resonance imaging (MRI) in adult survivors of radiotherapy-treated childhood BT compared to age and sex-matched population controls. Methods In this cross-sectional comparative study, 127 survivors were identified from hospital registries. After a mean follow-up of 20.7 years (range 5–33.1), 67 survivors (mean age 28.4, range 16.2–43.5) were investigated with MRI and compared to 75 sex-matched population-based controls. Evaluated MRI phenotypes included Pfirrmann grading, , intervertebral disc protrusions, extrusions, and high-intensity-zone-lesions (HIZ). Groups were also compared for known risk factors of lumbar intervertebral disc (IVD) degeneration. Results Childhood BT survivors had higher Pfirrmann grades than controls at all lumbar levels (all p < 0.001). Lumbar disc protrusions at L4-5 (p = 0.02) and extrusions at L3-4 (p = 0.04), L4-5 (p = 0.004), and L5-S1 (p = 0.01) were significantly more common in the BT group compared to the control. The survivor cohort also had significantly more HIZ-lesons than the controls (n=13 and n=1, p=0.003). Age at diagnosis was associated with lower degree of IVD degeneration (p < 0.01). Blood pressure correlated with IVD degeneration (P < 0.05). Conclusions Signs of early disc degeneration related to tumor treatment can be seen in the IVDs of survivors. Disc degeneration was more severe in children treated in adolescence.

Qina Ye, Shaodan Sun, Jian Deng et al.

Abstract To investigate the gut microbiota distribution and its functions in children with avoidant/restrictive food intake disorder (ARFID). A total of 135 children were enrolled in the study, including 102 children with ARFID and 33 healthy children. Fecal samples were analyzed to explore differences in gut microbiota composition and diversity and functional differences between the ARFID and healthy control (HC) groups via 16S rDNA and metagenomic sequencing. The gut microbiota composition and diversity in children with ARFID were different from those in heathy children, but there is no difference in the composition and diversity of gut microbiota between children at the age of 3–6 and 7–12 with ARFID. At the phylum level, the most abundant microbes in the two groups identified by 16S rDNA and metagenomic sequencing were the same. At the genus level, the abundance of Bacteroides was higher in the ARFID group (P > 0.05); however, different from the result of 16SrDNA sequencing, metagenomic sequencing showed that the abundance of Bacteroides in the ARFID group was significantly higher than that in the HC group (P = 0.041). At the species level, Escherichia coli, Streptococcus thermophilus and Lachnospira eligens were the most abundant taxa in the ARFID group, and Prevotella copri, Bifidobacterium pseudocatenulatum, and Ruminococcus gnavus were the top three microbial taxa in the HC group; there were no statistically significant differences between the abundance of these microbial taxa in the two groups. LefSe analysis indicated a greater abundance of the order Enterobacterales and its corresponding family Enterobacteriaceae, the family Bacteroidaceae and corresponding genus Bacteroides, the species Bacteroides vulgatus in ARFID group, while the abundance of the phylum Actinobacteriota and its corresponding class Actinobacteria , the order Bifidobacteriales and corresponding family Bifidobacteriaceae, the genus Bifidobacterium were enriched in the HC group. There were no statistically significant differences in the Chao1, Shannon and Simpson indices between the Y1 and Y2 groups (P = 0.1, P = 0.06, P = 0.06). At the phylum level, Bacillota, Bacteroidota, Proteobacteria and Actinobacteriota were the most abundant taxa in both groups, but there were no statistically significant differences among the abundance of these bacteria (P = 0.958, P = 0.456, P = 0.473, P = 0.065). At the genus level, Faecalibacterium was more abundant in the Y2 group than in the Y1 group, and the difference was statistically significant (P = 0.037). The KEGG annotation results showed no significant difference in gut microbiota function between children with ARFID and healthy children; however, GT26 was significantly enriched in children with ARFID based on the CAZy database. The most abundant antibiotic resistance genes in the ARFID group were the vanT, tetQ, adeF, ermF genes, and the abundance of macrolide resistance genes in the ARFID group was significantly higher than that in the HC group (P = 0.041). Compared with healthy children, children with ARFID have a different distribution of the gut microbiota and functional genes. This indicates that the gut microbiome might play an important role in the pathogenesis of ARFID. Clinical trial registration: ChiCTR2300074759

Chaoyi Wu, Jiayu Lei, Qiaoyu Zheng et al.

Driven by the large foundation models, the development of artificial intelligence has witnessed tremendous progress lately, leading to a surge of general interest from the public. In this study, we aim to assess the performance of OpenAI's newest model, GPT-4V(ision), specifically in the realm of multimodal medical diagnosis. Our evaluation encompasses 17 human body systems, including Central Nervous System, Head and Neck, Cardiac, Chest, Hematology, Hepatobiliary, Gastrointestinal, Urogenital, Gynecology, Obstetrics, Breast, Musculoskeletal, Spine, Vascular, Oncology, Trauma, Pediatrics, with images taken from 8 modalities used in daily clinic routine, e.g., X-ray, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Positron Emission Tomography (PET), Digital Subtraction Angiography (DSA), Mammography, Ultrasound, and Pathology. We probe the GPT-4V's ability on multiple clinical tasks with or without patent history provided, including imaging modality and anatomy recognition, disease diagnosis, report generation, disease localisation. Our observation shows that, while GPT-4V demonstrates proficiency in distinguishing between medical image modalities and anatomy, it faces significant challenges in disease diagnosis and generating comprehensive reports. These findings underscore that while large multimodal models have made significant advancements in computer vision and natural language processing, it remains far from being used to effectively support real-world medical applications and clinical decision-making. All images used in this report can be found in https://github.com/chaoyi-wu/GPT-4V_Medical_Evaluation.

Jared D. Huling, Jennifer P. Lundine, Julie C. Leonard

This work is motivated by the need to accurately model a vector of responses related to pediatric functional status using administrative health data from inpatient rehabilitation visits. The components of the responses have known and structured interrelationships. To make use of these relationships in modeling, we develop a two-pronged regularization approach to borrow information across the responses. The first component of our approach encourages joint selection of the effects of each variable across possibly overlapping groups related responses and the second component encourages shrinkage of effects towards each other for related responses. As the responses in our motivating study are not normally-distributed, our approach does not rely on an assumption of multivariate normality of the responses. We show that with an adaptive version of our penalty, our approach results in the same asymptotic distribution of estimates as if we had known in advance which variables were non-zero and which variables have the same effects across some outcomes. We demonstrate the performance of our method in extensive numerical studies and in an application in the prediction of functional status of pediatric patients using administrative health data in a population of children with neurological injury or illness at a large children's hospital.