BACKGROUND: Congenital heart defects afflict ≈1% of all births worldwide. Although deep learning has shown significant promise in automating and improving adult echocardiography analysis, existing pediatric-based models are often limited to single tasks and specific echocardiographic views. To address this, we introduce EchoAI-Peds, a multitask deep learning model for pediatric echocardiography. Our model was developed using the most comprehensive set of pediatric labels to date and is designed to integrate information from multiple views simultaneously. METHODS: We trained a video-based vision transformer to simultaneously detect 28 congenital heart defects, structural and functional abnormalities, repairs, and interventions directly from complete pediatric echocardiography studies with multiple videos. Our model was developed using >700 000 videos derived from >12 000 studies performed at Stanford Medicine between 2014 and 2021. Specifically, our model was trained on 10 815 studies (median age, 8 [IQR 2–14] years; 45% female) and validated on 1294 studies (median age, 9 [IQR 2–15] years; 46% female). Model efficacy was tested on an internal held-out data set of 1336 studies from that same period. In addition, model generalizability was tested on a spatially and temporally distinct patient cohort at the Children’s Hospital of Philadelphia using 2121 studies performed between 2024 and 2025. RESULTS: Our model achieved macroaveraged area under the receiver operating characteristic curve (AUROC) values of 0.91 (95% CI, 0.90–0.92) on the internal test set (median age, 8 [IQR 2–14] years; 47% female) and AUROC values of 0.89 (95% CI, 0.88–0.90) on the external test set (median age, 6 [IQR 0.92–13] years; 44.4% female). Moreover, EchoAI-Peds significantly outperformed adult-based echocardiography foundation models trained on substantially larger data sets, including EchoCLIP (internal AUROC=0.58, 95% CI, 0.56–0.60; external AUROC=0.61, 95% CI, 0.59–0.62) and EchoPrime (internal AUROC=0.58, 95% CI, 0.57–0.61; external AUROC=0.60, 95% CI, 0.59–0.62). Finally, our model demonstrated robust performance across patient age, patient sex, and studies with varying numbers of videos. CONCLUSIONS: Our findings demonstrate the remarkable potential for multitask deep learning models to aid the interpretation of pediatric echocardiograms. In addition, our results underscore the need for models that are specifically tailored to pediatric populations.

Deep learning requires extensive data to achieve robust performance. However, data availability is often restricted in medicine due to patient privacy concerns. Synthetic data from medical image generative models present a possible solution to these challenges. Unfortunately, many models are often task-specific or trained on private datasets, limiting scalability and reproducibility. To address this, we introduce MediSyn: an open-access, generalist, text-guided latent diffusion model. MediSyn can generate images across 6 medical specialties and 10 imaging modalities, while being trained exclusively on public data. Our study provides several key contributions using MediSyn. First, we demonstrate that a generalist model can generate images of comparable quality to task-specific models across multiple specialties. Second, we show that this generalist approach is more computationally efficient than a suite of task-specific models. Third, we establish that a generalist model can produce realistic, text-aligned images across visually and medically distinct modalities. Fourth, we empirically demonstrate that these synthetic images are visually distinct from real images used in training. Finally, we show that a generalist model can produce images that improve classifier performance in data-limited settings across multiple specialties. Altogether, our findings highlight the potential of generalist image generative models to accelerate medical algorithmic research and development.

BACKGROUND: The vascular system is the largest organ in the body and underlies most chronic diseases, yet the molecular mechanisms that govern its plasticity remain poorly defined. METHODS: We applied single-cell proteomics in vascular disease, integrating it with single-cell transcriptomics to map protein regulation in healthy and Marfan syndrome aortas. RESULTS: This approach uncovered cell type–specific and cell state–specific proteins missed at the transcriptional level. Notably, we identified a decoupling of fibrillin-1 RNA and protein abundance, suggesting altered protein regulation as a potential contributor to aortic degeneration. Single-cell proteomics further resolved modulated smooth muscle cell states enriched for matrix effectors (AEBP1 [adipocyte enhancer binding protein 1], HTRA1 [high-temperature requirement A serine peptidase 1], FN1 [fibronectin 1]) and uniquely protein-level regulators (GLIPR2 [GLI pathogenesis–related 2], ITGB2 [integrin beta-2], and CD151 [cluster of differentiation 151 antigen]). CONCLUSIONS: Together, these findings establish the human single-cell proteomic atlas of the aorta, reveal altered FBN1 (fibrillin-1) protein regulation in both human and mouse Marfan syndrome, and position single-cell proteomics as a transformative framework for decoding vascular plasticity and identifying actionable effectors in disease.

The innate immune system responds to virulent pathogens, yet many pathogens manipulate host signaling pathways which should limit immune activation. To understand how the immune system overcomes pathogenic manipulation we study the intracellular bacterium Legionella pneumophila, the cause of the severe pneumonia Legionnaire’s disease. Legionella utilizes a type IV secretion system to inject bacterial proteins into the cytosol of infected cells. Several of these bacterial effector proteins inactivate host cell factors involved in protein translation. Despite Legionella’s ability to block host protein translation, inflammatory cytokines are still made against Legionella. It is unclear how infected cells can mount a cytokine response when host protein synthesis is blocked. Our studies demonstrate that infected cells do not produce many cytokines, such as IL-6 and TNF, critical for controlling Legionella infection. Instead, uninfected, bystander cells produce these cytokines. Infected host cells do produce IL-1 de novo. These data suggest that infected cells have mechanisms to overcome protein synthesis inhibition to produce IL-1 and that uninfected, bystander cells are important contributors to the immune response during infection with Legionella. This mechanism of immune activation has broad significance as many other bacterial pathogens manipulate immune cell signaling.

This narrative review summarizes the published literature on robotic coronary revascularization, encompassing robotic-assisted minimally invasive direct coronary artery bypass (RA-MIDCAB), totally endoscopic coronary artery bypass (TECAB), and hybrid coronary revascularization (HCR). These are distinct procedures that differ in operative access, conduit harvesting, use of cardiopulmonary bypass (CPB), and typical patient selection, and are treated as such throughout this review rather than as interchangeable techniques. A structured, non-systematic literature search of PubMed/MEDLINE and major cardiothoracic surgical journals was performed. Articles were selected by the authors based on relevance, methodological quality, and recency. This is explicitly a narrative rather than a systematic review, and no PRISMA methodology was applied. In observational series from experienced centers, RA-MIDCAB and TECAB are associated with low perioperative mortality (0–2.8%), high early left internal thoracic artery (LITA)-to-left anterior descending (LAD) graft patency (>95%), and shorter hospital stay than conventional sternotomy coronary artery bypass grafting (CABG) in selected patients. HCR extends robotic revascularization to selected patients with multivessel and left main disease. However, nearly all of this evidence derives from retrospective, single-center, or registry-based observational studies performed at high-volume expert centers in selected patient populations, and comparative claims against conventional CABG are vulnerable to selection bias and residual confounding. Robotic coronary revascularization is a feasible option that has been reproduced across a growing number of high-volume centers for carefully selected patients, but current evidence does not support broad claims of superiority over conventional CABG, and outcome data remain concentrated among a limited number of expert programs. Prospective multicenter and randomized data, standardized outcome definitions, and structured training pathways are needed to define its long-term role in coronary revascularization.