Neonatal hypoxia ischaemia (HI) affects 1-3 per 1,000 live births, is a major cause of infant mortality and morbidity, and leads to adverse long-term neurological outcomes, while reliable biomarkers are scarce. Extracellular vesicles (EVs) are small membrane vesicles released from cells and play key roles in cellular communication through the transfer of diverse cargoes, including proteins, and can be isolated from various body fluids. Here, we developed a new non-invasive method of biofluid-EV profiling, isolating EVs from eye lavage. Our data demonstrate that in a neonatal HI mouse model of mild and severe insults, significant differences are found in EV eye lavage signatures. We identified increased EV numbers and modifications in EV size profiles and EV's proteomic cargo signatures in eye lavage from HI animals compared to controls. A protein-protein interaction network analysis of the EV proteome cargoes identified enrichment in Gene Ontology and Kyoto Encyclopaedia of Genes and Genomes (KEGG) pathways in the HI groups associated with various homeostatic and disease-related pathways. The specific changes in the mild HI group included pathways for ribosome biogenesis, translation, RNA processing, gene expression, blood coagulation, innate immunity, antioxidant activity, phospholipid binding, post-synapse, cell cortex, and HIF-1 signalling. The enriched pathways only associated with the EV proteome of the severe HI group included cytoskeleton organisation, peptide cross-linking, monosaccharide biosynthesis, peroxidase activity, extrinsic component of plasma membrane, the GAIT complex, mast cell granulation, ruffle, and sealing of the nuclear envelope by the endosomal sorting complex required for transport III. Here, we report a new non-invasive method using eye lavage EV signatures to identify changes in response to HI. Our results highlight eye lavage EVs as potential clinical biomarkers for predicting changes that occur in the brain and eye due to different neonatal HI injury severities.
Background and objective: Calcific obstruction of the pulmonary conduit is a late complication of surgical implantation of a homograft in congenital patients. Percutaneous pulmonary valve implantation (PPVI) is an effective alternative to surgical repair. However, this procedure is affected by several complications, with coronary artery (CA) compression being one of the most severe. High-fidelity finite element (FE) models can provide accurate predictions but are too computationally expensive for routine use, whereas simplified models sacrifice mechanical fidelity. This study proposes a novel FE-based framework to investigate conduit pre-stenting feasibility, while aiming to balance computational efficiency with predictive accuracy within clinically relevant timelines.
Methods: A semi-automated pipeline was developed, requiring manual input only for the segmentation of computed tomography (CT), virtual stent positioning, and simulation launch. Patient-specific geometries were meshed and processed through an automated in-house script, generating ready-to-run Abaqus input files. A multifactorial CA compression risk index was introduced, integrating baseline and post-expansion distances between the pulmonary artery and CA, and their changes during the procedure. The FE simulation of the pre-stenting procedure was tested on 10 PPVI candidates, simulating CP-stent implantation. Simulation accuracy was assessed against fluoroscopy-derived stent diameters.
Results: The full simulation process required less than 10 h per case, with minimal operator workload. FE-predicted stent configuration showed strong agreement with fluoroscopic measurements ( = 0.87), with a mean absolute error of 3.5 4.4%. Accuracy was highest in patients with calcific volumes <0.8 (error <0.5 mm). CA compression index identified 2 high-risk, 2 moderate-risk, and 6 negligible-risk patients. Peri-procedural fluoroscopy was not available for one negligible-risk patient; it excluded CA compression for the remaining negligible-risk patients (true negatives), for all moderate-risk patients, and for one high-risk patient (false positive); it highlighted CA compression for the remaining high-risk patient (true positive).
Conclusions: The proposed FE simulation framework enables patient-specific prediction of stent configuration and CA compression risk within clinically compatible timelines. The balanced trade-off between mechanical fidelity and computational efficiency supports its potential integration into pre-procedural planning of conduit pre-stenting and PPVI.
Preterm infants, particularly those born before 32 weeks of gestation, frequently face challenges in achieving full enteral nutrition due to underdeveloped sucking-swallowing-breathing coordination. Conventional feeding methods, such as the use of indwelling nasogastric tubes, overlook the importance of sucking activity, which is essential for the development of gastrointestinal motility and the secretion of digestive enzymes. To address this issue, we have developed a sucking-rewarded automatic feeding device specifically designed for preterm infants. The device features a specialized pacifier that detects sucking activity and triggers the delivery of a predetermined amount of milk into the stomach via a gastric tube. In addition to promoting sucking-induced satiety, the device continuously monitors sucking waveforms to assess infants' viability and sucking maturity. In a clinical pilot study involving 25 preterm infants, those fed with the device demonstrated a significant increase in intestinal oxygen saturation compared with conventional gavage feeding (p < 0.05). Complementary experiments in 12 newborn beagle puppies showed faster gastric emptying rates (p < 0.01) and elevated gastrointestinal hormone levels (p < 0.05) when using the device. These findings highlight the clinical potential of the proposed device in improving feeding safety, efficiency, and developmental outcomes in preterm infants, and warrant further large-scale clinical trials to validate its long-term efficacy and integration into neonatal care.
Introduction: Frailty, a multidimensional syndrome of reduced physiologic reserve, is associated with poorer outcomes following allogeneic hematopoietic cell transplantation (alloHCT), even among younger adults. This pilot study explores whether wearable sensor data reflecting physical activity and sleep are associated with pre-transplant frailty status in patients undergoing myeloablative alloHCT.
Methods: Adults undergoing first myeloablative alloHCT at the University of Minnesota from June 2022 to January 2023 were enrolled and given Fitbit® Sense devices. Frailty was assessed pre-transplant using Fried Phenotype criteria. Activity and sleep data were collected from hospital admission to day +30 post-transplant. Descriptive and inferential statistics assessed differences across frailty phenotypes.
Results: Nine patients were included: 2 not frail, 5 pre-frail, and 2 frail. Not frail patients demonstrated significantly higher daily steps and active minutes, and lower sedentary time compared to pre-frail and frail groups (all p < 0.01). Frail individuals had significantly reduced deep and REM sleep. The nadir for sleep and peak in sedentary behavior occurred around day +15 post-transplant.
Conclusion: Pre-transplant frailty was associated with decreased physical activity and less restorative sleep during the peri-transplant period. These findings support further study of wearable data to guide personalized supportive care strategies in alloHCT recipients.
Background: Gastric cancer remains a major cause of cancer-related morbidity and mortality. Despite advances in surgical and perioperative care, prolonged hospitalization continues to strain healthcare systems. Predicting postoperative length of stay (LOS) could support personalized care and efficient resource allocation. Japan's nationwide Diagnosis Procedure Combination (DPC) database provides real-world data for large-scale analysis, but no study has applied machine learning to predict LOS after gastrectomy.
Methods: This retrospective study included 26,097 patients who underwent gastrectomy between 2017 and 2022 at 472 hospitals in Japan. Using XGBoost, we developed a predictive model based on 1,433 admission-time variables extracted from the DPC database. Model performance was evaluated using Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE) in a five-fold cross-validation. SHAP values were used to interpret feature importance.
Results: The final model achieved an RMSE of 3.74 and MAE of 2.82 days. Key predictors of LOS included surgical procedure (laparoscopic distal gastrectomy and open total gastrectomy), designated cancer hospital, hospital size, peritoneal dissemination, and admission ADL score. SHAP analysis revealed that Laparoscopic distal gastrectomy and higher hospital volume were associated with shorter LOS, while open total gastrectomy was associated with longer LOS.
Conclusions: We developed a machine learning model that predicts postoperative length of stay with an error range of 2-4 days using admission data. This proof-of-concept study demonstrates the feasibility of predicting length of stay from admission data, showing that explainable AI can replicate intuitive patterns in surgical oncology while simultaneously identifying unexpected insights from administrative data. These findings highlight the clinical potential of explainable AI for perioperative workflow optimization.
The accurate detection of Alzheimer's disease (AD), a progressive and irreversible neurodegenerative disorder, remains a critical challenge in clinical neuroscience. The research aims to develop an advanced multimodal image fusion model for the accurate detection of AD using positron emission tomography (PET) and magnetic resonance imaging (MRI) techniques. The proposed method leverages structural MRI and functional 18-fluorodeoxyglucose PET (FDG-PET) information derived from the Alzheimer's Disease Neuroimaging Initiative (ADNI). After preprocessing, including Gaussian filtering, skull stripping, and intensity normalization, voxel-based morphometry (VBM) is applied to extract gray matter (GM) features relevant to AD progression. A GM mask generated from MRI is used to isolate corresponding metabolic activity in the PET scans. These features are then integrated using a mask-coding strategy to construct a unified representation that captures both anatomical and functional characteristics. For classification, the model introduces a Glowworm Swarm-Optimized Spatial Multimodal Attention-Enriched Convolutional Neural Network (GWS-SMAtt-ECNN), where the optimization enhances both feature selection and network parameter tuning. The Python was implemented, and the result demonstrates that the proposed multimodal image fusion strategy outperforms traditional unimodal and basic fusion approaches in terms of F1-score (94.22%), recall (96.73%), and accuracy (98.70%). These results highlight the therapeutic usefulness of the suggested improved fusion architecture in facilitating immediate and accurate AD detection by MRI and PET.
Endometrial cancer, accounting for over 90% of uterine malignancies, has experienced a significant global rise in incidence and mortality. Conventional therapies face limitations including fertility compromise, systemic toxicity, drug resistance, and poor outcomes in advanced/recurrent cases. Considering the unique physical and chemical properties of nanomaterials, the emerging drug delivery approaches based on nanomaterials are regarded as a promising pathway for enhanced therapeutic efficiency to combat endometrial cancer. Herein, this mini-review discusses emerging drug delivery approaches to overcome current treatment challenges. We classify common therapeutic nanomaterials into polymer-based nanocarriers, quantum dots, liposomes, and exosomes, analyzing their synthesis, mechanisms, and preclinical efficacy. Finally, scientific challenges and future perspectives for ongoing research in this field are presented.
[This corrects the article DOI: 10.3389/fmedt.2025.1695329.].
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：