J Am Chem Soc. 2025 Mar 26. doi: 10.1021/jacs.4c14474. Online ahead of print.

ABSTRACT

Ribonucleic acids (RNA) are commonly formulated into lipid nanoparticles (LNP) for in vivo use, but challenges exist with systemic delivery and low in vivo expression efficiency. Inspired by ribonucleoprotein complexes in cells, we created synthetic ribonucleocarbohydrate (RNC) complexes based on cyclodextrin nanoparticles with ferrocenyl fluorocarbons capable of carrying mRNA and additional small-molecule drug payloads, facilitating lysosomal escape and immune stimulation all in the same nanoparticle. We show that this strategy results in highly efficient myeloid cell targeting (dendritic cells and MHC expressing macrophages) and protein expression following systemic administration. The RNC platform should have broad applications in vaccine development, immunosuppression, and immunostimulation for various diseases.

PMID:40135499 | DOI:10.1021/jacs.4c14474

We have posted an Implementation of New Initiatives and Policies page on the NIH Grants & Funding Website to pull together the latest information on recent and upcoming changes that impact applications and grants administration.

Page Highlights

• NIH Grants and Funding Information Status. Keep up to date on how NIH grants and funding information is evolving as we align with new agency priorities (e.g., status of communications, funding opportunities, application guidance, and more).
• Upcoming Changes. Get the latest status on in-progress initiatives like our adoption of Biographical Sketch and Current and Pending (Other) Support.
• Recent Changes. Learn about key initiatives implemented in 2024 that culminated in numerous changes to grant application content and review for due dates on or after January 25, 2025 and any adjustments made to those initiatives.

Bookmark the link, share it widely, and visit regularly for updated information.

NIH has been hard at work over the last two years developing guidance and training to help prepare the community for the implementation of the simplified review framework. That framework went into effect on January 25, 2025 and applies to applications for most research project grants.

We are often asked how the new framework will affect the preparation of applications. The simplified review framework is a new way of reviewing the same research strategies you’ve always developed, which means neither the components nor the structure of your application are expected to change.

As always, applicants should be responsive to the Notice of Funding Opportunity (NOFO) that they are applying to, especially Sections IV. Application and Submission Information and V.  Application Review Information.

Other Information

• Applicant Guidance for Simplifying the Review Framework

• How to Apply – Application Guide

• Recent Changes Implemented for Due Dates on or after January 25, 2025

• NIH All About Grants Podcast: Simplified Review Framework

Pediatr Blood Cancer. 2025 Jun;72(6):e31667. doi: 10.1002/pbc.31667. Epub 2025 Mar 25.

ABSTRACT

BACKGROUND: Although research has improved the prognosis of childhood cancer, many challenges remain, especially for high-risk, recurrent, and rare cancers. The recognition that diverse cancer types may share molecular alterations that can be therapeutically targeted has stimulated "precision medicine" approaches in research. Understanding parent and patient interest in genomic-derived therapeutic options in the clinical setting is limited and offers a potential for improved care.

METHODS: A qualitative study was conducted to explore how young adult (YA) patients and parents of children and adolescent patients regard targeted therapy options. These patients had high-risk, recurrent, or rare cancer for which there was no known curative therapy and were enrolled in a study evaluating investigational genomic tumor profiling. Clinical data were retrieved from medical records, and interviews were conducted.

RESULTS: Seventeen participants were interviewed (11 parents and six YA patients). Six themes emerged: (i) Genomic Understanding, (ii) Partnerships in Decision-Making, (iii) Connection with and Trust in Providers, (iv) Quality-of-Life (QOL) Considerations, (v) Understanding of Prognosis, and (vi) Nurturing Hope. Treatment decision-making is complex and depends on the connection (with providers and others), understanding (of prognosis, genomic literacy), care goals (QOL considerations), and planning for the future (hope).

CONCLUSIONS: Participants favored partnership in decision-making, expressed trust in their providers, and recognized the value of research. Engaging parents and YA patients in the planning of precision medicine translational research may be a path to designing an integrated research and care model that maximizes translational research implemented in real time, leading to improved outcomes.

PMID:40130678 | DOI:10.1002/pbc.31667

Int J Pharm. 2025 Mar 23:125499. doi: 10.1016/j.ijpharm.2025.125499. Online ahead of print.

ABSTRACT

Bio-inspired nanotopographical carriers have emerged as innovative cancer therapy strategies, mimicking natural processes to enhance targeted delivery and reduce systemic toxicity. This study presents the development of virus-like mesoporous silica nanoparticles (MSN) as a delivery platform for repurposed metformin (MTF) in a topical multi-stimuli responsive system for melanoma treatment. Metformin-loaded virus-like MSN (MTF-MSN) were fabricated and incorporated into a thermo-responsive gelling system. The particles were evaluated for morphology, zeta potential (ZP), particle size (PS), entrapment efficiency (EE%), Fourier-transform infrared (FT-IR) spectroscopy, MTT cytotoxicity assay, in vitro release, and in a melanoma in vivo model. The particles exhibited a spherical morphology, a zeta potential of +31.9 ± 1.45 mV, and a particle size of 197 ± 3.47 nm, ideal for skin penetration. MTF-MSN demonstrated significant antiproliferative activity against melanoma A375 cells, with lower IC50 values (192 μg/mL) compared to free MTF (>300 μg/mL). Sustained, pH-sensitive MTF release was observed over 48 h at pH 7.4 and 6 h at pH 5.5. In vivo studies showed enhanced anti-cancer efficacy of MTF-MSN, evidenced by elevated caspase-3 and Neurofibromin Type-1 (NF-1) levels, along with suppressed angiogenesis markers VEGF and NRAS. The MTF-MSN-treated group exhibited superior outcomes compared to free MTF and controls (p < 0.05). This innovative bio-inspired MTF-MSN hydrogel system optimizes MTF delivery for melanoma therapy, pioneering advancements in drug repurposing and nano-oncology.

PMID:40132769 | DOI:10.1016/j.ijpharm.2025.125499

J Cyst Fibros. 2025 Mar 24:S1569-1993(25)00077-3. doi: 10.1016/j.jcf.2025.03.009. Online ahead of print.

ABSTRACT

BACKGROUND: We aimed to build a cohort of Maternal-Cystic Fibrosis (CF) fetal dyads treated in utero with Variant Specific Therapy (VST), to assess the efficacy on Meconium Ileus (MI) and potential adverse effects of treatment.

METHODS: Dyads were included if the foetus had a genetic diagnosis of CF and carried at least one variant responsive to VST. Standardized assessment included pre-VST Magnetic Resonance Imaging (MRI), repeated ultrasound (US), and VST drug concentrations in cord blood, maternal and infant plasma.

RESULTS: We enrolled 13 dyads. One withdrew from the study. VST therapies (Elexacaftor (ELX)/Tezacaftor (TEZ)/Ivacaftor(IVA) (ETI) n = 11, ivacaftor (IVA) n = 1) were administered to the pregnant women between 19 and 36 weeks' of gestation for a median[IQR] of 35[55] days, either as a curative indication of MI (n = 8) or as a tertiary prevention of fetal CF-related intestinal symptoms (n = 4). One foetus experienced increased bowel dilatation after ETI introduction. MRI revealed intestinal atresia. One dyad received only 2 doses. In the other 6 cases, resolution of MI was observed within 14[10] days of ETI. Fetal development and neonatal tolerance were excellent. Fecal elastase at birth was always below 200 ng/g even in the ETI breast-fed infant. Cord-to-maternal concentration yielded median ratios of 0.40 for ELX, 0.54 for IVA and 1.59 for TEZ.

CONCLUSION: ETI administration from the third trimester of pregnancy enables MI resolution. Trans-placental transfer is high. Fetal tolerance at ETI initiation needs to be monitored by a standardized assessment.

PMID:40133101 | DOI:10.1016/j.jcf.2025.03.009

BMJ Open. 2025 Mar 25;15(3):e091357. doi: 10.1136/bmjopen-2024-091357.

ABSTRACT

INTRODUCTION: Target trial emulation is a framework for evaluating the effects of treatments using observational data. The trial emulation approach involves specifying key elements of a protocol for a target trial (a randomised controlled trial designed to address the question of interest) and then describing how best to emulate the trial using observational data. Recent years have seen an uptake of target trial emulation in several disease areas, although there are limited examples in cystic fibrosis (CF). This protocol describes a study which aims to assess the applicability of target trial emulation in CF. We aim to emulate an existing trial in CF and assess to what extent the results from the trial can be replicated using registry data.

METHODS AND ANALYSIS: The target trial is a published randomised controlled trial which found evidence for beneficial effects of azithromycin use on lung function in young adults with CF. Two emulated trials are planned: one using data from the UK CF Registry and one using data from the US CF Registry. The inclusion and exclusion criteria, treatment and outcome definitions, follow-up period, and estimand of interest are all designed to match the published trial as closely as possible. The analysis step of the trial emulations will use causal inference methods to control for confounding. Results obtained in the emulated trials using registry data will be compared with those from the target trial.

ETHICS AND DISSEMINATION: Ethical approval has been granted by the London School of Hygiene and Tropical Medicine Ethics Committee (Ref: 29609). This study has also been approved by the UK CF Registry Research Committee and the North Star Review Board. The results of this study will be published in a peer-reviewed journal and presented at relevant scientific conferences.

PMID:40132841 | DOI:10.1136/bmjopen-2024-091357

Respir Med. 2025 Mar 23:108052. doi: 10.1016/j.rmed.2025.108052. Online ahead of print.

ABSTRACT

PURPOSE: Nontuberculous mycobacterial pulmonary disease (NTMPD) is a chronic, often progressive condition associated with a significant symptom burden and increased mortality. Goals of NTMPD treatment include microbiological eradication, symptom reduction, improved quality of life (QoL), and preventing disease progression. Antibiotics are used to reduce microbial burden and cultures of sputum are used to guide treatment. However, it is unclear whether achieving culture-negative status (often called "culture conversion") is associated with improved clinical outcomes. Studies use a variety of measures including symptom burden, radiological status, lung function, six-minute walk test distance, QoL assessments, and mortality to evaluate clinical outcomes related to changes in how patients feel, function, and survive. There is strong interest in more clearly understanding which clinical benefits may be associated with culture conversion. As NTMPD can cause sustained structural lung damage and declines in long-term pulmonary function, it is important to have clear evidence if prevention of these morbidities is associated with culture conversion.

METHODS: This targeted literature review summarizes the published evidence regarding associations between sputum culture conversion and clinical outcomes in patients with NTMPD. Identified studies used varying definitions of culture conversion and treatment success, making interpretation of outcomes across studies challenging.

RESULTS: Although some studies suggest an association between culture conversion and aspects of clinical improvement, overall, there are currently few high-quality studies supporting a link.

CONCLUSION: Further clarification of the relationship between culture conversion and clinical outcomes would be helpful in improving clinical monitoring and therapeutic decision-making during the treatment of patients with NTMPD.

PMID:40132751 | DOI:10.1016/j.rmed.2025.108052

Trends Biochem Sci. 2025 Mar 24:S0968-0004(25)00048-9. doi: 10.1016/j.tibs.2025.02.006. Online ahead of print.

ABSTRACT

Deaminases belonging to the AID/APOBEC family are known as ssDNA and mRNA mutators involved in innate/adaptive immunity, mRNA editing, genome stabilization by restricting retrotransposons, and carcinogenesis. Recent studies suggest that the repertoire of AID/APOBEC targets is more diverse than previously thought and imply a broader biological impact of these proteins.

PMID:40133172 | DOI:10.1016/j.tibs.2025.02.006

Dev Cell. 2025 Mar 18:S1534-5807(25)00120-0. doi: 10.1016/j.devcel.2025.02.015. Online ahead of print.

ABSTRACT

A population of putative mesendoderm progenitors that can contribute cellular descendants to both mesoderm and endoderm lineages is identified in the gastrulating mouse embryo. These progenitor cells are localized to the posterior epiblast, primitive streak, and nascent mesoderm of mid-streak- (E7.0) to late-streak-stage (E7.5) embryos. Lineage tracing in vivo identified that putative mesendoderm progenitors contribute descendants to the definitive endoderm and the axial mesendoderm of E7.75 embryos and to the endoderm of the foregut and hindgut of the E8.5-8.75 embryos. Differentiation of mouse epiblast stem cells identified that the choice between endoderm and mesoderm cell fates depends on the timing of Mixl1 activation upon exit from pluripotency. The knowledge gained on the spatiotemporal distribution of mesendoderm progenitors and the molecular drivers underpinning the divergence of cell lineages in these progenitors enriches our mechanistic understanding of the allocation of the tissue progenitors to germ layer derivatives in early development.

PMID:40132585 | DOI:10.1016/j.devcel.2025.02.015

Cell Rep Methods. 2025 Mar 24;5(3):101009. doi: 10.1016/j.crmeth.2025.101009.

ABSTRACT

We demonstrate a cybernetic approach to control the composition of a P. putida and E. coli co-culture that does not rely on genetic engineering to interface cells with computers. We first show how composition information can be extracted from different bioreactor measurements and then combined with a system model using an extended Kalman filter to generate accurate estimates of a noisy system. We then demonstrate that adjusting the culture temperature can drive the composition due to the species' different optimal temperatures. Using a proportional-integral control algorithm, we are able to track dynamic references with real-time noise rejection and independence from starting conditions such as inoculation ratio. We stabilize the co-culture for 7 days (∼250 generations) with the experiment ending before the cells could adapt out of the control. This cybernetic framework is broadly applicable, with different microbes' unique characteristics enabling robust control over diverse co-cultures.

PMID:40132542 | DOI:10.1016/j.crmeth.2025.101009

Nucleic Acids Res. 2025 Mar 20;53(6):gkaf207. doi: 10.1093/nar/gkaf207.

ABSTRACT

Human health depends on perplexing defensive cellular responses against microbial pathogens like Viruses. Despite the major effort undertaken, the (epi)genomic mechanisms that human cells utilize to tailor defensive gene expression programs against microbial attacks have remained inadequately understood, mainly due to a significant lack of recording of the in vivo functional cis-regulatory modules (CRMs) of the human genome. Here, we introduce the virus-responsive fate of the human (epi)genome as characterized in naïve and infected cells by functional genomics, computational biology, DNA evolution, and DNA Grammar and Syntax investigations. We discovered that multitudes of novel functional virus-responsive CRMs (vrCRMs) compose typical enhancers (tEs), super-enhancers (SEs), repetitive-DNA enhancers (rDEs), and stand-alone functional genomic stretches that grant human cells regulatory underpinnings for layering basal immunity and eliminating illogical/harmful defensive responses under homeostasis, yet stimulating virus-responsive genes and transposable elements (TEs) upon infection. Moreover, extensive epigenomic reprogramming of previously unknown SE landscapes marks the transition from naïve to antiviral human cell states and involves the functions of the antimicrobial transcription factors (TFs), including interferon response factor 3 (IRF3) and nuclear factor-κB (NF-κB), as well as coactivators and transcriptional apparatus, along with intensive modifications/alterations in histone marks and chromatin accessibility. Considering the polyphyletic evolutionary fingerprints of the composite DNA sequences of the vrCRMs assessed by TFs-STARR-seq, ranging from the animal to microbial kingdoms, the conserved features of antimicrobial TFs and chromatin complexes, and their pluripotent stimulus-induced activation, these findings shed light on how mammalian (epi)genomes evolved their functions to interpret the exogenous stress inflicted and program defensive transcriptional responses against microbial agents. Crucially, many known human short variants, e.g. single-nucleotide polymorphisms (SNPs), insertions, deletions etc., and quantitative trait loci (QTLs) linked to autoimmune diseases, such as multiple sclerosis (MS), systemic lupus erythematosus (SLE), Crohn's disease (CD) etc., were mapped within or vastly proximal (±2.5 kb) to the novel in vivo functional SEs and vrCRMs discovered, thus underscoring the impact of their (mal)functions on human physiology and disease development. Hence, we delved into the virus-responsive fate of the human (epi)genome and illuminated its architecture, function, evolutionary origins, and its significance for cellular homeostasis. These results allow us to chart the "Human hyper-Atlas of virus-infection", an integrated "molecular in silico" encyclopedia situated in the UCSC Genome Browser that benefits our mechanistic understanding of human infectious/(auto)immune diseases development and can facilitate the generation of in vivo preclinical animal models, drug design, and evolution of therapeutic applications.

PMID:40131776 | DOI:10.1093/nar/gkaf207

JCI Insight. 2025 Mar 25:e190655. doi: 10.1172/jci.insight.190655. Online ahead of print.

ABSTRACT

BACKGROUND: The graft-vs-leukemia (GVL) effect contributes to the efficacy of allogeneic stem cell transplantation (alloSCT). However, relapse, indicative of GVL failure, is the greatest single cause of treatment failure. Based on preclinical data showing that IFN-γ is important to sensitize myeloblasts to alloreactive T cells, we performed a phase I trial of IFN-γ combined with donor leukocyte infusions (DLI) in myeloblastic malignancies that relapsed post-HLA-matched alloSCT.

METHODS: Patients with relapsed acute myeloid leukemia or myelodysplastic syndrome after alloSCT were eligible. Patients self-administered IFN-γ for 4 weeks (cohort 1) or 1 week (cohort 2), followed by DLI and concurrent IFN-γ for a total of 12 weeks. Bone marrow samples were analyzed by single-cell RNA sequencing (scRNAseq) to assess in vivo responses to IFN-γ by malignant myeloblasts.

RESULTS: IFN-γ monotherapy was well tolerated by all subjects (n=7). Treatment-related toxicities after DLI included: grade I-II graft-versus-host disease (n=5), immune effector cell-associated neurotoxicity syndrome (n=2), and idiopathic pulmonary syndrome (n=1), all of which resolved with corticosteroids. Four of 6 DLI recipients achieved minimal residual disease-negative complete remissions and full donor hematopoietic recovery. Median overall survival was 579 days (range, 97-906) in responders. ScRNAseq confirmed in vivo activation of IFN-γ response pathway in hematopoietic stem cell-like or myeloid progenitor cells after IFN-γ in analyzed samples.

CONCLUSIONS: IFN-γ was safe and well tolerated in this phase I study of IFN-γ for relapsed AML/MDS post-alloSCT, with a promising efficacy signal when combined with DLI. Larger studies are needed to formally test the efficacy of this approach.TRIAL RESGISTRATION.

CLINICALTRIALS: gov NCT04628338.

FUNDING: The research was supported by The UPMC Hillman Cancer Center Cancer Immunology and Immunotherapy Program (CIIP) Pilot Award and Cure Within Reach: Drug Repurposing Clinical Trials to Impact Blood Cancers. Recombinant IFN-gamma (Actimmune®) was donated by Horizon Therapeutics.

PMID:40131369 | DOI:10.1172/jci.insight.190655

Am J Respir Crit Care Med. 2025 Mar 25. doi: 10.1164/rccm.202503-0579ED. Online ahead of print.

NO ABSTRACT

PMID:40132170 | DOI:10.1164/rccm.202503-0579ED

JCI Insight. 2025 Mar 25:e191098. doi: 10.1172/jci.insight.191098. Online ahead of print.

ABSTRACT

Type 2 inflammatory diseases are common in cystic fibrosis (CF) including asthma, sinusitis, and allergic bronchopulmonary aspergillosis. CD4+ T helper 2 (Th2) cells promote these diseases through secretion of IL-4, IL-5, and IL-13. Whether the cystic fibrosis transmembrane conductance regulator (CFTR), the mutated protein in CF, has a direct effect on Th2 development is unknown. Using murine models of CFTR deficiency and human CD4+ T cells, we show CD4+ T cells expressed Cftr transcript and CFTR protein following activation. Loss of T cell CFTR expression increased Th2 cytokine production compared to control cells. Mice with CFTR-deficient T cells developed increased allergic airway disease to Alternaria alternata extract compared to control mice. Culture of CFTR-deficient Th2 cells demonstrated increased IL-4Rα expression and increased sensitivity to IL-4 with greater induction of GATA3 and IL-13 compared to control Th2 cell cultures. The CFTR potentiator ivacaftor reduced allergic inflammation and type 2 cytokine secretion in bronchoalveolar lavage of "humanized" CFTR mice following Alternaria alternata extract challenge and decreased Th2 development in human T cell culture. Together, these data support a direct role of CFTR in regulating T cell sensitivity to IL-4 and demonstrate a potential CFTR-specific therapeutic strategy for Th2 cell-mediated allergic disease.

PMID:40131363 | DOI:10.1172/jci.insight.191098

PLoS One. 2025 Mar 25;20(3):e0320219. doi: 10.1371/journal.pone.0320219. eCollection 2025.

ABSTRACT

Neuroinflammation is a key feature of human neurodisease including neuropathy and neurodegenerative disease and is driven by the activation microglia, immune cells of the nervous system. During activation microglia release pro-inflammatory cytokines as well as reactive oxygen species (ROS) that can drive local neuronal and glial damage. Phytocannabinoids are an important class of naturally occurring compounds found in the cannabis plant (Cannabis sativa) that interact with the body's endocannabinoid receptor system. Cannabidiol (CBD) is a prototype phytocannabinoid with anti-inflammatory properties observed in cells and animal models. We measured ROS in human microglia (HMC3) cells using CellROX, a fluorescent dynamic ROS indicator. We tested the effect of CBD on ROS level in the presence of three known immune activators: lipopolysaccharide (LPS), amyloid beta (Aβ42), and human immunodeficiency virus (HIV) glycoprotein (GP120). Confocal microscopy images within microglia were coupled to a deep learning model using a convolutional neural network (CNN) to predict ROS responses. Our study demonstrates a deep learning platform that can be used in the assessment of CBD effect in immune cells using ROS image measure.

PMID:40131976 | DOI:10.1371/journal.pone.0320219

PLoS One. 2025 Mar 25;20(3):e0319858. doi: 10.1371/journal.pone.0319858. eCollection 2025.

ABSTRACT

This study aims to explore the integration of the Faster R-CNN (Region-based Convolutional Neural Network) algorithm from deep learning into the MobileNet v2 architecture, within the context of enterprises aiming for carbon neutrality in their development process. The experiment develops a marine oil condition monitoring and classification model based on the fusion of MobileNet v2 and Faster R-CNN algorithms. This model utilizes the MobileNet v2 network to extract rich feature information from input images and combines the Faster R-CNN algorithm to rapidly and accurately generate candidate regions for oil condition monitoring, followed by detailed feature fusion and classification of these regions. The performance of the model is evaluated through experimental assessments. The results demonstrate that the average loss value of the proposed model is approximately 0.45. Moreover, the recognition accuracy of the model for oil condition on the training and testing sets reaches 90.51% and 93.08%, respectively, while the accuracy of other algorithms remains below 90%. Thus, the model constructed in this study exhibits excellent performance in terms of loss value and recognition accuracy, providing reliable technical support for offshore oil monitoring and contributing to the promotion of sustainable utilization and conservation of marine resources.

PMID:40131882 | DOI:10.1371/journal.pone.0319858

IEEE Trans Pattern Anal Mach Intell. 2025 Mar 25;PP. doi: 10.1109/TPAMI.2025.3554560. Online ahead of print.

ABSTRACT

Recent advancements in model pruning have focused on developing new algorithms and improving upon benchmarks. However, the practical application of these algorithms across various models and platforms remains a significant challenge. To address this challenge, we propose ONNXPruner, a versatile pruning adapter designed for the ONNX format models. ONNXPruner streamlines the adaptation process across diverse deep learning frameworks and hardware platforms. A novel aspect of ONNXPruner is its use of node association trees, which automatically adapt to various model architectures. These trees clarify the structural relationships between nodes, guiding the pruning process, particularly highlighting the impact on interconnected nodes. Furthermore, we introduce a tree-level evaluation method. By leveraging node association trees, this method allows for a comprehensive analysis beyond traditional single-node evaluations, enhancing pruning performance without the need for extra operations. Experiments across multiple models and datasets confirm ONNXPruner's strong adaptability and increased efficacy. Our work aims to advance the practical application of model pruning.

PMID:40131753 | DOI:10.1109/TPAMI.2025.3554560

IEEE Trans Cybern. 2025 Mar 25;PP. doi: 10.1109/TCYB.2025.3549583. Online ahead of print.

ABSTRACT

The neural activity patterns of localized brain regions are crucial for recognizing brain intentions. However, existing electroencephalogram (EEG) decoding models, especially those based on deep learning, predominantly focus on global spatial features, neglecting valuable local information, potentially leading to suboptimal performance. Therefore, this study proposed a dynamic hierarchical convolutional attention network (DH-CAN) that comprehensively learned discriminative information from both global and local spatial domains, as well as from time-frequency domains in EEG signals. Specifically, a multiscale convolutional block was designed to dynamically capture time-frequency information. The channels of EEG signals were mapped to different brain regions based on motor imagery neural activity patterns. The spatial features, both global and local, were then hierarchically extracted to fully exploit the discriminative information. Furthermore, regional connectivity was established using a graph attention network, incorporating it into the local spatial features. Particularly, this study shared network parameters between symmetrical brain regions to better capture asymmetrical motor imagery patterns. Finally, the learned multilevel features were integrated through a high-level fusion layer. Extensive experimental results on two datasets demonstrated that the proposed model performed excellently across multiple evaluation metrics, exceeding existing benchmark methods. These findings suggested that the proposed model offered a novel perspective for EEG decoding research.

PMID:40131750 | DOI:10.1109/TCYB.2025.3549583

IEEE J Biomed Health Inform. 2025 Mar 25;PP. doi: 10.1109/JBHI.2025.3554495. Online ahead of print.

ABSTRACT

Training deep learning models for photoplethysmography(PPG)-based cuff-less blood pressure estimation often requires a substantial amount of labeled data collected through sophisticated medical instruments, posing significant challenges in practical applications. To address this issue, we propose Physiological Knowledge-Aware Contrastive Learning (PhysCL), a novel approach designed to reduce the dependence on labeled PPG data while improving blood pressure estimation accuracy. Specifically, PhysCL tackles the semantic consistency problem in contrastive learning by introducing a knowledge-aware augmentation bank, which generates positive physiological signal pairs using knowledge-based constraints during the contrastive pair generation. Additionally, we propose a contrastive feature reconstruction method to enhance feature diversity and prevent model collapse through feature re-sampling and re-weighting. We evaluate PhysCL on data from 106 subjects across the MIMIC III, MIMIC IV, and UQVS datasets under cross-dataset validation settings, comparing it against state-of-the-art contrastive learning methods and blood pressure estimation models. PhysCL achieves an average mean absolute error of 9.5/5.9 mmHg (systolic/diastolic) across the three datasets, using only 2% labeled data combined with 98% unlabeled data for pre-training and 5 samples for personalization, which represents a 6.2%/4.3% improvement, respectively, over the current best supervised methods. The ablation study provides further convincing evidence that the unlabeled data can be utilized to improve the existing cuff-less blood pressure estimation models and shed light on unsupervised contrastive learning for physiological signals.

PMID:40131744 | DOI:10.1109/JBHI.2025.3554495