Results
- Showing results for:
- Reset all filters
Search results
-
Conference paperSifaoui S, Angelini E, Toupin S, et al., 2027, , Pages: 451-460, ISSN: 0302-9743
Dense self-supervised learning (SSL) is a powerful paradigm for learning without annotations the local descriptors required to solve dense medical imaging tasks. We present Pix2Rep-v2, a framework for SSL of pixel- and voxel-level representations suitable for few-shot downstream applications. Pix2Rep-v2 addresses the main challenges of dense SSL by leveraging a redundancy reduction objective at the pixel-level with a principle of equivariance of dense representations, that scales efficiently to 3D or wide field-of-view applications. We evaluate our method on four datasets, across multiple tasks, multiple modalities and anatomical structures using multiple backbones in 2D and 3D, and under various data regimes. As an alternative to linear probing or full fine-tuning on the downstream task, we also propose an in-context variant, without downstream training, based on a dense prototype approach. Pix2Rep-v2 shows substantially higher data-efficiency in few-shot scenarios compared to fully supervised baselines, and is competitive with the state-of-the-art e.g.,+9.3 Dice points in one-shot segmentation on the M&Ms-2 dataset. Our code and pre-trained models are publicly available at https://github.com/BioMedTP/pix2rep-v2.
-
Conference paperLi Z, Zhao W, Yu K, et al., 2027, , Pages: 143-152, ISSN: 0302-9743
Identifying robust associations between cardiac imaging phenotypes and clinical diseases is fundamental to population-scale cardiovascular research and reliable risk stratification. However, current phenome-wide association studies rely on pre-defined, single-variable phenotypes or expert-crafted features, which limits their ability to capture clinically meaningful non-linear effects and cross-phenotype interactions. To address this, we propose CPAgents, an iterative phenotype-Composition framework for cardiovascular Phenome-wide association study (PheWAS) that automatically constructs and validates interpretable composite phenotypes (e.g., polynomial, ratio, and interaction forms) from base imaging features. Specifically, our system coordinates three agents: (i) an Analyst that identifies statistical pathologies and nominates candidate transformations; (ii) a Proposer that generates constrained, medically and statistically motivated expressions under numerical safety rules; and (iii) a Verifier that evaluates candidates using multi-stage criteria and produces transparent evidence trails for accepted phenotypes. Evaluated on a population-scale cardiac imaging cohort, the discovered composite phenotypes markedly improve disease discrimination: across 72 classifier–disease–metric combinations, our variants achieve the top rank in 56 cases versus 18 for baselines, with gains observed across all nine clinical disease categories. Our framework yields compact, clinically interpretable phenotype formulas with transparent evidence trails, enabling scalable discovery of stronger phenotype-disease associations beyond expert-driven feature selection.
-
Journal articleBoyle B, 2026,
Skin care interventions in infants for preventing eczema and food allergy
, Cochrane Database of Systematic Reviews, ISSN: 1469-493X -
Journal articleBruno F, Langhorne J, Lo Celso C, 2026, , STAR Protocols, Vol: 7, ISSN: 2666-1667
In malaria-endemic regions, individuals experience repeated malaria infection throughout life. Here, we present a protocol to investigate the immunological mechanisms triggered by repeated exposure using a murine model of sequential infections with the rodent parasite Plasmodium chabaudi chabaudi. We provide a detailed and standardized series of steps covering primary infection, natural recovery without treatment, and subsequent secondary infection. This approach offers a robust framework to investigate the effects and long-term consequences of repeated malaria infections on the host.
-
Journal articleAhmad AK, Ooi BMF, Ofili-Yebovi D, et al., 2026, , Endocrines, Vol: 7, ISSN: 2673-396X
Background/Objectives: Tubal ectopic pregnancy (TEP) remains a significant cause of maternal morbidity and mortality. Early access to pregnancy assessment units now allows for the detection of many TEPs at a stage at which they could potentially resolve spontaneously. The current selection of women for expectant management is guided by human chorionic gonadotrophin (hCG) concentrations, alongside ultrasonographic and clinical features, although reported success rates vary widely in the literature. We evaluated whether progesterone, another biomarker secreted during pregnancy, can predict successful expectant management alone or in combination with hCG. Methods: A narrative review was undertaken using Ovid MEDLINE and Embase, searched from 1 January 1990 to 1 May 2026, with a supplementary grey literature search. Of 1363 records identified, 27 studies were included. Results: Progesterone is associated with pregnancy viability but performs poorly in isolation for distinguishing ectopic from non-ectopic pregnancy. However, in studies of expectant management success, it has been identified as a second-tier variable that may improve prediction of successful expectant management of TEP. Conclusions: Current evidence does not support the use of progesterone as an independent predictor of expectant management success. It may provide additional prognostic information within an hCG-defined stratum, but this requires further validation before implementation in clinical practice.
-
Journal articleTruter C, Hamilton MS, Le Roux C, et al., 2026, , BMC Pediatrics, Vol: 26, ISSN: 1471-2431
BackgroundOne million children develop tuberculosis (TB) each year and a quarter of these die. TB meningitis (TBM) is the most severe form of TB disease and even if diagnosed and treated, 20% die and over 50% of survivors are left with permanent neurological disability. Much of the morbidity and mortality associated with TBM is due to infarction yet, despite this, our understanding of the pathogenesis of infarction in TBM remains limited, especially in children. MethodsThe iThemba study (Understanding ischemia in children with tuberculous meningitis), aims to recruit 100 children with probable or confirmed TBM and obtain samples of blood and cerebrospinal fluid (CSF). All children will undergo MRI and FDG PET/CT at baseline and will have repeat MRI with further blood and CSF samples collected at 2 weeks’ follow-up. MRI will then be carried out at 24 weeks with neurodevelopmental assessment at 48 weeks. Neuroimaging will focus on methods to identify and characterize ischemic penumbra and evaluate how this correlates with clinical outcomes. RNA sequencing of blood and CSF will be used to identify differentially expressed genes and identify implicated biological pathways between children with and without infarction. Targeted proteomic profiling will be performed on plasma and CSF to determine differences in protein abundance, with a focus on proteins involved in coagulation and endothelial function. Finally, we will integrate transcriptomic, proteomic and radiomic data to generate a comprehensive understanding of the pathogenesis of infarction in children with TBM. We aim to group children into relevant biological/anatomical phenotypes, each of which may benefit from a different therapeutic approach. Using computer simulation, we will then explore the impact of potential therapeutic interventions on biological pathways for each distinct phenotype. This work may pave the way for the development of point-of-care tests at diagnosis that could allow for stratified no
-
Journal articleWan Y, Wong JLC, Sanchez-Garrido J, et al., 2026, , BMC Genomic Data, Vol: 27, ISSN: 2730-6844
Background Resistance to carbapenems and third-generation cephalosporins is increasing in Klebsiella pneumoniae globally, restricting therapeutic options. The β-lactam/β-lactamase inhibitor combinations are widely used to circumvent β-lactamase-mediated resistance. In 2021, an unusual K. pneumoniae clinical isolate, KpMVR1, was recovered from a hospitalised patient in England, exhibiting resistance to meropenem-vaborbactam, imipenem-relebactam, and ceftazidime-avibactam. To investigate this phenomenon, we characterised the genome and antimicrobial susceptibility of KpMVR1 alongside two clonally related isolates susceptible to all three β-lactam/β-lactamase inhibitor combinations: KpMVS1, collected from the same patient 42 days earlier, and KpMVS2, from another patient in the same hospital. Methods Illumina and MinION whole-genome sequencing were conducted for these three isolates, followed by hybrid genome assembly. Annotated genome assemblies were compared to identify genetic variation. Mutagenesis experiments were performed to verify predicted functional alterations. Results All isolates belonged to clone ST8134 and carried blaKPC-2 alleles (KpMVR1: blaKPC-157; KpMVS1 and KpMVS2: blaKPC-2) in plasmids predicted to be conjugative. Insertion sequence ISEc68 caused a frameshift mutation in KpMVR1’s ompK36 gene, reducing susceptibility to meropenem-vaborbactam and imipenem-relebactam. KPC-157 demonstrated decreased hydrolysis of imipenem and ceftazidime when compared with KPC-2. KpMVR1 also encoded a disrupted transcriptional repressor MarR and a destabilising mutation in AcrB, a component of the AcrAB-TolC multidrug efflux pump. An intact, iron-transporting fec operon was identified on a novel IncFII(pKP91)/IncFIB(K) plasmid unique to KpMVS2, possibly accounting for the cefiderocol resistance observed in this isolate. ConclusionsKpMVR1 carried multiple resistance-associated genetic alterations and likely developed its resistance profile
-
Journal articleJebreel WMA, Abdel Hamid MM, Mustafa SA, et al., 2026, , BMC Infectious Diseases, Vol: 26, ISSN: 1471-2334
BackgroundMalaria causes high morbidity and mortality in Sudan. Malaria control efforts have been disrupted by conflict and displacement, which affected the whole health system. This study aimed to characterize malaria and glucose-6-phosphate dehydrogenase deficiency in conflict-affected zones of southern and eastern Sudan.MethodsThis cross-sectional study was conducted between 2023 and 2024, enrolling 717 patients with clinical symptoms suggestive of malaria in Kosti (southern Sudan, n = 252) and Kassala (Eastern Sudan, n = 465). Malaria infection was confirmed by light microscopy as a standard test, and qPCR was used as a reference method. Haematological indices were analysed on automated analysers, and PCR-RFLP was used to determine G6PD genotypes.ResultsMalaria prevalence was 62.6% (291/465; 95% CI 58.2–67.0%) in Kassala and 52% (133/252; 95% CI 46.6–59.0%) in Kosti using PCR. In Kassala, 157 cases (54%; 95% CI: 48.2–59.7%) were Plasmodium vivax (P.v), 99 (34%; 95% CI: 28.6–39.8%) were Plasmodium falciparum (P. f), and 35 (12%; 95% CI: 8.6–16.2%) were P. f/P. v infections. In Kosti, P. f was detected in 130 (97.7%) subjects, and P. v was detected in 3 (2.7%; all were negative by microscopy). There were 37 (8.7%) subjects not detected by microscopy but positive by PCR (submicroscopic), 20 (15%) in Kosti and 17 (5.8%) in Kassala. The G6PD B variant predominated in Kassala (438/94.2%) and Kosti (200/79.4%). The African A− variant was detected in 9 (3.5%) individuals in Kosti (7 males, 2 females). In Kosti females, BA, BA−, AA, and AA− were observed in 11 (7%), 4 (2.5%), 4 (2.5%), and 9 (5.7%), respectively, compared to 10 (4.1%), 4 (1.6%), and 7 (3%) BA, BA−, and AA cases in Kassala females. No significant association was observed between G6PD genotype and parasite density. Malaria prevalence did not differ significantly between Internally Displaced Persons (IDPs) and residents.Conclu
-
Journal articlePashley SL, Hair M, Ukegbu CV, et al., 2026, , Life Sci Alliance, Vol: 9
Plasmodium parasites rely on diverse protein kinases to regulate critical processes, yet many remain uncharacterised. Protein kinase 2 (PK2) is required for asexual blood-stage proliferation and merozoite invasion, but its role in sexual stages and transmission is unknown. Here, using live-cell and super-resolution imaging, ultrastructural analysis, functional phenotyping, and quantitative phosphoproteomics, we define the role of PK2 in Plasmodium berghei transmission. PK2 was expressed in merozoites, ookinetes, and sporozoites: the three invasive forms of the parasite. Conditional depletion of PK2 impaired ookinete motility, disrupted apical microneme positioning, abolished oocyst formation, and prevented mosquito-to-host transmission. Direct haemocoel injection of mutant ookinetes failed to restore sporozoite production, demonstrating an additional requirement for PK2 downstream of midgut traversal. Although global protein abundance was largely preserved, phosphoproteomic profiling identified altered phosphorylation of proteins associated with cytoskeletal organisation, host-cell interaction, midgut invasion, and gene regulation. These findings highlight PK2's critical role across multiple invasive stages, advancing understanding of malaria parasite transmission and identifying putative PK2-regulated pathways as potential targets for transmission-blocking interventions.
-
Journal articleCelsa C, Pressiani T, Nishida N, et al., 2026, , JHEP Reports, Vol: 8, ISSN: 2589-5559
Background & AimsDurvalumab plus tremelimumab (STRIDE) has emerged as a first-line systemic treatment option for unresectable hepatocellular carcinoma (HCC). This international multicentre study aimed to evaluate the efficacy and tolerability of STRIDE or durvalumab monotherapy in routine clinical practice, comparing outcomes between patients within and outside key eligibility criteria for the HIMALAYA trial.MethodsFrom a database of 1,423 patients with advanced/unresectable HCC treated with immunotherapy across 35 centres, we analysed 233 patients receiving STRIDE or durvalumab monotherapy. Patients were categorized as HIMALAYA-IN or HIMALAYA-OUT based on key trial eligibility criteria (no prior systemic therapy, ECOG-PS 0–1, Child-Pugh class A, no Vp4 thrombosis). Baseline characteristics were assessed for overall survival (OS) and hepatic decompensation using a multivariable Cox model and competing-risk analysis, respectively. Objective response rates and treatment-related adverse events were recorded.ResultsOf the 233 patients, 123 (53%) were HIMALAYA-IN and 110 (47%) were HIMALAYA-OUT. STRIDE was given in 95% of HIMALAYA-IN patients. After median follow-up of 6.0 months, median OS was 20.4 months (95% CI 11.7-NR) in the overall population. HIMALAYA-IN patients achieved significantly longer OS than HIMALAYA-OUT patients (23.0 vs. 12.2 months; hazard ratio 0.61; 95% CI 0.39-0.96; p = 0.03). Macrovascular invasion and hepatic decompensation were independent negative prognostic factors in the whole cohort. Hepatic decompensation occurred in 10.5% of patients within 12 months from treatment start. Objective response rate was 23.7% and 17.8% of HIMALAYA-IN and -OUT patients, respectively. Patients achieving disease control (whole cohort: 59.4%) demonstrated 24-month OS of 58.2% in HIMALAYA-IN and 44.8% in HIMALAYA-OUT groups. Grade 3-4 treatment-related adverse events occurred in 16.3% of patients.ConclusionsSTRIDE shows reproducible effectiveness and an ac
This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.