53417 research outputs found
Sort by
Towards biomarker-based diagnosis of Parkinson disease.
No full textThe current clinical diagnostic criteria for Parkinson disease (PD) have limitations and are inherently insensitive to the earliest stages of disease, when classical motor signs can be absent. Imaging and genetic tests are currently used to support or establish a diagnosis of PD, but no validated biomarker-based diagnostic framework currently exists. Substantial progress has been made in the field of molecular disease markers, most notably with the development and validation of seed amplification assays (SAAs), which enable detection of very low levels of pathological α-synuclein in the cerebrospinal fluid and other biofluids and tissue. In this Review, we discuss the potential of α-synuclein SAAs and other biomarkers to improve diagnostic accuracy and enable earlier diagnosis of PD. We consider biological disease definitions that have been proposed on the basis of these biomarkers, highlighting their merits, limitations and implications for PD research and clinical management. Research is ongoing to determine the predictive value of PD biomarkers in healthy people and people with prodromal PD and to develop markers that are sensitive to disease progression, both of which are key for implementation of trials involving drugs designed to modify or prevent disease. Integrating clinical, genetic, molecular and imaging biomarkers should enable earlier, more accurate diagnosis of PD and characterization of PD subtypes, thereby enabling personalized treatment to slow or even prevent PD
An Investigation on Lidar-Based Relative Pose Estimation and Image-Based Absolute Pose Estimation
No full textAccurate and efficient estimation of sensor pose is a key requirement in autonomous driving, robotics, and mixed-reality applications. Lidar-based relative pose estimation (point cloud registration) must cope with sparse and partially overlapping point clouds acquired under changing viewpoints, while image-based absolute pose regression (APR) has to infer camera pose from a single RGB image in environments with limited training data and appearance changes. Existing geometric registration methods, such as variants of Iterative Closest Point (ICP) and robust correspondence filtering, degrade under extreme sparsity and cross-scene generalization. APR methods based on convolutional networks or transformers reduce storage compared to structure-based localization but often exhibit poor robustness outside the training distribution. Recent geometry-aware 3D representations, ranging from Special Euclidean Group(SE(3))-equivariant graph neural networks for point cloud registration to implicit and explicit scene models such as Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS), provide richer priors over scene geometry and appearance, but still face limitations in computational cost, robustness to sparsity, and their integration with APR pipelines. This thesis investigates how geometry-aware neural representations can improve the robustness and efficiency of both lidar-based relative pose estimation and image-based APR. First, an SE(3)-equivariant graph neural network is introduced for sparse point cloud registration. By jointly learning point features and poses in an equivariant manner, the model improves rotation and translation accuracy over non-equivariant deep baselines on 3DMatch and KITTI, and shows better cross-dataset generalization with sparse point clouds as input. Second, a NeRF-based multi-modality supervision framework is proposed for APR. A NeRF model is trained to reconstruct RGB, depth, and high-level feature fields, and the resulting synthetic views supervise a pose regressor. On the indoor (7-Scenes) dataset, this multi-modal NeRF supervision yields lower median translation and rotation errors than RGB-only APR baselines. Finally, to overcome the efficiency and quality limitations of NeRF, the final contribution introduces a framework that uses 3DGS as a superior data synthesis engine. By leveraging the real-time, high-fidelity rendering of a multi-modality supervised 3DGS model, a large-scale, diverse dataset with geometric and appearance perturbations is generated. Training an APR model on this synthetic data advances the state-of-the-art (SOTA) on challenging indoor (7-Scenes) and outdoor (Cambridge Landmarks) benchmarks, demonstrating significant gains in accuracy and generalization. Overall, the results demonstrate that combining equivariant geometric models with implicit and explicit 3D scene representations is an effective strategy for achieving robust and efficient pose estimation across lidar-based relative and image-based absolute settings, addressing key limitations of existing registration and APR methods
Unmet Medical Needs in Pediatric Endocrinology
No full textPediatric endocrinology has made remarkable advances over recent decades, transforming the lives of countless children and families. Yet, major challenges persist. Many rare and complex endocrine disorders remain difficult to diagnose, monitor, and treat effectively. Disparities in access to specialized care, limited research investment, and fragmented health systems continue to create inequities in outcomes across regions and populations. This document arises from the collective effort of the European Society for Paediatric Endocrinology (ESPE) to highlight these unmet medical needs and to chart a path forward. It underscores the necessity of harmonized diagnostic standards, innovative research, and sustainable policies that support both patients and professionals. By identifying key barriers and proposing strategic directions, ESPE aims to foster collaboration among clinicians, researchers, policymakers, and patient communities. Only through coordinated action can we ensure that every child with an endocrine disorder receives equitable, timely, and high-quality care—regardless of where they live.</jats:p
Defining the bellwether procedures and processes for global trauma care: an international Delphi study.
Full text linkBACKGROUND: The complexity of delivering trauma care makes the assessment of its provision challenging. The identification of bellwether procedures has previously been successful in the evaluation of global surgical care; however, any equivalent in assessing trauma care is currently lacking. Through a Delphi process, we aimed to produce the bellwether procedures and processes for global trauma care. METHODS: A global Delphi process was undertaken with healthcare professionals and academics involved in trauma care from across the world. A list of potential procedures and processes was identified through literature review and expert opinion, along with subsequent additional options suggested by respondents. Three successive rounds were completed, with respondents rating the importance of each procedure or process to be undertaken at any hospital that cares for trauma patients using a five-point Likert scale. RESULTS: A total of 411 respondents from 78 countries completed the initial round of the Delphi process, with minimal attrition observed across rounds. Following three successive rounds of the Delphi and functional aggregation, nine bellwethers of global trauma care were determined, subdivided into three functional categories: 'Resuscitation & Stabilisation'-(1) Advanced Airway Management, (2) Short-term C-spine Immobilisation, (3) Long Bone Immobilisation; 'Diagnosis & Monitoring'-(4) Blood Gas Analysis, (5) Focused Assessment with Sonography in Trauma (FAST) Scanning, (6) Continuous Access to CT Imaging; 'Optimisation & Intervention'-(7) Blood Transfusion, (8) Tube Thoracostomy, (9) Laparotomy and Splenectomy. CONCLUSION: The Global Trauma Care Delphi study has produced nine metrics that provide pragmatic indicators for the overall assessment of trauma care capabilities at any healthcare setting worldwide. These bellwethers of global trauma care can enable hospitals, local managers and health ministries to identify institutions or regions that may require more in-depth assessment, allowing standards in the management of traumatic injuries to improve
Investigating the dynamics and characteristics of neutrophils infiltrating the draining lymph nodes following Ischaemia-Reperfusion Injury, and their role in regulating CD4+ T cells
No full textIschaemia-Reperfusion Injury (IRI) is a deleterious inflammatory response associated with cardiovascular pathologies, such as myocardial infarction (MI), stroke, or organ transplantation. IRI involves the recruitment and activation of both innate and adaptive immune cells, particularly neutrophils and T helper (CD4+) cells. In MI, CD4+ T cells are surprisingly detected in the infarcted tissue within 24hrs. However, the mechanisms of this early CD4+ T cell activation is unknown as it precedes the entry of classical professional antigen presenting cells (pAPCs) to secondary lymphoid organs. Interestingly, preliminary work from our group demonstrated the presence of neutrophils into draining lymph nodes (dLNs) of ischaemic tissues at this early time-point of the inflammatory response. This project aims to define the unique characteristics of neutrophils within this secondary lymphoid organ and to investigate their role in regulating CD4+ T cell responses during the acute phase of IRI. Using 2 distinct murine models of IRI, we reveal by flow cytometry that neutrophils transiently infiltrate the dLNs within 8–24hrs post-reperfusion. Complementary qPCR analyses identified a rapid Th1/Th17-skewed activation of CD4+ T cells post-reperfusion; a response suppressed in neutrophil-depleted animals. Furthermore, we demonstrate neutrophils and CD4+ T cells engage into dynamic interactions in the dLNs, as observed by confocal microscopy. Moreover, flow cytometry and NanoString transcriptome analyses identified that dLN-infiltrating neutrophils exhibit phenotypic and transcriptomic features of pAPCs, including the expression of MHC class II (MHC-II) molecules and co-stimulatory receptors, supporting their potential role in direct antigen presentation to T cells in vivo. To directly interrogate this mechanism, we have generated and characterised a novel neutrophil-specific MHC-II knockout mouse model. While neutrophil MHC-II deletion does not alter overall neutrophil dynamics in the dLNs or inflamed tissue, it significantly impairs Th1 and Th17 like responses in a sex dependent manner. Collectively, these findings identify neutrophils as critical regulators of early, pro-inflammatory CD4⁺ T cell activation in the context of IRI. By mediating cytokine production, chemokine-driven recruitment, and direct antigen presentation, neutrophils emerge as potential therapeutic targets for modulating adaptive immunity during acute ischaemic events
BCR::ABL1-Induced Enhancer Reprogramming Uncovers Hypersensitivity of Ph+B-ALL Cells to Enhancer-Targeting Drugs.
No full textCancer is driven by genomic lesions and malignancy-promoting transcriptional programs. In blood cancers, both are often interconnected as lesions frequently affect transcription factor (TF)-encoding genes. TFs largely function through enhancers, and enhancer deregulation is linked to cancer initiation and progression. Consequently, enhancer-targeting drugs are in trials for several advanced hematologic cancers. However, for cancers not driven by TF-related lesions, it is less clear how their transcriptional programs are established; if oncogenesis involves enhancer-deregulation, and if they are sensitive to therapeutic enhancer-targeting. Here, we explore this for Philadelphia chromosome-positive (Ph+) B-lineage leukemia (B-ALL), the most common B-ALL in adults with a historically poor prognosis. Ph+B-ALL is driven by BCR::ABL1, a kinase without TF-related function. We report that malignant transformation and transcriptional reprogramming by BCR::ABL1 is indeed defined by enhancer reprogramming and that enhancer signatures differentiate Ph+B-ALL from other leukemias. Mechanistically, we show that BCR::ABL1 itself induces enhancer activation, through its kinase activity and via kinase-dependent activation of STAT5, ETV5, and MYC. Consequently, BCR::ABL1-induced genes are hypersensitive to enhancer inhibition, and Ph+B-ALL cells are hypersensitive to enhancer-targeting drugs. Enhancer-targeting further improves the efficacy of BCR::ABL1 kinase inhibitors used for Ph+B-ALL therapy, especially in cells from IKZF1PLUS patients that most frequently relapse from current treatment, suggesting enhancer-targeting as a potential promising addition to current therapy
An efficient implementation of legofit software to infer demographic histories from population genetic data
No full textHealthcare needs and priorities of older people living with heart failure and frailty: a multi-perspective study of patients, carers and clinicians.
No full textLet's be emotional! Experimental evidence for an alternative mediation logic of emotions in business relationships
No full textThis study conceptualizes and provides experimental evidence for the mediating role of emotions in business relationships. Drawing on Relational Norm Theory, we explore how (positive, negative) emotional responses to the (non)fulfillment of relational norms influence trust as a key characteristic of business relationships. We employ a thematic analysis of the literature, augmented by sensitizing interviews with boundary spanners (i.e., organizational representatives that manage interactions in business relationships) from buying and selling firms, to develop a conceptual dual-pathway model, and an experimental design for causal testing. Our results suggest that emotions of positive and negative valence partially mediate the relationship between relational norm fulfillment and trust, in extension to the established mediation models of business relationships in the literature. While positive and negative emotions do not differ in their degree of influencing trust outcomes, relational performance norm (non)fulfillment was found to generate stronger emotional reactions than that of relational interaction norms. Our research contributes to a more nuanced understanding of business relationships by integrating an alternative explanatory mechanism complementing structural and attitudinal mechanisms, operating on the level of boundary spanners within established models. Way markers for novel research avenues are proposed, and practical insights for managers in interorganizational contexts are derived
