324139 research outputs found
Sort by
Developing acute respiratory infection severity indicators for public health surveillance using primary care computerised medical records
Get PDFIntroduction: Acute respiratory infection (ARI) surveillance systems generate essential intelligence to help health authorities protect the public from the consequences of epidemic and pandemic-prone pathogens such as influenza. This Doctor of Philosophy thesis explores how data from primary care computerised medical records (CMRs) can be used to strengthen surveillance of ARIs. Specifically, it describes the development and evaluation of timely population-level severity indicators of ARIs derived from primary care CMRs. Methods: The thesis consists of four pieces of work: (1) defining an algorithm for the identification of episodes of ARI from the CMR; (2) a systematic review to identify possible markers of severe disease relevant to ARIs in primary care; (3) an assessment of the data quality of these severity markers in the primary care CMR; and (4) a retrospective evaluation of these severity markers to determine their suitability for use in prospective public health surveillance of ARIs. Key findings: The case detection algorithm provided a unified and flexible approach that increased sensitivity for identifying ARIs and established a suitable cohort for assessing severity. The systematic review identified 30 potential severity markers, comprising seven severe outcomes and 23 more timely predictors of severe outcomes. Severe outcomes included death, hospitalisation, intensive care admission, and complications, while predictors included symptoms, signs, investigations, treatments, and healthcare utilisation markers. The data quality of severity markers varied significantly and was heavily affected by the pandemic. Several predictors showed strong potential as timely severity indicators, with some symptoms, signs, and healthcare utilisation markers demonstrating significant associations with severe outcomes. Conclusions: This thesis demonstrates that primary care CMR data can be used to create timely severity indicators for ARIs. Future work should focus on piloting severity indicators prospectively and in near real time and improving the recording of severity markers. This could provide a pathway to the implementation of reliable and timely severity indicators for routine public health surveillance
Quantization of the Willmore energy in Riemannian manifolds
No full textWe show that the quantization of energy for Willmore spheres into closed Riemannian manifolds holds provided that the Willmore energy and the area be uniformly bounded. The analogous energy quantization result holds for Willmore surfaces of arbitrary genus, under the additional assumptions that the immersion maps weakly converge to a limiting (possibly branched, weak immersion) map from the same surface, and that the conformal structures stay within a compact domain of the moduli space
Tobacco Smoke Exposure From Prenatal To Adolescent Periods Drives IBD Pathogenesis: Dynamic DNA Methylation Signatures Across Lifespan Stages
Get PDFThe association between early life exposure to smoking and the risk of inflammatory bowel disease (IBD) needs to be further verified, and the potential role of DNA methylation in the association is unclear. Through an integrated study design, this study demonstrates that maternal smoking during pregnancy (MSDP) is potentially associated with increased risk of IBD, Crohn's disease (CD), and ulcerative colitis (UC) in offspring. In addition, individuals who started smoking in adolescence have a higher risk of developing CD and IBD. Mechanistically, MSDP‐associated DNA methylation alterations in ADCY7 (newborn), AKAP8L (newborn), TIGD7 (newborn), and TNF/LTA (across life stages) are significantly correlated with increased risk of CD in offspring; MSDP‐induced DNA methylation changes in PRRT1 (newborn), AHRR (across life stages), and MYO1G (across life stages) show significant associations with UC risk in offspring. Notably, the alterations of DNA methylation status within AHRR, MYO1G, and TNF/LTA loci associated with smoking exposure are present throughout the life course. Collectively, MSDP may serve as an independent risk factor for IBD in offspring, and active smoking during adolescence may cause increased risk of developing CD and IBD. MSDP may contribute to IBD susceptibility by inducing persistent DNA methylation alterations at multiple developmental stages
Development and validation of a parsimonious AI-based risk score for mortality in heart failure: a UK cohort study
Get PDFBackground: Accurate risk stratification in heart failure (HF) populations is crucial. Existing simple scores, such as MAGGIC, show limited discrimination (C-index < 0.80) and require specialised tests like echocardiography. Conversely, while advanced Artificial Intelligence (AI) models offer greater accuracy, their operational complexity and data access requirements often hinder clinical adoption. Purpose: To develop and validate a parsimonious, high-performing AI-based risk model for all-cause mortality in HF patients by leveraging insights from a complex AI model and feature engineering. Methods: Using a cohort of 373,389 patients with HF (≥18 years; 1,153 English practices for derivation, 289 for validation) from the Clinical Practice Research Datalink (CPRD) Aurum dataset, we developed and validated an AI-based risk prediction model for all-cause mortality. A MAGGIC-based Cox proportional hazards model adapted for electronic health records (EHR) was established as a performance benchmark. An initial Multi-layer Perceptron (MLP) model with a survival framework was trained with MAGGIC variables. This MLP model was then enhanced by incorporating highly predictive comorbidity features—identified by an explainable Transformer model trained on longitudinal electronic health record (EHR) data—and feature distillation to derive a final, optimised 11-variable model (named MLP-Distilled). The model employed readily ascertainable variables including age, BMI, year of birth (as a proxy for birth cohort effects), and key comorbidities such as cancers. Results: MLP-Distilled demonstrated significantly improved discriminatory performance (C-index: 0.801, 95% CI [0.796, 0.806]) compared to the benchmark MAGGIC-EHR model (0.735, [0.729, 0.742]), while maintaining excellent calibration. This parsimonious model uses fewer and more accessible variables than MAGGIC (11 vs. 13) and requires no specialised tests. The model was validated for secondary outcomes of cardiovascular events and rehospitalisation. Conclusion: Utilising knowledge from a complex EHR-trained AI model combined with feature distillation yielded a parsimonious yet powerful risk score for HF mortality. This novel approach offers a pathway to clinically implementable tools that balance predictive accuracy with pragmatic utility, potentially improving routine HF risk stratification.Discrimination_Calibration.jpeg Impact_Analysis_50_Threshold.jpe
India, colonialism and “energy justice”: Dinabandhu Mitra, Rudyard Kipling and Rokeya Sakhawat Hossain
No full textThe praxis of justice is central to the processes of both decolonization and post-capitalism. But the desired transition from colonialism and capitalism is hampered because such praxis often confines itself within the classical limits of “distributive justice”. While distributional issues are self-evidently important, justice cannot be wholly equated to them since they are finally determined by historical conditions of domination, resistance and recognition. The contemporary movement for “energy justice” has been one important source of reconfiguring a praxis of justice aligned to such conditions. Ranging from the examination of racial and ethnic dimensions of toxic hazards and “natural disasters” to offering new structural analyses of colonialism and capitalism, “energy justice” compels a twinned recognition of the historicity of environment and the environment of history. But such contemporary models of “non-distributive energy justice” also have a long genealogy rooted in nineteenth-century anti-colonial and colonial imaginations. Hence, this chapter reads Rudyard Kipling’s “The Bridge Builders” (1898), Dinabandhu Mitra’s Neel Darpan (1858) and Rokeya Sakhawat Hossain’s “Sultana’s Dream” (1905) to trace some aspects of this genealogy. Across a variety of forms and genres, these texts test out the relationship between energy, justice, colonialism and capitalism as well as the limits and possible mutations of their current configurations. Such fictional experiments and registrations are indispensable to our collective efforts to live in a world where climate change has now decisively redrawn the boundaries and extents of both power and resistance
News from everywhere: fiction and the Victorian world wide web
No full textThere is a general agreement regarding the relationship between the technologies of media/communications and global modernity - that the epochal developments in the former are both indices and drivers of the latter. But as Friedrich Kittler, Stuart Hall and others have also long argued from a variety of different critical positions, modern media & communications embody and enable specific kinds of power, domination and inequality - social, political, economic, cultural and aesthetic. Many of these media networks were enabled by, and enabled, the Victorian world order. How did Victorian world literature register such the presence of such networks and their work of mediation? In this essay, I range across a variety of genres - crime fiction, science fiction, sensation fiction and adventure tales - to show how the worlding of Victorian writing was achieved by enfolding media, mediation and empire into its form and content
Targeting ferroptosis pathway to overcome erlotinib resistance
No full textNon-small cell lung cancer (NSCLC) is the leading cause of cancer-related mortality worldwide. For advanced NSCLC, targeted therapies are often recommended to improve prognosis by acting on specific genetic mutations. Epidermal growth factor receptor (EGFR) is frequently hyperactivated in NSCLC, and EGFR tyrosine kinase inhibitors have been used clinically for decades to treat patients harbouring EGFR mutations. However, acquired resistance commonly develops, reducing therapeutic effectiveness. To address this challenge, we investigated whether erlotinib-resistant NSCLC could be targeted through ferroptosis, an iron-dependent cell death that can be promoted by radiation. Our RNA sequencing data revealed that genes involved in lipid metabolism and iron transport, key components of the ferroptosis pathway, were significantly dysregulated in erlotinib-resistant cells, along with those associated with tumour cell invasion. Western blot analysis further demonstrated that these cells have a compromised capacity to manage lipid peroxidation, likely due to lower xCT/SLC7A11 expression, a key regulator of ferroptosis resistance. Our data showed that erlotinib-resistant cells were more sensitive to erastin- and radiation-induced ferroptosis. We also found that erlotinib-resistant cells exhibited significantly higher lipid peroxidation and lipid droplet accumulation, which were further enhanced following ferroptosis induction. To validate our data in a physiologically relevant setting, we performed in vivo studies using mouse subcutaneous xenograft models. Consistent with in vitro data, erlotinib-resistant tumours were more sensitive to ferroptosis induced by the combination of radiation with the ferroptosis inducer IKE. Our research demonstrated increased ferroptotic susceptibility in erlotinib-resistant NSCLC, suggesting a novel therapeutic vulnerability. It provides a potential strategy for overcoming erlotinib resistance in NSCLC and highlights the role of ferroptosis in the treatment of erlotinib-resistant cancer
Automatic classification of lung cancers from histopathology images
Get PDFLung cancer accounts for more deaths than any other type of cancer. Currently, most lung cancers are diagnosed in symptomatic patients using CT scans and CT-guided biopsies or bronchoscopies. The latter two involve surgical excision of a small piece of tissue. To make a diagnosis, a pathologist examines the tissue under a microscope at different magnifications, noting cytological features and architectural patterns. These observations are aggregated into lung cancer subtypes, which may exhibit multiple characteristic patterns.
My research contributed to the broader DART Lung Health programme, which was based on the Targeted Lung Health Check programme conducted by NHS England. DART's main goals were to generate large datasets to enhance lung cancer diagnosis through quicker, less invasive, and more accurate methods while identifying research opportunities for treatments that could improve survival rates. I worked on automatically classifying lung cancers from histopathology images and creating an annotated histology dataset that would enable connecting histology and CT modalities. My main contributions are:
1. I developed a three-stage protocol for annotating lung cancer histology images from DART. I showed that it is possible to optimise the annotation process by selecting slides or regions with under-represented subtypes or patterns. My work resulted in a multi-centre dataset annotated to the degree unavailable in the public domain.
2. I curated a public lung cancer dataset and proposed using pretext tasks to choose promising patch-level histopathology foundation models for any custom dataset at a fraction of the computational cost of a rigorous benchmarking study. The choice of a good pretext task remains an open avenue of research.
3. I showed that incorporating prior pathology knowledge into model architecture and training pipelines enables models to learn both the dependencies between cancer subtypes and the relative importance of different regions on the whole slide images, improving the lung cancer classification performance as a result
Geometric and topological inference from random samples
Get PDFWe study statistical inference problems in geometry and topology inspired by data science. Generally we consider an independently, identically distributed random sample X = (X1, . . . Xn) drawn from a measure µ over a set M ⊆ RD. Then we consider the class of problem of inferencing a property P of M only using X. The following pairs of (M,P) are considered:
(1) M = Manifold, P = Dimension. We locally apply PCA (principal components analysis), and infer the dimension of M by counting how many of the variances fall under a threshold. While this method is widely used, a rigorous mathematical theorem guaranteeing its correctness was not established. We prove such a theorem.
(2) M = Manifold, P = Tangent spaces. The local PCA algorithm above can also be used to infer tangent spaces: a tangent space is estimated as a linear span of top principal components. Again for this standard well-known algorithm, we prove theorems guaranteeing the correctness of the algorithm.
(3) M = Stratified space, P = Singular points. A stratified space generalises manifolds, and possesses singularities at which there is no local resemblance to a Euclidean space. We present a fast algorithm that detects singularities using local hypothesis testing. The kernel method used in the algorithm is significantly faster than previous topological methods. Experimental results on both synthetic and real data are presented. Furthermore, we prove a theorem that guarantees the algorithm’s correctness in the case of union of two manifolds.
(4) M = Manifold, P = Homotopy type. A standard way to infer homotopy type of a manifold from a finite sample is via constructing a simplicial complex at a small distance threshold. However instead of stopping at a small threshold, we consider arbitrarily large connectivity thresholds and study anomalous topology arising from this. In particular, we study a very specific case of circle M = S1 , and show that Cech complexes arising from finite samples on M are homotopic to bouquets of high-dimensional spheres with high probability
Moving towards high-dose primaquine or single-dose tafenoquine for Plasmodium vivax treatment in Cambodia: a meeting report from dissemination of results of the EFFORT trial to stakeholders
Get PDFCambodia has targeted malaria elimination by 2025. As the malaria burden has decreased in Cambodia, transmission has become more focal, and Plasmodium vivax has become the predominant species. The recurrent nature of P. vivax, due to its dormant liver stages causing relapses, is the main obstacle to malaria elimination in Cambodia. In 2021, Cambodia’s National Center for Parasitology, Entomology and Malaria Control (CNM) rolled out low-dose 14-day primaquine (total dose 3.5 mg/kg) supported by point-of-care quantitative testing for glucose-6-phosphate dehydrogenase deficiency. However, this treatment is limited by poor adherence to its prolonged duration and suboptimal efficacy of the low total dose. The EFFORT clinical trial was conducted in four malaria-endemic countries, including Cambodia, to assess the safety and effectiveness of a 7-day unsupervised high-dose course of primaquine (7 mg/kg total dose) and single dose tafenoquine (300 mg) compared to 14-day unsupervised low-dose primaquine for the treatment of patients presenting with P. vivax malaria. In addition, data were collected on the feasibility and cost-effectiveness of these treatment options. CNM organized the national dissemination of the EFFORT study results on March 27, 2025, to inform key stakeholders and discuss the implications of the study findings for policy and practice in Cambodia