This project seeks to address a key UK societal challenge in healthcare concerning neurological disorders. It is estimated that neurological conditions are a leading cause of disability in the UK and cause around one-fifth of all deaths yearly. Growing evidence points to the pivotal role of mechanics in neurological disorders and experiments show that neurodegenerative diseases, such as multiple sclerosis, Alzheimer's disease, and demyelination, lead to changes in nervous tissue microstructure. For instance, autism has been linked to structural plasticity-associated changes leading to alterations in dendritic spines' (tiny protrusions from dendrites, which form functional contacts with neighbouring axons of other neurons) shape and number. From the modelling point of view, efforts have been made to understand the mechanisms underlying such neurological conditions through different theoretical and computational approaches and to elucidate innovative treatment strategies. While advances have been made in the investigation of the function and dysfunction of nervous tissues, many common challenges still need to be addressed and realistic computational simulations based on a multiscale and multiphysics framework tying together the individual pieces of information are still missing. The methodologies developed in this proposal will provide new mathematical infrastructures and generate a fundamental scientific understanding of the nerve tissue by explaining how the relationship among electro-chemo-mechanical interactions at individual scales contributes to its evolution through mathematical modelling and high-performance computing. The outcomes and knowledge generated by the proposal will be of most benefit to the scientific community and, particularly, to healthcare as it aims at understanding some of the mechanisms concerning the progression of neurological disorders and shed light on new information that can be useful for the conception of novel treatment strategies. A long-term goal is to reinforce the scientific community in terms of the technological transfer towards biomedicine through accessible, but sophisticated, computer software.

Stroke is a major killer, and has a greater disability impact than any other chronic disease. Patients occupy one of the largest numbers of acute hospital bed days of any patient group (2.6 million per year), and the disease consumes between 4-6% of NHS expenditure (approx £2.8 billion) with costs projected to rise 30% by 2010 as life expectancy and survive rates increase. Treatment options are limited (clot busting drugs like rt-PA or aspirin) with less than 2% of stroke patients receiving rt-PA since it must be given within 3 hours of stroke for safety reasons, and patients must have a brain scan first to rule out brain haemorrhage. rt-PA is very effective in patients where brain damage has not fully evolved. In the first few hours after stroke, injured but potentially salvageable tissue (penumbra) can be saved by rt-PA. Penumbral tissue can be identified by magnetic resonance imaging (MRI), but current techniques underestimate penumbra size. If a rapid, more accurate MRI technique was available, more patients could be treated and recruitment into clinical trials to study new therapies, improved. This proposal describes development and validation of such a technique and provides preliminary MRI data from animal models and man.

Currently, there is no established methodology for implementing Doughnut Economics that fully accounts for the dynamic interaction between the targets. Researchers in GALLANT are working on it, and this PhD is a part of this activity. On this project, Jonathan is using a mixed-methods approach, combining both qualitative methods (e.g. The Soft Systems Methodology) and quantitative approaches (e.g. The NK fitness models), and applying them using the City of Glasgow as the case study. They are working in a multidisciplinary environment, linking knowledge across healthcare, economics, and environmental systems to support decision-making by Glasgow City Council.

Antibiotics such as ciprofloxacin and metronidazole are used to reduce inflammation in the intestines [1]. However, due the rise of antibiotic resistance, the inflammation remains. Continuous bacterial infections or large bacterial burdens can overwork the immune system and cause dysregulation and lead to inflammatory diseases such as irritable bowel disease (IBD) [2]. Patients with a history of these intestinal ailments are at a higher risk of developing intestinal cancers, including colon cancer, due to the existing inflammation in the intestinal mucosa [3]. Additionally, the overuse of antibiotics can modulate gut microbiome and host-microbial interactions. Consequently, these may influence the barrier junctions of the intestine, inducing inflammatory and pathogenic effects. A "leaky" intestine would impose risks to bloodstream infections. Current organ-on-a-chip (OoC) models have expressed in vitro intestine devices to absorb nutrients and drugs. However, there are difficulties imitating the complexities of the physiological conditions. They lack the imitation of the four layers of the intestinal wall and can only culture the epithelial lining and gut bacteria both together and separately for less than a week [4]. Gut microbiomes-on-a-chip devices have also been exhibited with growth of bacteria and have great potential for modeling IBD and other intestinal diseases. The Yin Lab at the University of Glasgow has developed an intestine-on-a-chip model for drug screening through microfluidic extrusion of channels to create hollow microfibers. These microfibers use an alginate-based outer layer to mimic the stiffness of vessels and have a collagenbased core [5]. This promotes cell growth from a bioactive microenvironment. The microfluidic device

Physics-informed and physics-constrained machine learning for next generation imaging. A team of researchers from the University of Glasgow and the "big image data" company, Dotphoton. The student will be part of a multidisciplinary team of physicists and working on novel imaging techniques, of computer scientists working on novel machine learning approaches that encode the physics of the imaging problem and bio-engineers working with cutting edge microscopy techniques. The ambition is to develop new-generation, physics-constrained AI that can image better, faster and more intelligently that current systems. By embedding physical constraints in to the design of the AI, you will develop better microscopes and biological imaging techniques that will be tested on new-generation fluorescence microscopes and healthcare monitoring devices. The research will be carried out at the Advanced Research Centre (ARC) in Glasgow, where you will work with a team of physicists, computing scientists, engineers and biologists. The project will be in close collaboration with Dotphoton (Switzerland) and will ideally involve also in-person visits to the company premises to work with their team of data scientists.