Chronic hepatitis B is one of the world’s unconquered diseases and presents as a spectrum of liver disease, reflecting a dynamic interaction between the virus and immune system. Although current treatments suppress hepatitis B virus (HBV), they are not curative and patients are at risk of developing progressive liver disease. The persistence of the HBV DNA genome and inadequate host immune responses limit progress towards a cure. HBV replication varies spatially within the liver and temporally between disease phases, suggesting localised hepatocyte-intrinsic and liver-resident immune resistance mechanisms. We have shown that oxygen levels influence HBV replication, leading to the hypothesis that hypoxia is a central unifying pathway regulating hepatocyte susceptibility to HBV infection and local immune control. We have developed single-molecule sensitive methods to visualise HBV RNA molecules, offering unprecedented sensitivity to quantify viral transcription in situ and new insights into the behaviour of individual cells that are masked in population-based studies. Combining this with spatial transcriptomics provides an integrated approach to study both viral and immune parameters in the liver and a step-change in our understanding of this chronic disease and paving the way for new therapies.

Tuberculosis (TB) is an important infectious disease which affects nine million people and causes two million deaths every year. The human immune system can protect against TB but is also responsible for causing the tissue damage that may result from TB disease. I aim to increase our understanding of the parts of the immune system that influence the balance of beneficial and harmful responses to TB infection in order to identify new targets for more effective treatments and vaccines. To do this I will combine experiments in a fish model that closely resembles human TB with experiments in patients with active TB disease. Initially, I will focus on the role of a specific part of the immune system called interleukin (IL)10 because this is a key mediator that controls the immune system during its response to infections. Thus far the role of IL10 in TB has almost exclusively been evaluated in mouse models that provide incomplete information because they do not accurately reflect human TB disease. In addition, I will take advantage of the massive increase in genetic data that has become available, in order to discover new components of immune responses to TB which vary most between people. I postulate that variable immune responses cause differences in outcome of TB infection. I will test this theory by investigating the effects of deficiency or excess of potential new regulatory factors that I identify in people with TB, using the fish model of TB infection. I hope to gain new insights that help to develop novel interventions that will significantly shorten the length of anti-TB treatment, which will be important to reduce spread of TB and minimise development of drug resistance. I anticipate that my work may also lead to design of new vaccines that prevent TB disease altogether.

Conventional Magnetic Resonance Imaging (MRI) is very time consuming (taking over 1hour/scan) and requires patients to remain still and perform multiple breath-holds. This is particularly difficult for children, and my work focusses on developing fast imaging techniques using Machine Learning (ML), to speed up MRI in children and reduce the need for breath-holding. I have shown that it is possible to use ML to 'learn' the best way to speed up the collection of MRI data, allowing each image to be collected up to 50x faster. To achieve these speed-ups, it is necessary to reduce the amount of data that we collect, however this results in significant errors in the images. I have shown that it is possible to reconstruct clinically useful images very quickly using ML; up to 100x faster than current state-of-the-art mathematical methods. In addition, I have developed fully-automated ML tools for analysis of the MRI images; enabling clinical metrics to be calculated whilst the patient is in the scanner, up to 7000x faster than conventional methods. Combined, these tools have allowed MRI scans in the heart and abdomen to be performed quickly (in as little as 10 minutes) and without the need for breath-holding in children, as well as in sick adults. In initial studies, these tools have been tested in small patient groups. This work focusses on large-scale international clinical testing, to make sure that these ML tools work reliably and accurately across all children's diseases and in different hospital settings. This work will build trust in these tools, enabling them to be shared them with different hospitals across the world, to maximise the benefits to all patients and hospitals. I will also continue benefitting from new developments in ML to further improve these technologies and overcome any limitations encountered during clinical testing. Most hospitals only have conventional MRI scanners, with field strengths of 1.5T or 3.0T. However, recently low-field strength (0.55T) MRI scanners have become available. Although these have not yet been clinically established, they offer significant financial benefits, including lower initial cost (~50% of 1.5T), easier/cheaper installation (70% of 1.5T) and lower running/maintenance costs (~45% of 1.5T). This makes these scanners highly desirable, and may enable MRI to become affordable in some countries for the first time. Low-field scanners, also address some of the remaining challenges of MRI in children; i) They have a bigger bore, so children find them less daunting, ii) They are much quieter, which means children may be able to remain asleep, and iii) There are less concerns over heating in the body. However, the measured signal at low-field strength is <25% of that on conventional scanners, resulting in lower quality images. Therefore, the second part of this extension will build on the ML tools that I have developed for imaging children on conventional scanners, to enable good quality images from rapid scans at low-field MRI. This includes the use of ML to 'find' the best way to collect the data, and ML methods to improve image quality. This extension will work towards clinical validation of these fast scans in children on conventional scanners, making MRI less difficult or daunting for children, improving availability and reducing risks. Quicker scans would help reduce waiting lists and costs for the NHS, and improve diagnostic accuracy and outcomes in childhood diseases. It would also mean that MRI scanning would be used far more often, so it could help many more children. Additionally, by investigating the use of rapid imaging tools at low-field MRI, I will continue to push novel research ideas, enabling improved image quality near to the lungs, in the abdomen and in fetus', with lower risk to patients and significant cost benefits for hospitals.

Over 120,000 people in the USA and over 6000 people in the UK are waiting for organ transplants, and many more are suffering from organ failure. Many donor organs are not transplanted (approximately 60% of donor hearts and 20% of kidneys are not used), often because they can only be kept for a short time. Long term preservation would mean better matching with recipients over larger geographical areas, reducing the chances of rejection and increasing the number of organs that could be used. One potentially transformative method of preserving organs for a longer time is cryopreservation. This involves freezing the organs at very low temperatures and then defrosting them when needed. However, this is currently limited to small volumes (<3 ml), largely due to the difficulty in rewarming the tissues without damage after freezing. To avoid damage on rewarming, tissues must be heated quickly and uniformly. This is not possible with existing water bath methods so the development of new methods for volumetric rewarming of large tissue volumes is critical. The aim of this fellowship is to develop a novel method of tissue rewarming using ultrasound. As ultrasound passes through frozen tissue, it loses energy which is deposited as heat. By controlling the pattern of the ultrasound waves entering the tissue, heat can be deposited as needed to raise the temperature of the tissue quickly and uniformly. First, the ultrasound parameters will be optimised for maximum cell viability and optimal heating rate using small volumes of cells. An ultrasound array based on these parameters will then be developed with methods of steering and shaping the acoustic field to uniformly and rapidly heat larger volumes of cells. This will be extended to warming tissues with inhomogeneous acoustic and thermal properties and larger volumes, using real time feedback to control the heating distribution, with the ultimate vision of creating a fully flexible tool that can be used to rewarm whole organs. Ultrasonic volumetric warming has the potential to enable long-term storage of tissues and organs which would transform the availability of organs for transplant. It would also have many other applications such as increasing access to therapies involving implanting cells and tissues in the body for diseases such as type 1 diabetes or for restoration of fertility after cancer therapy.

Ultrasonic imaging is a safe, inexpensive way of looking inside thebody. Unfortunately, not everything shows up clearly in anultrasound scan. Tumours can be hard to see, becausethey often reflect sound in much the same way as the surroundingtissue. Even when they are detectable, their boundaries can beindistinct. This makes it difficult for surgeons to plan preciselywhat to cut out, or for clinicians to assess how well a tumour isresponding to treatment. However, tumours are often stiffer thantheir surroundings. If ultrasound could show the tissue'sstiffness, instead of the way it reflects sound, then tumours would bemuch easier to spot and delineate.This is what ultrasonic elastography sets out to achieve. There areseveral flavours of elastography, but we're going to focus on onewhich involves taking a series of conventional ultrasound pictureswhile the clinician presses down with varying pressure. If we comparetwo images in the sequence, stiff structures (like tumours) won'tchange much, whereas less stiff structures will be deformed. Imageprocessing algorithms can look at the two images and deduce thedeformation of each bit of tissue. We can therefore build up a map ofthe tissue's elasticity.Clinicians can already purchase equipment offering real-timeelastography, but what they get are two-dimensional (2D) pictures,corresponding to slices through the anatomy, and not a 3D map of thetissue's elasticity. Unfortunately, without the 3D map, it isdifficult to plan surgery and monitor a tumour's response totreatment. This is where this research proposal comes in. It bringstogether internationally leading groups in the areas of ultrasonicelastography (London) and 3D ultrasound (Cambridge) with the goal ofdeveloping 3D ultrasonic elastography.The research will progress on parallel high and low risk paths. Thelow risk work will look at ways of recording a series of 2Delastograms, at closely packed locations in space, and then stackingthem together to make a 3D image. We could get the clinician to sweepthe probe over the area of interest, recording elastograms all thewhile: this is the freehand approach. Or we could use a special 3Dprobe, inside which the innards of a 2D probe are mounted on a rockermechanism driven by a stepper motor. In this mechanical approach, theclinician holds the probe still, while the motor sweeps the beam overthe target area. We will implement both approaches and compare theireffectiveness in terms of imaging quality and ease of use. We willalso look at ways of exploiting the 3D nature of the data to improvethe clarity of the elastograms. This low risk research will interfaceclosely with the project's clinical objectives, to evaluate 3Delastography in the context of cancers of the breast andbrain. Feedback from the collaborating clinicians is important if theengineers are to develop technology which could actually affect theeveryday management of cancer patients.Meanwhile, the high risk path will attempt to build more detailedelastograms by measuring tissue deformation in 3D. Currently,elastography algorithms assess tissue deformation only in thedirection of the applied pressure. However, the tissue actuallydeforms in all three dimensions, and by measuring this weshould be able to make better elastograms and glean moreclinically useful information about the material's properties. Butmeasuring 3D deformation is hard, mostly because we can only make highresolution measurements in the direction of the ultrasound wave'spropagation, which is perpendicular to the skin surface. Tomeasure deformation in other directions, we will need tocontrol the ultrasound scanner to steer the waves moretangentially. Our aim is to image each bit of tissue from differentdirections while the applied pressure is varied. We will then need todevelop algorithms to deduce the 3D deformation from this rich data.