For people with insufficient hearing to effectively use hearing aids, cochlear implants can provide great benefit, although they do not restore normal hearing. Cochlear implants stimulate the hearing nerve directly with electrical currents from a set of electrodes surgically implanted into the inner ear (cochlea). Although cochlear implants are a great success, a wide range in speech perception outcomes are observed. The aim of this programme of research is to reduce the variability in performance as well as increasing the levels of speech recognition achieved by all. We propose to achieve this by using a special machine to measure brain activity in response to sound for cochlear implant users. This technique is used routinely in clinical practice and involves wearing a special hat to make the recordings. We will use this method to record responses at different parts of the hearing system between cochlea and brain. By understanding the relationship between these responses and speech understanding it will help us to appreciate the importance of different stages in the hearing pathway. These measures will allow us to observe the brain changes that take place once cochlear implants are switched on. We will define 'biomarkers' that can be used by clinicians to guide implant fitting and rehabilitation. A biomarker is an objective measurement of typical changes in the body. We will develop guidance for adjusting the ways that sounds are sent through the cochlear implant or for switching off electrodes. For those with bilateral cochlear implants we will also develop guidelines for adjusting the implants to make the sounds more similar in each ear to improve the way that the ears work together. This research will be conducted with both adults and children to fully understand how people hear with a cochlear implant.

One in 17 people have a rare disease. Rare diseases can be extremely difficult to diagnose, but they often have an unidentified genetic cause. Recent advances in clinical imaging, pathology, and genomic technologies have led to remarkable progress in understanding disease - particularly rare diseases. However, the power of these technologies cannot be fully realised until the immense volume of data generated can be integrated with NHS data, then analysed by researchers in a secure environment that protects the privacy of individuals. Working across the NHS, academia and industry we will use existing tools to transfer data from NHS Trusts to a secure environment that interfaces with the NHS network and shares data with Public Health England. NHS information will then be combined with research data in a cloud-based platform. Initially, we will involve patients with rare diseases recruited to the NIHR BioResource; a national resource of volunteers who have already provided consent that information retrieved from their health records can be used for medical research. This will create a rich research resource with the potential to transform our understanding of rare genetic disorders, drive improvements in diagnosis and management, and provide proof of principle for use in other diseases.

"Crohn's Disease and Ulcerative Colitis are the main forms of IBD. They cause debilitating symptoms affecting 0.78% of the UK population (500,0001 people), and costing UK health budgets approximately £1.5 Billion2 each year. Treatment is with steroids, immunosuppressants and antibody therapies, but results are variable. Over 70% of patients with Crohn’s and 15% with colitis require major surgery3. There is an urgent need to better understand why patients respond differently to treatments in order to improve outcomes and reduce costs. Recent advances in clinical imaging, pathology, and genomic technologies have produced remarkable progress in understanding IBD. However, the power of these technologies cannot be clinically realised until these data can be combined and used in a meaningful way. Our DIH will integrate data from multiple sources and create a secure research resource that allows approved researchers to access data, whilst protecting the privacy of individuals. Patient and public involvement is key to our success. 25,000 IBD patients have already provided consent for their health records to be retrieved and used for medical research. Working together, we will transform our understanding of IBD, drive improvements in diagnosis and treatment, and deliver a data framework to reproduce in other disease areas."

Delivery of pathology tissue diagnoses, most of which are cancer, in the current format is unsustainable. Advances in genomic medicine and immune-oncology have shown that the classification of tumours into subtypes allows selection of patients for specific treatments but also spares patients unnecessary toxic and expensive therapies. Still, making such diagnoses has become more time-consuming, involving the selection and interpretation of ancillary tests which requires an ever-growing specialist knowledge for each cancer type. Whilst the need for diagnostic expertise is increasing, there is already a shortfall of 25% of pathologists who are able to report results: this is set to decline. We propose that the use of AI can ensure that the delivery of tissue diagnoses by pathologists is sustainable and supports delivery of personalised treatments. The benefits of AI in pathology are beginning to be seen, e.g. identification of high-grade areas of prostate cancer shows a reduction in errors and pathologists' time. The development of AI for diagnoses is timely as full adoption of digitised histological images, allowing them to be interrogated by both humans and artificial intelligence (AI), is expected in the UK by 2025. AI is a data-hungry process; it is unrealistic to provide 100,000s images that are required to train a model. Even the most common cancers (e.g. breast) have multiple subtypes; identification of these is required for selection of patients for personalised treatments. To address this challenge, we propose to develop a novel AI strategy using a relatively small sample size (~1000 images per class). Such a model could be adapted to any cancer type. A multiple-instance learning framework will be developed, using transformers for feature extraction and classification. A tool that flags samples that cannot be confidently classified will be applied thereby alerting the pathologist of potentially unseen diseases. The deep learning model will be strengthened by the injection of pathologists' domain knowledge. Soft tissue and bone tumours We will develop the AI model on tumours of soft tissue (muscle, fat, blood vessels, etc.) and bone, an area considered to be one of the most challenging diagnostically. These tumours comprise approximately 100 different subtypes, and represent some of the most common cancers in children and young adults. We will build on our existing deep learning model of 15 different subtypes trained on 2122 images, which predicts the correct diagnosis in 87% of cases. Selection of confirmatory ancillary tests is then prompted by the algorithm and streamlines the diagnostic pathway. 17,000 images that have already been scanned will be added to the library and allow the rapid development and extension of the classification model. The image library will be linked to clinical outcomes and expanded to 35,000 images during the project. Added to this is the commitment of the established Sarcoma Network of at least 20 pathologists from across all countries in the UK, to provide the additional 20,000 images mentioned above. Additional benefits The study and infrastructure will serve as the framework for the continued development of the model which can rapidly be expanded prospectively with the introduction of digital pathology in the NHS and globally. The model can be developed over time in response to new advances. The image library will be available for training future pathologists, research, validation of other AI algorithms, and contribute to the Sarcoma Genomics England Clinical Interpretation Partnership (GeCIP) offering a valuable resource for future multi-modal multi-omic research. Working closely with Sarcoma charities, and partners, we will involve and engage patients, their families, and the public, to build trust in the use of AI in health care. Development of AI models for digitised pathology images can avert the crisis facing this medical specialty.

Many aspects of a child or young person's life can affect their mental health. If someone has a serious mental health problem their general practitioner (GP) may refer them to mental health (psychiatry) services for assessment and treatment by professionals. Mental health services are stretched so often intervene late, leaving people to suffer unnecessarily with problems that therefore may last longer, be more severe, or be harder to treat. Early warning signs of mental health problems may be noticed by the person themselves or by others (e.g. school staff, social workers). Many things can suggest a mental health problem, such as difficult early experiences, bullying, changes in behaviour, poor school attendance or grades, or risk-taking. Not all who experience one or more of these will have a mental health problem, so we need to take them together to spot patterns that show who is developing problems and may need professional help. However, this information (data) is stored in different places, e.g. by schools, GPs and social workers and so it may be impossible to spot problems early. Some researchers have joined data from two or more sources to find patterns suggesting mental health problems. Their success indicates good potential in this approach, but they have not made a practical difference for two main reasons: 1) the models are not yet accurate enough, probably because they omit many factors that can lead to problems; 2) the results cannot be used directly to help young people as they are based on anonymous data. We will develop a system that can be used by health, education, or social workers to identify adolescents showing early signs of mental health problems, to offer them help sooner. At the same time we want to provide better anonymous data for research into predicting mental health problems. Data must be held securely (most likely in the NHS), and only people involved in a person's care should be able to see it, but we need to understand how best to do this. To use data for research while protecting privacy it will be anonymised, removing anything that directly identifies a person (e.g. name, address, date of birth, NHS number) and access will be restricted to approved researchers. But we do not yet know what technical problems there may be in linking the databases, or what data the system will need in order to detect people showing early signs of a problem. The final challenge is how to make this work within the NHS, schools, and social care settings to enable earlier identification of young sufferers of mental health problems. Over the next year, we want to tackle these challenges by creating a group including mental health researchers, psychologists, schools, the NHS, councils, computer scientists, security experts, mathematicians, people who provide services, and policy makers, many of whom are doing ground-breaking work in other areas. We want to turn their attention to jointly solving these problems. We must involve young people, their carers, and people with lived experience: it is their data and we need to understand their views. We would like their help thinking about which professionals can see their data, and what should happen when a young person is thought to be developing mental health problems. We will hold workshops about these questions. We also have permission to create an initial data set with data from health, social services, and education. We will anonymise these, and practise linking and analysing them. These will help us understand the challenges, so that our final plan will be more detailed and likely to succeed. In the future we want to test if a computer program makes it easier to identify mental health problems and offer young people treatments earlier, and if they get better quicker because of this. This might have a range of benefits including helping with school, relationships, home life, and getting jobs or into university, and we want to test this theory.