Aims: Our research project aims to create a powerful new tool for detecting and predicting brain and abdominal diseases like Alzheimer’s, Parkinson’s, liver cirrhosis, and different types of cancers, much earlier than we can today. We’re doing this by combining modern artificial intelligence (AI) with a huge collection of medical data from the UK Biobank. This data includes detailed scans of people’s brains and abdomens, their genetic information, and lots of other health records.
Scientific rationale: The reason behind our work is simple: these diseases can sneak up on people, and catching them early is key to better outcomes. But to do that, doctors often rely on specialized medical imaging, which can be tricky. At the same time, AI has made huge strides in analyzing medical images, showing it can help spot diseases earlier.
The UK Biobank gives us a golden opportunity – access to a treasure trove of information from many thousands of individuals. We’ll use this data to train our AI model to recognize patterns in brain and abdominal scans that indicate disease, even in its very early stages. We’ll also connect these image patterns with genetic data to understand how they relate to the biology of the diseases. Plus, we’ll mix in blood test results to make our predictions even more accurate.
Project duration: We’re planning to spend three years developing and perfecting this AI model, making sure it works really well.
Public health: The impact on public health could be massive. Once we’ve got our model working perfectly, we’ll make it freely available for anyone to use. This means doctors could diagnose these diseases faster and start treatment sooner, leading to better treatments working more effectively, and lower healthcare costs from treating advanced stages of diseases. Ultimately, we want our model to become a tool that helps predict, prevent, and personalize medicine for millions of people around the world, making their lives healthier.