Imaging and modelling the brain to help earlier dementia diagnosis

A major European project, led by the University of Sheffield, will develop sophisticated image-based biophysical models of the human brain to enable early diagnosis of dementia and help doctors predict the progression of the disease.

The model will be the first to combine a wide range of physiological data from medical imaging – such as blood flow, brain tissue properties and cellular activity – with psychological measures such as memory and cognitive function. It will also bring in demographic, genetic, lifestyle and environmental factors, making it much more sensitive than existing diagnostic tools.

The aim is to develop a way for doctors to identify at an early stage – before obvious symptoms appear – not only whether a patient has or is developing dementia, but which form of dementia it is. Diagnosis of the most common form of dementia – Alzheimer's – in the UK still takes an average of 32 months after symptoms are first noticed compared to 20 months in Europe as a whole.

''We currently rely on basic cognitive tests and conventional brain scans for diagnosis of dementia but these aren't good enough to recognise the disease at an early stage,'' explains the project's scientific director, Professor Alex Frangi, from the University's Faculty of Engineering.

''If we can identify a much wider range of markers which provide better and earlier diagnosis, then that gives doctors time to delay the progression of the disease. This would be not only improving patients' quality of life, but it would also reduce the burden on carers and the enormous costs of supporting people with dementia.''

Over 800,000 people have dementia in the UK alone, with numbers expected to rise to over a million by 2021. The Alzheimer's Society calculates the cost of dementia to the UK in 2012 was £23 billion.

Less than half of those with the disease in the UK are properly diagnosed. The engineers, scientists and clinicians who together hope to tackle this issue will draw on historical data from over 20,000 patients across Europe to build their model. In addition, three studies will gather new data from 150 patients to test hypotheses on contributory factors for the disease.

One of these studies will take place in Sheffield and will involve a new brain imaging technique which can identify changes in the elasticity and rigidity of brain tissue.

Iain Wilkinson, professor of magnetic resonance (MR) physics in the University's Faculty of Medicine, Dentistry and Health, explains, ''The temporal lobes, part of the brain affected in Alzheimer's, lose tissue as the disease progresses. Their rigidity or 'softness' is likely to alter as part of this process before the tissue loss occurs. A technique called MRI elastography, developed to scan for these properties in liver tissue, is being adapted by colleagues in the UK and Switzerland for use in the brain. We'll be applying this technique with patients in Sheffield, to see if it could be used in the model to help earlier diagnosis.''

The research will also take into account the possible impact of other diseases on the onset of dementia, particularly those such as diabetes and lung disorders.

''Changes in the vascular system, such as the amount of blood flow from the brain, may be a factor in dementia,'' says professor of clinical translational neuropsychology, Annalena Venneri, also from the Faculty of Medicine, Dentistry and Health. ''These changes can be caused by diabetes or respiratory illnesses – so we want to see if we can use this in diagnosing dementia. The strength of our model is that it will be able to draw useful information from so many different variables and factors to make the diagnosis more definitive and more specific.''

The four-year, €18 million project is funded through the European Union and involves universities and industrial partners from the UK, France, Germany, Switzerland, Finland, Spain, Norway, Austria, the Netherlands and Portugal. The research in Sheffield will involve the University's Faculty of Engineering and Faculty of Medicine, Dentistry and Health, in collaboration with the Sheffield Teaching Hospitals NHS Foundation Trust.

The project is one of several underway at the University of Sheffield aiming to create a 'virtual physiological human' – a computer model of a patient which can be personalised to test potential treatments. The University's INSIGNEO Institute for In-Silico Medicine is one of the major European hubs for this field of research.