MRI improves its image
New software for processing magnetic resonance images is being used to quantify brain anomalies in schizophrenics and to monitor multiple sclerosis and its treatment. Background Magnetic resonance imaging (MRI) is the best scanning technique available for viewing soft tissues such as those of the brain. It is essential to the diagnosis and management of many brain diseases. To fully exploit its potential in brain research, better image processing tools are needed to determine the exact size and shape of structures and lesions, to distinguish normal variability from the effects of disease, and to monitor lesions over time. In the `Biomorph' project, "Development and validation of techniques for brain morphometry", five European teams are developing such tools. They are validating their methods in clinical situations of major medical importance, and their work may lead to quicker, cheaper, testing of drugs which are potentially effective against brain diseases. Description, impact and results Biomorph focuses mainly on two major brain diseases, schizophrenia and multiple sclerosis (MS). A fundamental cause or property of schizophrenia may be a loss of asymmetry between the two sides of the brain, so algorithms are created to detect planes of (near) symmetry and to quantify that symmetry. In MS, both lesions and symptoms tend to appear and disappear. It is important to understand how changes in the lesions relate to disease progression. Here the emphasis is on automatic segmentation to determine the size of regions of abnormal intensity (lesions), and on detection of active, changing lesions in a series of images taken at different times. Compared with assessment by experts, these automated methods are much quicker and they may help cut the time and cost of testing drugs, such as interferon beta, which is known to reduce the rate of development of MS lesions seen on MR scans. The new methodology makes it possible to detect changing lesions with greater sensitivity and uses fewer patients to demonstrate therapeutic effects. Other diseases to which the new tools are applicable include Alzheimer's (shape and symmetry of brain structures), Creutzfeldt-Jakob Disease (intensity abnormalities), and brain tumours (size). Working partnerships Clinicians and computer scientists collaborate closely within Biomorph. The project includes psychiatrists (University of Oxford); experts in registration - positioning images so that they can be compared with others or referred to an anatomical map (INRIA, University of Leuven (KUL)); segmentation - determining the limits of visualised structures (KUL, University of Oxford); brain symmetry (INRIA, University of Oxford); and detecting changes over time (Eidgenossische Technische Hochschule). At the UK's University of Kent segmentation is evaluated and clinical and computer aspects coordinated. Outside collaborators include the State University of New York at Stony Brook, the University of North Carolina, and Guy's, Kings' and St. Thomas' School of Medicine in London.