Magnetic Resonance Pattern Recognition in White Matter Disorders

J Valk

Hong Kong J Radiol 2005;8:136-40

Magnetic resonance pattern recognition constitutes a ‘conceptual’ approach to radiological assessment, in which all of the radiological and other clinical features present are considered in combination. Systematic analysis of magnetic resonance images improves the specificity of the modality as a diagnostic tool. Computer-aided analysis of radiological findings can be used to complement clinicians’ knowledge. In recent years, use of magnetic resonance techniques has facilitated characterization of several new forms of white matter disorders in children. Pattern recognition software aided the identification of common features among unclassified leukoencephalopathies, resulting in 7 major categories and 2 newly described entities, megaencephalic leukencephalopathy with cysts and vanishing white matter disease. Magnetic resonance imaging has a central role in diagnosis and clinical management.