Image processing in the neurologic sciences

Imaging modalities like MRI and PET play an important role in neurologic diagnostics. More than just displaying the examination data as an image, a computer-aided processing can reveal quantitative descriptions of tissue lesions and properties, the nature of the participating structures and their changes in time. On a quantitative basis, more profound statements about structure function relationships and therapeutic implications are possible. Here, the authors describe the implementation of a clinical-wide multimedia patient database system. Emphasis is laid on the description of analysis and display facilities of 3D volume data. Since these processes still require much time-intense (expert) user interaction, the authors' current work is devoted towards a more automated preprocessing of volume datasets. The outline of a knowledge-based framework and its connected algorithms for automatisation of this processing chain are discussed.<
>