Automatic Liver Classification with Multi-Slice CT Data

Publication, 2007


G. Zwettler, W. Backfrieder, R. Swoboda, F. Pfeifer, H. Kratochwill, F. Fellner - Automatic Liver Classification with Multi-Slice CT Data - Tagungsband des ersten Forschungsforum der österreichischen Fachhochschulen, Fachhochschule Salzburg, Campus Urstein, Austria, 2007, pp. 425-426


Multi slice CT (MSCT) effected major improvement in spatial and temporal resolution of radiologic diagnosis. Accurate high contrast imaging of liver parenchyma allows early diagnosis of even small-sized liver tumors, improving prognoses of therapy. Modern imaging methods for liver classification, automated segmentation of entire liver vessel tree and lesions of the parenchyma benefit from state of the art computerized methods like level sets. Liver lobe classification is performed according to Couinaud utilizing graph analysis. Information about the position of the lesions and their size is used for operation planning and monitoring of therapy progress.