Method for the fully automatic segmentation of an organ, in particular of the renal parenchyma, of volume data sets of the medical imaging

摘要

The present invention relates to a process for the segmentation of an organ, in particular of the renal parenchyma, of volume data etching of the medical imaging. On the basis of segmentation results of training data etching for this purpose is a probability data set is generated, on the basis of its intensity in the retentate values each voxel of the volume data set is associated with a probability, to the organ to include. With the aid of a threshold value method, a binary data set is produced, in which the organ region to be eroded in a stepwise fashion. In the case of a splitting of the organ region in a plurality of portions, it is in each case only one of the partial areas, of the container with high probability the organ, selected for the further erosion or eroding. This partial area, on the basis of geometrical features of the organ recognized, the etching from the training data are known. In the probability data set are then the probability values of voxels of the in the binary data record split off in the partial regions of reduced, so that a corresponding corrected probability data set is obtained. On the basis of this corrected probability data set is then the member from the probability data set segmented. With the proposed method, the segmentation of organs from native mr - data records with a high degree of reliability makes it possible to.