| 摘要 |
In the method according to at least one embodiment of the invention, an image data record having a structure to be segmented is first of all displayed by display equipment. Using an input apparatus, a segmentation algorithm to be used is selected from a group of different segmentation algorithms, including a contour-based segmentation algorithm, a region-based segmentation algorithm and manual segmentation, based on the local image contrast in a region to be segmented in the image data record. A region to be segmented in the image data record is marked, and the structure to be segmented in the marked region is segmented using the selected segmentation algorithm, and a segmentation result of the segmentation is displayed. This procedure (selecting a segmentation algorithm/marking a region/segmenting the region/displaying) is repeated until the structure to be segmented is completely segmented in the displayed image data record and a boundary line of the structure is produced as the final segmentation result. Lastly, the final segmentation result is saved and/or displayed. Furthermore, an image processing unit is disclosed for carrying out the method of at least one embodiment.
|
