| 摘要 |
The invention relates to a method for determining the contours of an object during imaging analysis methods, according to which the object is scanned from several projection angles and a three-dimensional representation of the analysis object is reconstructed from the scanned values. Volume data, which represents the analysis object, is obtained in said imaging analysis, a threshold value for the reconstruction values is defined to determine the position of the contour of the three-dimensional object representation and the position of the contour points is determined. A first threshold value (G<SUB>i</SUB>) of the reconstruction values is defined and to determine the position of a contour point a respective first contour point (K<SUB>i</SUB>) of the analysis object is determined. A profile line, which stretches from the exterior of the analysis object into the latter and which runs through the first contour point, is also defined and reference points are assigned to said line. The corresponding reconstruction values are determined along the profile line and a new local threshold value (G<SUB>1</SUB>) of the scanning values is determined from the gradient of the latter along and/or adjacent to the profile line in the vicinity of the first contour point. The first contour point on the profile line is then shifted to a second contour point (K<SUB>1</SUB>) at the site of the local threshold value. |
