Information Technology Reference
In-Depth Information
Fig. 5.5 Segmentation results obtained from the first stage of the proposed algorithm. a 3D
surface reconstruction of the coronary arteries with a touching non-arterial vessel (shown by the
arrow). b The cross-sectional image taken from the volume. The closed up image shows the
intensity characteristics of the vessel in an axial image. The non-arterial vessel (arrowed) appears
darker than the artery
Fig. 5.6 3D reconstructed
image of the coronary arteries
and part of the aorta. a and
b are 3D surface rendering
images reconstructed from
different view of points,
respectively. They have
shown that the arteries
originate from the aorta
3. Evolve the curves in terms of the energy function defined in Eq. 5.9 . To speed
up convergence, the segmentation mask obtained from this slice is used as an
initial estimate in the following slice.
4. Detection of non-arterial objects. We compare the initialization against the
segmentation mask obtained from the first stage. If there is a connected
component which does not touch any known components in the initialization
mask, then the object is considered as a non-arterial component, and a level set
is assigned to it.
5. If the area of the level sets is less than one pixel, we remove such level set and
decrease the number of level sets functions.
6. If the current frame is not the last frame containing coronaries, go to step 3.
Search WWH ::




Custom Search