Invisible bifurcation detection within vessel tree images

1. An image registration system, comprising:

an endoscope ( 12 ) operable for generating an intra-operative endoscopic image ( 14 ) of a vessel tree within an anatomical region,

wherein the intra-operative endoscopic image ( 14 ) of the vessel tree includes a plurality of branches of the vessel tree visible within the intra-operative endoscopic image ( 14 ) as an indication of a furcation of the vessel tree invisible within the intra-operative endoscopic image ( 14 ); and

an endoscope controller ( 22 ) operable for image registering the intra-operative operative endoscopic image ( 14 ) of the vessel tree to a pre-operative three-dimensional image ( 44 ) of the vessel tree within the anatomical region,

wherein the image registration includes an image matching of a graphical representation of the furcation of the vessel tree as indicated by the visible branches of the vessel tree within the intra-operative endoscopic image ( 14 ) of the vessel tree to a graphical representation of the furcation of the vessel tree within the pre-operative three-dimensional image ( 44 ) of the vessel tree.

2. The image registration system of claim 1 , wherein the image matching includes:

generating a main graph derived from a geometrical representation of the pre-operative three-dimensional image ( 44 ) of the vessel tree, the main graph including a main set of nodes representative of each furcation of the vessel tree within the pre-operative three-dimensional image ( 44 ) of the vessel tree;

generating a subgraph derived from a geometrical representation of the intra-operative endoscopic image ( 14 ) of the vessel tree, the subgraph including a subset of the main set of nodes, wherein a first node of the subset is representative of the furcation of the vessel tree as indicated by the visible branches of the vessel tree within the intra-operative endoscopic image ( 14 ) of the vessel tree; and

node matching the subgraph to the main graph.

3. The image registration system of claim 2 , wherein the image matching of the subgraph to the main graph includes:

establishing at least one of a vertical ordering and a horizontal ordering of the nodes in the main graph.

4. The image registration system of claim 1 , wherein the image registration further includes:

delineating a polygonal area on the intra-operative endoscopic image ( 14 ) enclosing the furcation of the vessel tree invisible within the intra-operative endoscopic image ( 14 ).

5. The image registration system of claim 4 , wherein a delineation of the polygonal area includes:

marking each branch of the vessel tree visible within the intra-operative endoscopic image ( 14 ) as a branch segment;

extending rays from each branch segment; and

delineating overlapping rays of different branch segments.

6. The image registration system of claim 4 , wherein the image registration further includes:

defining a furcation candidate within the polygonal area as a detection of the furcation of the vessel tree invisible within the intra-operative endoscopic image ( 14 ).

7. The image registration system of claim 6 , wherein a definition of the furcation candidate within the polygonal area includes:

identifying a center of the polygonal area as the furcation candidate.

8. The image registration system of claim 6 , wherein a definition of the furcation candidate within the polygonal area includes:

identifying a point within the polygonal area as the furcation candidate based on an anatomy of the anatomical region.

9. The image registration system of claim 6 , wherein a definition of the furcation candidate within the polygonal area includes:

identifying a point within the polygonal area as the furcation candidate based pre-operative three-dimensional image ( 44 ) of the vessel tree.

10. The image registration system of claim 1 ,

wherein the image registration further includes:

selecting all visible furcations within the intra-operative endoscopic image ( 14 ); and

wherein the image matching includes:

generating a geometrical representation of the intra-operative endoscopic image ( 14 ) of the vessel tree utilizing the visible furcations and the invisible furcation.

11. The image registration system of claim 10 , wherein the image registration further includes:

delineating a polygonal area on the intra-operative endoscopic image ( 14 ) enclosing the invisible furcation of the vessel tree, wherein the visible furcations are external to the polygonal area.

12. The image registration system of claim 1 , the endoscope controller ( 22 ) is further operable for refining an image registration of the intra-operative operative endoscopic image ( 14 ) of the vessel tree to a pre-operative three-dimensional image ( 44 ) of the vessel tree based on a scale of the vessel tree within the pre-operative three-dimensional image ( 44 ) of the vessel.

13. The image registration system of claim 1 , wherein a refinement of the image registration includes:

utilizing the scale of the vessel tree within the pre-operative three-dimensional image ( 44 ) of the vessel to scale of a shape of each branch of the vessel tree visible within the intra-operative operative endoscopic image ( 14 ); and

utilizing a scale a shape of each branch of the vessel tree visible within the intra-operative operative endoscopic image ( 14 ) to execute a point-based registration of the intra-operative endoscopic image ( 14 ) of the vessel tree to the pre-operative three-dimensional image ( 44 ) of the vessel tree.

14. The image registration system of claim 1 , wherein the endoscope controller ( 22 ) is further operable for overlaying the geometrical representation of the pre-operative three-dimensional image ( 44 ) of the vessel tree onto the intra-operative endoscopic image ( 14 ) of the vessel tree in accordance with the image registration.

15. The image registration system of claim 1 , wherein the vessel tree is one of an arterial tree and a venous tree.