Automatic crown and gingiva detection from the three-dimensional virtual model of teeth
A method is provided for automatically separating tooth crowns and gingival tissue in a virtual three-dimensional model of teeth and associated anatomical structures. The method orients the model with reference to a plane and automatically determines local maxima of the model and areas bounded by the local maxima. The method automatically determines saddle points between the local maxima in the model, the saddle points corresponding to boundaries between teeth. The method further positions the saddle points along a dental arch form. For each tooth, the method automatically identifies a line or path along the surface of the model linking the saddle points to each other, the path marking a transition between teeth and gingival tissue and between adjacent teeth in the model. The areas bounded by the lines correspond to the tooth crowns; the remainder of the model constitutes the gingival tissue.
1. A method for automatically detecting a ravine between a first saddle point and a second saddle point in a virtual three-dimensional dentition model in a dental arch; wherein said model is in the form of a single mesh comprising points interconnected by triangles, comprising the steps of:
a) storing said model in a memory accessible to a general-purpose computer, said computer including a processing unit;
b) providing machine readable instructions for execution by said processing unit, said instructions comprising instructions operating on said model in the following respects:
1) performing a tree search starting from said first saddle point along paths traversing edges of said triangles;
2) determining a quality for each of said edges along said paths; and
3) selecting a path which reaches said second saddle point, wherein said selected path has the highest quality at said second saddle point; wherein the quality of a path of the search tree is obtained by adding the qualities of all edges from said first saddle point to the end of the edge—multiplied by their lengths—and dividing them by the total length of the path;
wherein said selected path is the ravine between said first saddle point and said second saddle point.
2. The method of claim 1 , wherein said quality of said edge is the curvature of said edge.
3. The method of claim 2 , wherein said curvature is measured in terms of concavity of said edge.
4. The method of claim 1 , wherein said quality of said edge is the angle formed between the edge and the vector field at the starting point of the edge.
5. The method of claim 1 , wherein said tree search is terminated after a specified amount of computation time performed by said general-purpose computer.
6. The method of claim 1 , wherein said path comprises a series of branches, and said tree search along a branch is not continued if another branch leading to the same branch ends but one with a better quality has been found.
7. The method of claim 1 , wherein said path comprises a series of branches and said tree search along a branch is not continued if a maximum length is exceeded, a minimum quality is fallen short of, or a certain environment related to said first saddle point or said second saddle point is departed from.
8. The method of claim 1 , wherein said ravine is used to identify tooth crowns.
9. The method of claim 1 , wherein said ravine is used to identify gingival tissues.