Method for teeth segmentation and alignment detection in CBCT volume
View Patent ↗A method of automatic tooth segmentation, the method executed at least in part on a computer system acquires volume image data for either or both upper and lower jaw regions of a patient and identifies image content for a specified jaw from the acquired volume image data. For the specified jaw, the method estimates average tooth height for teeth within the specified jaw, finds a jaw arch region, detects one or more separation curves between teeth in the jaw arch region, defines an individual tooth sub volume according to the estimated average tooth height and the detected separation curves, segments at least one tooth from within the defined sub-volume, and displays the at least one segmented tooth.
1. A method of automatic tooth segmentation, the method executed at least in part on a computer system and comprising:
acquiring volume image data for either or both upper and lower jaw regions of a patient;
identifying image content for a specified jaw from the acquired volume image data and, for the specified jaw:
(a) estimating an average tooth height for teeth within the specified jaw;
(b) finding a jaw arch region;
(c) detecting one or more separation curves between teeth in the jaw arch region;
(d) defining at least one individual tooth sub volume according to the estimated average tooth height and the detected separation curves; and
(e) segmenting at least one tooth from within the defined sub-volume; and
displaying, storing, or transmitting the volume image data for the at least one segmented tooth.
2. The method of claim 1 wherein segmenting the at least one tooth comprises generating foreground and background seeds for the at least one tooth.
3. The method of claim 1 further comprising defining and displaying a centroid position for the at least one segmented tooth.
4. The method of claim 3 further comprising defining and displaying two or more additional centroid positions for segmented teeth in the specified jaw and highlighting one or more displayed centroid positions that are indicative of tooth misalignment.
5. The method of claim 3 wherein defining and displaying two or more additional centroid positions comprises identifying an inertia center of at least a portion of the segmented tooth.
6. The method of claim 3 wherein defining and displaying two or more additional centroid positions comprises identifying a geometric center of at least a portion of the segmented tooth.
7. The method of claim 4 further comprising identifying tooth misalignment according to a convex hull conformance of the displayed centroid positions.
8. The method of claim 1 wherein acquiring volume image data comprises acquiring cone beam computed tomography image data.
9. The method of claim 1 wherein finding the jaw arch region comprises using adaptive thresholding.