Robust and fully automated segmentation of mandible from CT scans
2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017), pp. 1209–1212
Abstract
Mandible bone segmentation from computed tomography (CT) scans is challenging due to mandible's structural irregularities, complex shape patterns, and lack of contrast in joints. Furthermore, connections of teeth to mandible and mandible to remaining parts of the skull make it extremely difficult to identify mandible boundary automatically. This study addresses these challenges by proposing a novel framework where we define the segmentation as two complementary tasks: recognition and delineation. For recognition, we use random forest regression to localize mandible in 3D. For delineation, we propose to use 3D gradient-based fuzzy connectedness (FC) image segmentation algorithm, operating on the recognized mandible sub-volume. Despite heavy CT artifacts and dental fillings, consisting half of the CT image data in our experiments, we have achieved highly accurate detection and delineation results. Specifically, detection accuracy more than 96% (measured by union of intersection (UoI)), the delineation accuracy of 91% (measured by dice similarity coefficient), and less than 1 mm in shape mismatch (Hausdorff Distance) were found.
Abstract sourceThis paper presents a two-stage method for locating and delineating the mandible in computed tomography scans, evaluated on the MICCAI 2015 Head-Neck segmentation challenge dataset.
Summary
The study separates mandible segmentation into recognition and delineation. Random forest regression first localizes the mandible in three dimensions; gradient-based fuzzy connectedness then delineates it within the localized region, followed by a state-machine heuristic to refine potential leakage near the teeth and skull joints. The authors evaluated the method on the MICCAI 2015 Head-Neck segmentation challenge dataset. They report detection accuracy above 96% by union of intersection, a Dice similarity coefficient of 91%, and less than 1 mm of shape mismatch by Hausdorff distance.
The authors note that performance was not tested on scans with fractures or bone loss in or near the mandible, and that the anatomy-informed refinement may not generalize to congenital disease or trauma. These results describe an evaluation on a challenge dataset; they do not establish prospective clinical validation or clinical readiness.
Cite this work
Neslisah Torosdagli, Denise K. Liberton, Payal Verma, Murat Sincan, Janice Lee, Sumanta Pattanaik, Ulas Bagci (2017). Robust and fully automated segmentation of mandible from CT scans. 2017 IEEE 14th International Symposium on Biomedical Imaging (ISBI 2017), pp. 1209–1212 https://doi.org/10.1109/ISBI.2017.7950734
BibTeX