Ellipse Fitting demo

Digital Pathology: Segmentation

Jun Kong, P.Y. Zhang, Y.H. Liang, G. Teodoro, D.J. Brat, F.S. Wang, Robust Cell Segmentation for Histological Images of Glioblastoma, IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI), pp. 1041-1045, Prague, Czech Republic, April, 2016. (Oral).
P.Y. Zhang, Y.H. Liang, G. Teodoro, F.S. Wang, D.J. Brat, Jun Kong, Automated Level Set Segmentation of Histopathologic Objects with Sparse Shape Prior Support and Dynamic Occlusion Constraint, IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI), pp.718-722, Melbourne, Australia, April 2017.
Jun Kong, F.S. Wang, G. Teodoro, Y.H. Liang, Y.Y. Zhu, C. Tucker-Burden, D.J. Brat, Automated Cell Recognition with 3D Fluorescence Microscopy Images, IEEE International Symposium on Biomedical Imaging: From Nano to Macro (ISBI), pp.1212-1215, Brooklyn, NY, April 2015. (Oral)
P.Y. Zhang, F.S. Wang, G. Teodoro, Y.H. Liang, M. Roy, D.J. Brat, Jun Kong, Effective Cell Segmentation with Sparse Shape Prior and Dynamic Occlusion Constraint for Pathology Images, SPIE Journal of Medical Imaging, 6(1), 017502, pp.1-15, 2019.

Schema of the fully automated cell segmentation method.
The cell seed detection algorithm utilizes joint information of spatial connectivity, distance constraint, image edge map, and a shape-based voting map derived from eigenvalue analysis of Hessian matrix across multiple scales.
Variational level set method with a new shape prior term, an adaptive contour occlusion penalty term, and a boundary attraction term.

we have applied our 2D method to a randomly selected 2D image slice of a 3D testing fluorescent imaging volume.
The ex-vivo 3D fluorescent cell segmentation results:

3D cell surfaces are projected to 2D image planes for qualitative cell segmentation assessment:
3D segmented cells are overlaid with a 2D image plane for qualitative cell segmentation assessment:
3D Hepatic artery segmentation.