The Publication Database hosted by SPL

All Publications

Upload

Tubular Surface Segmentation for Extracting Anatomical Structures from Medical Imagery

Mohan V.¹, Sundaramoorthi G.², Tannenbaum A.¹


Institution:	¹Georgia Institute of Technology, School of Electrical and Computer Engineering, 777 Atlanta Drive NW, Atlanta, GA 30332-0250 ²Computer Science Department, University of California, Los Angeles, CA, USA
Publication Date:	May-2010
Citation:	IEEE Transactions in Medical Imaging
Keywords:	Blood Vessel Segmentation, fiber bundle segmentation, Shape Prior, tubular shape, branch detection, Projects:TubularSurfaceSegmentation
Appears in Collections:	NA-MIC, NAC
Generated Citation:	Mohan V., Sundaramoorthi G., Tannenbaum A. Tubular Surface Segmentation for Extracting Anatomical Structures from Medical Imagery. IEEE Transactions in Medical Imaging
Downloaded:	577 times. [view map]
Paper:	Download, View online
Export citation:

This work provides a model for tubular structures and trees, and devises the algorithm to automatically extract (starting with a single seed point) anatomical structures from medical imagery that fit the model. Our model fits many anatomical structures in medical imagery, in particular, various fiber bundles in Diffusion-Weighted Magnetic Resonance Imagery (DW-MRI) of the brain such as the cingulum bundle, and blood vessel trees in computed tomography angiograms (CTA). Extraction of the cingulum bundle is of interest because of possible ties to Schizophrenia, and extracting blood vessels are helpful in the diagnosis of cardiovascular diseases. The tubular model we propose has advantages over many existing approaches in literature: fewer degrees of freedom over a general deformable surface that energies defined on such tubes are less sensitive to undesirable local minima, and the tube (in 3D) can be naturally represented by a 4D curve (a radius function and centerline), which leads to computationally less costly algorithms and has the advantage that the centerline of the surface is obtained without additional effort. Our model also generalizes to tubular trees, and the extraction algorithm that we design automatically detects and evolves branches of the tree. We demonstrate the performance of our algorithm on 20 datasets of DW-MRI data and 32 datasets of CTA, and quantify the results of our algorithm when expert annotations are available.

Additional Material

1 File (221.395kB)

Mohan-TMI2010-fig4.jpg (221.395kB)