29 April 2005 Semiautomatic segmentation of textured laser range scans for use in image-guided procedures
Author Affiliations +
Laser range scanners produce high resolution surface data of anatomic structures, which facilitates the determination of intraoperative soft tissue deformation and the performance of surface based image-to-physical space registration. Segmentation of the range scans is required for the data to be effectively incorporated into current image-guided procedures. Due to time constraints in the operating room, manual segmentation methods are not feasible. We propose a novel segmentation algorithm based on the level set method that uses information from the texture map and curvature of the acquired point cloud to provide an accurate edge map for computation of the speed image. Specifically, the edge image is created by combining the curvature values, computed from a surface fitted to the acquired point cloud using radial basis functions, and gradients of the RGB intensities in the texture map. Preliminary results, obtained from comparing the semiautomatic segmentations of intraoperatively acquire liver LRS data with manual gold standard segmentations, shows the method to be a significant first step towards the implementation of semiautomatic LRS segmentation routine during image-guided surgery.
© (2005) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Logan W. Clements, Logan W. Clements, David Marshall Cash, David Marshall Cash, Tuhin K. Sinha, Tuhin K. Sinha, Robert L. Galloway, Robert L. Galloway, "Semiautomatic segmentation of textured laser range scans for use in image-guided procedures", Proc. SPIE 5747, Medical Imaging 2005: Image Processing, (29 April 2005); doi: 10.1117/12.595814; https://doi.org/10.1117/12.595814


Back to Top