Quick Review of - S. You, L. Hong, M. Wan, K. Junyaprasert, A. Kaufman, S. Muraki, Y. Zhou, M. Wax, and Z. Liang (1997). "Interactive Volume Rendering for Virtual Colonoscopy ". Proc. IEEE Visualization '97, 433-436.

    This paper examines the basics of their interactive viewing system in addition to the system they used to achieve a realtime virtual colonoscopy, and an explanation of their optimized volume renderer. The problem present in this paper is similar to the problem I worked in my MQP. When a patient has a tumor in their colon or trachae the physican sends a camera attached to an optical camera into the tubes which is called a colonoscopy or bronchoscopy depending on where the camera is going. This procedure has a number of faults to it such as patient discomfort, high cost, can create comlications, and the optical camera can only see straight ahead. The alternative discussed in this paper is a virtual colonoscopy which consists of taking a CT scan of the patient and using doing a virtual fly through of that field data. Although this procedure is still slow, not user friendly enough for a physician, has a limited resolution, and limited colors, it is still a very promising procedure.

    To collect the data, a set of  2D CT images were made and combined into a 3D volume set and the surface is determined using the Marching Cubes algorithm. In order to make the visualization realtime, the authors implemented two existed techniques: surface-assisted ray casting and parallelization. Surface-assisted ray casting is the process in which ray-casting through empty space is avoided. Since the colon is tubular, there is a lot of empty space that can be avoided. Using this algorithm saves a lot of unnecessary processing time. Parallelization was simply implementing the algorithm using multiple processors.
 
    The equipment used here was a SGI Power Challenge equipped with 16 R10000 processors using the Visible Male Data set giving an average of 1frame/sec for a 512x512 volume set, 4 frames/sec for a 256x256 volume, and 16.6 frames/sec for a 128x128 volume set.
 
Virtual Colonoscopy by Visualization Lab at SUNY at Stony Brook
Virtual Colonoscopy by Bowman Gray School of Medicine
Virtual Bronchoscopy by WPI Sydney MQP Team
Virtual Bronchoscopy by WPI Sydney MQP Team
Actual Colonoscopy by Virtual Endoscopy Center
 
 

[Return to cs563 '99 Biomedical Visualization Talk]