Technology

The following text is an extract from "A Parallel PC-Based Visible Human Slice Web Server", presented at the Visible Human Conference 98. This article describes the slice extraction server used by the Slice, Surface and Animation extraction applet. For the complete article, as well as other articles related to the Visible Human Server published by the EPFL Peripheral Systems Lab, please consult the bibliography below.

Visualization of 3D tomographic images by slicing, i.e. by intersecting a 3D tomographic image with a plane having any desired position and orientation is a tool of choice both for learning and for diagnosis purposes.

A parallel Visible Human Slice Web server has been developed, which offers to any Web client the capability of interactively specifying the exact position and orientation of a desired slice and of requesting and obtaining that slice from a 3D tomographic volume, made of either CT, MRI or cryosection images (digital color photographs of cross-sections).

For enabling its parallel storage and access, the Visible Human 3D volume is segmented into volumic extents of size 32x32x17 RGB voxels, i.e. 51KBytes, which are striped over a number of disks possibly residing on several server PC's. In order to extract an image slice from the 3D image, the extents intersecting the slice are read and the slice parts contained in these volumic extents are extracted and projected onto the display space (Figure 1).

Figure 1. Extraction of slices from the Visible Human dataset

The parallel Visible Human slice server application was synthesized using the CAP Computer Aided Parallelization tool [1] and the library of reusable parallel file system components [2] which enable building pipelined-parallel applications combining high-performance computing and I/O intensive operations. The parallel server architecture has been interfaced to a Netscape Web Server using FastCGI. A Java 1.1 applet runs on Web clients and enables users to specify slice position and orientation and generate image slice extraction requests. Replies of the Web server are compressed using the JPEG standard and send back to the Web clients for display.

The parallel slice server application consists of a server interface residing on the Web server PC and of server processes running on the server's parallel PC's. The server interface interprets the slice location and orientation parameters defined by the user and determines the image extents which need to be accessed. It sends to the concerned servers (servers whose disks contain the required extents) the extent reading and image slice part extraction requests. These servers execute the requests and transfer the resulting slice parts to the server interface, which asssembles them into the final displayable image slice (Figure 2).

Figure 2. General schematic for slice extraction

The parallel slice server application is described by the diagram of Figure 3. The program written in the CAP formalism is precompiled into C++ code which is then compiled to run on the server's PC's.

Figure 3. Diagram of parallel slice extraction

The CAP computer-aided parallelization tool greatly simplifies the creation of parallel distributed memory server applications. Application programmers create separatly the serial program parts and express the parallel behavior of the program with CAP constructs. The parallel part of the program can be easily modified by changing the sequence of operations or by building hierarchical CAP constructs.

Bibliography

From the Gigaserver project (for articles not directly related to this site, consult the complete list):

• The Visible Human Slice Sequence Animation Web Server, Jean-Christophe Bessaud, Roger D. Hersch, The Third Visible Human Project Conference Proceedings, Bethesda, Maryland, 5&6 October 2000
  
• The Visible Human Slice Web Server: A first Assessment, R.D. Hersch, B. Gennart, O. Figueiredo, M. Mazzariol, J. Tarraga, S. Vetsch, V. Messerli, R. Welz, L. Bidaut, Proceedings IS&T/SPIE Conference on Internet Imaging, San Jose, Ca, Jan. 2000, SPIE Vol. 3964, 253-258
  Abstract Full paper: PDF 1.4 MB
• Computer-Aided Synthesis of Parallel Image Processing Applications, Benoit Gennart, Roger D. Hersch, Proceedings Conf. Parallel and Distributed Methods for Image Processing III, SPIE International Symposium on Optical Science, Engineering and Instrumentation, Denver, Colorado, July 1999, SPIE Vol-3817, 48-61
  Abstract Full paper: PDF 874 KB
• Parallelizing I/O intensive Image Access and Processing Applications, V. Messerli, O. Figueiredo, B. Gennart, R.D. Hersch IEEE Concurrency, Vol. 7, No. 2, April-June 1999, 28-37
  Abstract Full paper: PDF 1.7 MB
• Synthesizing Parallel Imaging Applications using the CAP Computer-Aided Parallelization tool, B. A. Gennart, M. Mazzariol, V. Messerli, R.D. Hersch, IS&T/SPIE's 10th Annual Symposium, Electronic Imaging'98,Storage & Retrieval for Image and Video Databases VI. San Jose, California, Paper 3312-28, pp. 446-458
  Abstract Full paper: PDF 387 KB
• Performances of the PS² Parallel Storage and Processing System for Tomographic Image Visualization, V. Messerli, B. Gennart, R.D. Hersch, Proc. 1997 Int'l Conference on Parallel and Distributed Systems, Seoul, Korea, December 1997, pp 514-522
  Abstract Full paper: Postscript 230 KB (gzipped)

From the Discrete Geometry project (for articles not directly related to this site, the complete list is here):

• Advances in Discrete Geometry Applied to the Extraction of Planes and Surfaces from 3D Volumes, Oscar Figueiredo, Ecole Polytechnique Fédérale de Lausanne, Ph.D. Thesis No 1944, 1999
  Abstract Full paper: Acrobat PDF 5MB
• Digitization of Bezier Curves and Patches using Discrete Geometry, Oscar Figueiredo, Jean-Pierre Reveillès and Roger D. Hersch, Discrete Geometry for Computer Imagery, Proceedings of the 8th International Conference DGCI'99, Marne-la-Vallee, France, March 1999, Gilles Bertrand, Michel Couprie, Laurent Perroton, Eds., LNCS 1568, Springer Verlag, pp. 388-398
  Abstract Full paper: Acrobat PDF 100KB