doi:10.1016/j.media.2004.11.008
Copyright © 2004 Elsevier B.V. All rights reserved.
Real-time synthesis of bleeding for virtual hysteroscopy
aComputer Vision Laboratory, Swiss Federal Institute of Technology Zurich, CH-8092 Zurich, Switzerland
bDepartment of OB/GYN, Clinic of Gynecology, University Hospital of Zurich, Switzerland
Received 2 February 2004;
revised 20 September 2004;
accepted 9 November 2004.
Available online 14 January 2005.
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Abstract
In this paper we present a method for simulating bleeding in a virtual reality hysteroscopic simulator for surgical training. The simulated bleeding is required to be visually appealing while at the same time instantaneously responsive to any feedback that the surgeon may be conducing to the virtual environment. In order to meet these real-time requirements, we have based the simulation on graphical fluid solvers. These solvers primarily work best over a 2D domain. For correct visualization in the hysteroscopic simulator, it is, however, necessary to perform the simulation fully in 3D. Therefore in this paper we also present the design modifications for 3D graphical fluid solving and show how to use parallelization to maintain real-time behavior. We also discuss how the incorporation of massless particles into the simulation can improve the visual quality of the results by limiting numerical dissipation effects.
Keywords: Surgical simulation; Virtual reality; Flow simulation; Hysteroscopy; Bleeding
Fig. 1. Flowchart of the solution. Top line contains the evolution of the velocity field, Eqs. Figs. (1) and (6). Bottom line contains the evolution of the substance field, Eq. (7).
Fig. 2. Spatially discretized 2D staggered grid configuration.
Fig. 3. Top row: original images recorded during hysteroscopic surgery in the case of bleeding when fluid inflow and outflow were stopped. Bottom row: frames from the real-time 2D synthesis. See Movies 1 and 2 in Appendix A.
Fig. 4. Spatially discretized 3D staggered grid cell.
Fig. 5. 3D simulation of bleeding with pure semi-Lagrangian method. Notice the change of the camera position during the simulation. See Movie 3 in Appendix A.
Fig. 6. The setup of parallelization.
Fig. 7. Simulated bleeding for virtual hysteroscopy in 3D with the marker particle approach. See Movie 4 in Appendix A.
Fig. 8. Interacting with the environment. Inflow and outflow were switched on in Frame 2. See Movie 4 in Appendix A.
Fig. 9. Average calculation times in ms.
Table 1.
Average calculation time in ms (τn), speedup (S) and efficiency (ε) in percentage for the 3D numerical solver

Table 2.
Average calculation time in ms (τn), speedup (S) and efficiency (ε) in percentage for the marker update
