home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.graphics.visualization
- Path: sparky!uunet!ukma!darwin.sura.net!news.Vanderbilt.Edu!vuse.vanderbilt.edu!se.vanderbilt.edu!brandon
- From: brandon@se.vanderbilt.edu (Brandon S. Dewberry)
- Subject: Red blood cell simulation/visualization - geometry question
- Message-ID: <BxGJ2E.3AC@vuse.vanderbilt.edu>
- Sender: news@vuse.vanderbilt.edu
- Nntp-Posting-Host: taacman
- Reply-To: brandon@se.vanderbilt.edu (Brandon S. Dewberry)
- Organization: Vanderbilt University School of Engineering
- Date: Mon, 9 Nov 1992 16:27:50 GMT
- Lines: 44
-
- Hi,
- I'm beginning a project which will simulate and visualize red blood cell
- (RBC) passage in an arteriole and capillary.
-
- I want to start out as simple as possible so the model will be 2D, the
- RBC's and vessel walls will have no elastance. Plasma flow rate will
- vary with position but not with time (not pulsitile.)
-
- The vessel model is a 2D channel which gets smaller, with subsequent
- increasing plasma speed, until the channel only slightly wider than the
- RBC diameter (represented as circles in flatland.)
- This is a pretty standard basis for more complex models.
-
- RBC centers will be at points Xi, Yi. The vessel walls will be either
- b-spline curves (flexible) or functions F(x,y), whichever work the best.
- At each major time slice the simulation will determine if any of the RBCs
- have collided with each other or the vessel wall. Momentum will be
- computed based on collisions as well as the plasma velocity.
-
- Questions:
-
- Determining RBC collisions is easy, though time consuming - an n^2
- determination of distances between centers and comparison to RBC diameter.
- if( sqrt( (x1-x2)^2 + (y1-y2)^2) == rbc_diameter) then collision...
- Is there a quicker way?
-
- The more pressing question is what is the quickest way to determine collision
- between RBC's and the capillary wall?
- Finding the distance (as near as I can tell) involves iteratively searching for the
- perpendicular which includes the RBC center, then finding the distance from its
- intersection to the RBC center. For all the RBCs.
- Is there a quicker way?
- Does 'Graphics Gems' include an algorithm?
-
- A cohort suggests finding a map of all distances in the vascular space before
- beginning the simulation. This would be quickest, but limits interaction with
- the simulation.
-
- Any suggestions?
-
- Brandon
- --
- Brandon S. Dewberry Biomedical Engineering Department
- brandon@vuse.vanderbilt.edu Vanderbilt University
-