I am working on a research project in clinical medicine part of which
requires mapping of the points on one 2-D body image to the
corresponding points on a second differently shaped 2-D body image.

As part of our project (written in C++ and MatLab) our team has written
MatLab code that displays two TIF images of a human body on a computer
screen and allows a user to click on a point on each image in succession
to identify a common anatomic landmark in each image. We will thus
collect data on, say, 30 landmark pairs, some along the outline of the
body and some inside the outline.

The next steps will be to:

1. Map the pixels along the outline (i.e., map each pixel on the outline
of one body image to a corresponding pixel on the outline of the other
body image).
2. Map the interior pixels: (i.e., map each pixel inside one body image
to a corresponding pixel inside the other body image).

Our current thinking is to use geometric morphometric approaches to map
both the outline and the interior of the images.

For mapping of the outline, we had originally thought we would just map
the outline pixels according to the proportional distance between two
common landmarks on the outline (e.g., map the pixel 10% of the way
along the outline from Landmark A to Landmark B in image 1 to the pixel
10% of the way along the outline from Landmark A to Landmark B in image
2).  However, it seems that Booksteinss sliding semilandmark method
might be a superior approach.

We had then thought to use the original user-specified common landmarks
together with the calculated common pixels along the outlines as the
basis for a thin-plate spline interpolation (or similar interpolation
technique) to map pixels on the interiors of the body images.

I would welcome any recommendations as to the best way to do the
mapping. In addition, if anyone with mapping expertise in a MatLab
setting would like to join our team, we would be most interested  some
funding support can be provided.

Colin R. Taylor, M.D.
[EMAIL PROTECTED]

_________________________________________________________________
Fast. Reliable. Get MSN 9 Dial-up - 3 months for the price of 1! 
(Limited-time Offer)
http://click.atdmt.com/AVE/go/onm00200361ave/direct/01/
==
Replies will be sent to list.
For more information see http://life.bio.sunysb.edu/morph/morphmet.html.

Reply via email to