Hi, As a few people have asked for more information, I am posting some details about using confocal microscopy for morphometrics. The first challenge is to get the 3D data out of the confocal scan. There are two ways to do this, and which one you choose depends on the type of surface you are interested in. If you have a confocal scan of a fully 3D surface (something like a whole skull), then it is best to analyse the surface using something like CT scanning software to capture landmark points by importing the image stack from the confocal into it. Chris, you mentioned that you get an image stack from the confocal. This stack can be imported into most CT software such as MIPAV (http://mipav.cit.nih.gov/). It easily imports Leica stacks but I haven't tried it with a Zeiss confocal stack before. I usually don't use CT software to gather landmarks, however. You can use the second option if your surface can be represented from a single direction, what is called 2.5D, so you can use geographic information systems (GIS) software. To do this, I convert the image stack information into 3D data points that represent the surface and then analyse the surface using GIS. Most confocal software will allow you to export a depth map or digital elevation model that represents the surface of interest. This represents the height of the surface as a grey level in a single image. Unfortunately, this then loses a lot of the 3D information but then represents the surface in a more useable form. If you then output the pixel values for the image using NIH image (Mac: http://rsb.info.nih.gov/nih-image/Default.html), Scion Image (PC: http://www.scioncorp.com/frames/fr_download_now.htm) or ImageJ (platform-independent Java: http://rsb.info.nih.gov/ij/) the depth map can be converted to x,y,z points that represent the surface by using the information on the voxel sizes or fields of view in the x, y and z directions. The x,y,z points can then be imported into GIS software such as Surfer (http://www.goldensoftware.com/products/surfer/surfer.shtml), ArcView (http://www.esri.com/software/arcview/) or MFWorks (http://www.thinkspace.com/software/mfworks/index.html). These will then give a representation of the surface that you can capture landmarks from. It sounds like a long and involved process, but all steps can be automated to produce a useable 3D surface in a minute or two using a simple VisualBasic program I have written. I have a few papers where I have used the results of the process in getting landmarks and measurements from the confocal scans which I can send to anyone interested. These do not use geometric morphometrics but measure other characteristics of the surface. If you have any questions about this I'd be happy to answer them as best I can.
Best wishes, Alistair -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
