Methods and Techniques Discussion => Image Analysis => Topic started by: Neurochica on May 12, 2010, 01:29:34 PM

Title: Sampling and Quantifying Puncta
Post by: Neurochica on May 12, 2010, 01:29:34 PM
Hi Everyone :)

I have been doing some immunohistochem on glutamate receptors within the vestibular system. I am using an antibody to the glutamate receptor subunit 4, along with carboxy terminal binding protein 2 (for ribbon synapse dense bodies) and phalloidin as a marker for different pre-defined zone borders within my studied organ. I take confocal images of the tissue that are multiple optical sections deep and then montage between 9-12 of these image stacks together to get a complete cross section of my organ. The problem I am running in to is exactly how to quantify my data. I am basically looking for regional differences in these pre-defined zone borders in intensity and number of glutamate receptor puncta. I have looked in to a few different methods that use thresholding (ImageJ), but I have found none that incorporate a Z-dimension in their analysis. Does anyone happen to know of some research that uses methodology that may solve or come close to solving my quantification problem?

I am also hoping to look at "colocalizations" between the carboxy terminal binding protein 2 and the glutamate receptor subunit 4, but this is proving to be a little challenging. The carboxy terminal binding protein 2 and glutamate receptor subunit 4 are not actually colocalized but are adjacent to each other (one is pre-synaptic and one is post-synaptic) Does anyone know of a way to count these "colocalizations" while incorporating a z-dimension in to the analysis?

Is it possible that I may just have to analyze one dimension only? Could I possibly use a maximum intensity projection of my image stacks?

Any help would be greatly, greatly appreciated!! Thanks so much!!