Thread Subject: Gaussian surface fit for image intensity peak finding

First my thank for anyone who is reading this.

Basically I have an image of particles shined with laser. So the intensity of each particle is close to a Gaussian Surface.

I have already define and isolated each particle into a small intensity matrix, and pixels that are not likely to be part of the Gaussian are already be set as zero, what I want to do is use Gaussian Surface fit to find the peak location of the particle intensity.

I have searched many places and I think I have end up with the lsqcurvefit function, but I do not quite understand how this function should be used. Any suggestion and help would be really appreciated.