v = fminsearch(@(v) sum(abs(Data - YourGammaConvolutionFunction(v(1), v(2), v(3))).^2), [v01 v02 v03]);
This often works quite nicely and is quick and easy. It really helps if you get a good estimate of the initial parameters - for example find the peak, RMS width, background etc. first).
Note that your gamma convolution may only require three parameters, but for it to fit your data, you may need to rescale and shift it (two additional search parameters), unless these are two of your three parameters your want to estimate.