Extensive use of simulations with a few experiments thrown in for good measure. Upshot of this article is that the standard 15-20% energy windows most people use aren't optimal for Ga-67 imaging. What I also found interesting was that the optimum windows changed depending on the task being performed (detection vs estimation). Results were based on SNR calculations, but did not look at any changes in image quality. Only planar images were examined in this paper. It would be interesting to see how the modified windows affect Ga-67 SPECT imaging. With 5/8" crystals coming into vogue now, it would also be interesting to see if the optimum windows change with a thicker crystal. I'll have to file that away on my list of projects to get to. That might make an interesting subject for a summer student to tackle.