Application of deconvolution algorithms to astronomical images is often limited by variations in PSF structure over the domain of the images. One major difficulty is that Fourier methods can no longer be used for fast convolutions over the enitre images. However, if the PSF is modeled as a sum of orthogonal functions that are individually constant in form over the images, but whose relative amplitudes encode the PSF spatial variability, then separation of variables again allows global image operations to be used. This approach is readily adapted to the Lucy-Richardson deconvolution algorithm. Use of the Karhunen-Loeve transform allows for a particularly compact orthogonal expansion of the PSF. These techniques are demonstrated on the deconvolution of Gemini/Hokupa'a adaptive optics images of the galactic center.