Fluorescence microscopy is a popular technique for multidimensional analysis of biological specimens. Within this framework, hyperspectral imaging allows to provide additional information on the sample of interest. However, the acquisition process induces various degradations on the image which can prevent quantitative post-processing treatment algorithms from producing accurate results. Fluorescence imaging thus benefits from restoration techniques such as deconvolution. To the best of our knowledge, only a few works specifically focus on the problem of hyperspectral image restoration. We present here a restoration method that is specifically designed to process large multidimensional hyperspectral images.