A method of fusing two volume representations wherein the fused information is created by blending the information of datasets corresponding with the volume representations by means of a blending function with a blending weight that is adjusted locally and/or dynamically on the basis of the information of either of the datasets.