Abstract: A method for use in scale-based visualization of an image dataset is provided. The method comprises identifying a first set of voxels of the image dataset, wherein the voxels of the first set of voxels comprises gray values that are statistically frequently present in the image dataset, identifying a second set of voxels, wherein the voxels of the second set of voxels comprises gray values that are not statistically frequently present in the image dataset, and calculating a scale based on the first set of voxels and the second set of voxels using a transfer function, wherein the transfer function is non-linear. The method offers high manipulation accuracy where required within a limited amount of display space, by changing the linear interaction scale into a non-linear scale. Important image/volume gray values are given a higher percentage of interaction space on the available display space than other less important gray values.