Volume visualization based on the intensity and SUSAN transfer function spaces

Designing transfer functions is a challenging task for medical volume data visualization, especially when an arch of the same boundary disperses seriously and adjacent arches are intersected in the intensity and gradient magnitude (IGM) transfer function space. In this paper, a novel transfer function space is proposed to better highlight and differentiate different materials in realistic volume datasets. The proposed method combines the intensity values and three-dimensional (3D) SUSAN (Smallest Univalue Segment Assimilating Nucleus) edge responses of the original data to define the intensity and SUSAN (IS) transfer function space. The results of various datasets in volume rendering show that boundary of different materials exhibits a trapezoidal shape in the proposed IS space, and boundary information is much better brought out in comparison to the IGM space. Thus the IS space provides much more intuitive clues than the IGM space in order that transfer functions can be more easily designed. Meanwhile, more details of materials of interest are visible in the rendering images.