Improving Separability of Structures with Similar Attributes in 2D Transfer Function Design

Abstract

The 2D transfer function based on scalar value and gradient magnitude (SG-TF) is popularly used in volume rendering. However, it is plagued by the boundary-overlapping problem: different structures with similar attributes have the same region in SG-TF space, and their boundaries are usually connected. The SG-TF thus often fails in separating these structures (or their boundaries) and has limited ability to classify different objects in real-world 3D images. To overcome such a difficulty, we propose a novel method for boundary separation by integrating spatial connectivity computation of the boundaries and set operations on boundary voxels into the SG-TF. Specifically, spatial positions of boundaries and their regions in the SG-TF space are computed, from which boundaries can be well separated and volume rendered in different colors. In the method, the boundaries are divided into three classes and different boundary-separation techniques are applied to them, respectively. The complex task of separating various boundaries in 3D images is then simplified by breaking it into several small separation problems. The method shows good object classification ability in real-world 3D images while avoiding the complexity of high-dimensional transfer functions. Its effectiveness and validation is demonstrated by many experimental results to visualize boundaries of different structures in complex real-world 3D images.

Separation and visualization of arteries and heart in 3D computed tomography angiography images

“…In this paper, we provide a priori knowledge based method to roughly determine the region, and then we slightly adjust and modify the region in the SG-TF space according to the visual feedback results. This method is very friendly for region selection in SG-TF space when compared with the previous method [25] . Now, the region selection does not need to be done manually by the user through trial-and-error.…”

“…In cardiac system visualization, Zhang et al [22] proposed real-time visualization of 4D cardiac MR images, Sera et al [23] also provided a cardiovascular visualization of high-resolution 4D images, but these two methods cannot remove the aorta and spine which are connected to the heart. Although Lan's method [25] can separate and visualize the arteries and heart, the region selection of SG-TF space is very difficult and tedious, and it is time-consuming and tedious to extract the structures of interest from the determined voxels. These shortcomings can bring great inconvenience in clinical visualization.…”

“…We can easily extract all connected sets from Φ by spatial connectivity [25] , but it is consuming-time and tedious work in volume rendering. So we modify the spatial connectivity algorithm to directly extract the arteries or heart instead of extracting all connected sets by specifying the seed point on the aorta or heart.…”

“…These structures cannot be easily separated due to the partial volume effects (PVEs) [24] ; therefore, the previous transfer functions usually fail in visualizing the arteries and heart separately. Although Lan's method [25] can separate and visualize the arteries and heart in volume rendering, it faces the region selection problem in (scalar-gradient transfer function) SG-TF space [8] . It is usually done manually by the user through trial-and-error and is a tedious work.…”

“…The second step is to directly extract the arteries or heart from the determined voxels by specifying the seed point in volume rendering. Compared with Lan's method [25] , these two steps can greatly improve the region selection method and quickly extract the structures of interest. Finally, the set operations of boundary voxels are still applied to separate the arteries and heart and remove the spine.…”