Interactive Volume Modeling And Illustration Of Muscle

Medical illustration has been widely recognized as an effective way to depict salient anatomical features while hiding irrelevant details. This paper presents two solutions to the challenging problem of modeling and illustrating muscle volume by taking advantage of existing medical imaging techniques. First, an interactive volume modeling method is proposed to build skeletal muscles within scalar volume data from computer tomography or magnetic resonance imaging. We present an intuitive anatomy classification approach to accommodate spatial constraints on the muscle volume. Provided with an easy-to-use mod-eling interface, users are able to produce compelling results that conform to the character-istic anatomy in the input volume. We experiment our approach with four medical datasets and obtain results that are comparable to hand-drawn muscle illustration. This work is useful in illustrative visualization that involves muscle modeling, where the spatial con-text should be faithfully preserved. Second, this paper presents a new muscle illustration approach to effectively visualize fibrous structure by leveraging diffusion tensor imaging (DTI) data and example-based solid texture synthesis techniques. To represent the structure and orientation of muscle bundles, we reformulate the volumetric DTI data into a scalar field and an auxiliary guidance vector field. A muscle mask derived from DTI data is used to identify the muscle region. To simulate the internal appearance of muscle, we propose a new 2D example-based solid texture synthesis algorithm that generates a solid texture constrained by a guidance vector field. Rendering the constructed scalar field along with solid texture highlights the global appearance of the muscle as well as the local shape and structure of muscle fibers in an illustrative fashion. We demonstrate plausible illustration and expressiveness by applying the proposed approach to five example DTI datasets.