Large Medical Data Manipulation for Bone Surgery Simulation


Medical image data obtained from Computed Tomography (CT) are used as input to reconstruct and visualize 3-D structures of human bones for the purpose of developing a virtual reality (VR) based bone surgery system. These data are used for geometric modeling, force modeling, and model update to perform simulation of material removal with graphic and haptic rendering. One important issue in bone surgery simulation is to handle the large, complex, and often poor-quality data. Although the processing power of personal computer has increased greatly over the years, improper data handling can still cause implementation problems such as excessive memory consumption, low data processing speed, and incapability of real-time simulation. This paper presents a method for managing large CT scan data based on the consideration of implementation complexity, memory storage and computational overhead. Besides medical data acquisition and image processing, two important computer graphics concepts, i.e. bounding volume and adaptive subdivision, are applied to remove irrelevant data and to organize the rest data. Two data structures, a complex linked list and a Quadtree list, are developed to store and organize the image data. These data are processed before VR simulation so as to reduce the data update time. With the proposed method, the memory bandwidth requirement is reduced drastically and real-time simulation performance is achieved.


Mechanical and Aerospace Engineering

Keywords and Phrases

Bone Surgery System; Computed Tomography (CT); Computer Simulation; Medical Applications; Virtual Reality (VR)

Document Type

Article - Conference proceedings

Document Version


File Type





© 2005 American Society of Mechanical Engineers (ASME), All rights reserved.

Publication Date

01 Jan 2005