TRUE VOLUME VISUALIZATION OF MEDICAL DATA

这个应用程序是基于三维（3D）

This application is based on the hypothesis that 3 dimensional (3D) structures are easier to identify for both humans and computers when data is presented in true 3D. 3D is a term with many meanings, and in this application, the true adjective means that the viewer can sense depth using stereopsis. This application deals primarily with human perception, however, the results form computer perception must be presented to humans for validation, and the visualization tools needed will be similar to those described for raw or processed 3D data. A volumetric workstation is under developed. The true 3D display uses the liquid crystal stereo shutter technology. The workstation has a unique direct 3D pointer and voice input and output. Volumetric software consists of a volume window manager, Mayo Clinics' Analyze program converted to the stereo display, and the Digital Stereo Microscope algorithms that allow discrete 3D data to be examined in much the same way an optical stereo microscope is used to examine transparent real samples. This application proposes to complete the workstation by upgrading a donated Explorer LX workstation containing the Odyssey multiprocessor subsystem with additional memory and disk storage, and converting numerically intensive algorithms to the existing multiprocessor. The feasibility of the technology will be demonstrated by processing and/or displaying two classes of medical data: cubes of 3D sampled data (NMR, SPECT and simulated doppler echocardiograms) and data obtained form spatially oriented transducers (multielectrode epicardial electrograms, stereo electron micrographs and stereo cineangiograms). Investigators experience with the data types will evaluate the results. The perception of medical 3D data has received little attention and psychophysical studies are proposed that will examine the investigators ability to resolve detail in 3D space. Subjects will presented a variety of application oriented stimuli and asked to respond by touching perceived structures with the 3D pointer providing unusually quantifiable data for analysis. Extensive calibration studies will also be performed to determine how accurately a subject can point to objects of precisely known dimensions.