Volumetric Image Processing for Three-Dimensional Display of the Skeletal Anatomy of the Sea Otter (Enhydra lutris)
IAAAM 1987
Michael K.Stoakopf1,2; Elliot K. Fishman3; Thomas D. Williams4
1National Aquarium in Baltimore; 2Div. of Comp. Med., Johns Hopkins University; 3Dept. Radiology, Johns Hopkins Med. Inst.; 4Monterey Bay Aquarium

As part of an anatomical study of the sea otter, Enhydra lutris, volumetric images were generated from whole body computerized tomography (CT) scans of otters found dead on northern California beaches. These three dimensional color images preserved all original CT data, not just surface boundries, and could be displayed as static images in either a 1024x768 or a 640x488 NISC format as static images. Alternatively they could be rotated about any axis in "real-time" by displaying sequential rotation sequences generated by the Pixar imaging computer. Preprocessing control of the CT data also allowed the removal of rectangular areas of anatomy to provide unimpeded view of underlying detail in the rotating images. CT scans consisted of overlapping transaxial scans taken at 3mm intervals scanning for 3 seconds at 230 mAs, 125 kVp with 4 mm collimation. Data from the CT scans were recorded on 1/2 inch tape and transferred to an imaging system consisting of a Sun 3/160 computer workstation (Motorola 68020 microprocessor; 4 Mbyte RAM) connected through a Unix operating system to a Pixar Imaging Computer with four concurrent channel processors able to execute ten million instructions per second (24 Mbyte fram buffer RAM; 2048x48 bit picture elements/side). Program initiation and interaction with the PIC utilized the "DOCTOR" program, a menu based interface for computer-naive operators, written at Johns Hopkins Hospital by Derek Ney. Three dimensional images were rendered by processing stacks of sequential CT images as volumes, while replacing the grey scale intensity information of each pixel with gels of varying color and transparency. To create realistic depth of field through overshadowing and differential opacification, an opacification algorithm traced rays from a selected viewing plane through the colored gel volume, calculating absorbance percentages as a function of the transparency of the gels it penetrated.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Michael K. Stoskopf, DVM, PhD
Baltimore, MD


MAIN : All : Volumetric Image Processing
Powered By VIN
SAID=27