We first used the facet model surface detection program to process the CT images of the fresh female cadaver. The window size is 5x5x5. Fig.4(a), Fig.4(b), Fig.4(c), Fig.4(d), and Fig.4(e) are CT images and Fig.4(f) shows the detected surface points from these CT images.  Fig.4(g) shows the detected surface points of the female head including skin and skull. About 210 CT images were processed and it took about 40 hours to finish the surface detection process. The extracted surfaces can then be seperated using finite element model [4].

    Then we used the active contour model contour extraction and registration program to extract the external surfaces of the fresh male and female cadaver. The surfaces of the arms, legs and the body are extracted seperately using different initial contours. Then these surfaces are put together to construct the external surface of the cadaver. Fig.5(a) shows the extracted surface points of the male cadaver and Fig.5(b) shows its head section.  Fig.6(a), Fig.6(b), Fig.6(c), and Fig.6(d) show the generated surface viewing from different angles. For most part of the body, contours can be extracted sequentially without re-initialization. For the total of 521 CT images, 73 initial contours were generated. (some initial contours were generated to extract seperate anatomical entities) Heavy human intervention occurred at the section of neck and hand because of irregular shapes and rapid shape change shown in these sections. (For example, see Fig.7(a) Fig.7(b) Fig.7(c) Fig.7(d) which show a sequence of CT images at the neck section). This problem could be solved by decreasing the thickness of the CT images in these sections.  Fig.8(a) shows a partially extracted surface points of the female cadaver and Fig.8(b) shows it head section.  Fig.9(a), Fig.9(b), Fig.9(c), Fig.9(d) show the generated surface viewing from different angles. For a total of 240 CT images, 18 initial contours were generated.

    We also used this program to extract the skull surface points of the male cadaver. Fig.10 shows the external surface points of the skull. The red part represents the skull. The yellow part is the mandible. The cyan part is the tooth.  Fig.11(a) and Fig.11(b) show the generated surface. Fig.12 shows the internal surface points of the skull. (Only two major parts are extracted) Fig.13(a) and Fig.13(b) show the generated surface. Fig.14 shows an overlapped image of both external and internal surface points of the skull. Because of the complex structure of the skull, more initial contours were generated. For a total of 167 CT images, 67 initial contours were generated. (some are generated to extract seperate part of the skull).