Surface modeling of craniofacial form in human embryos with a limited graphics terminal

Three-dimensional morphology of the human embryo typically is visualized through computerized modeling techniques utilizing planar contours as the data base. Through this approach, tissue outlines are digitized, and contour lines are superimposed, providing a depth perspective. However, these techniques represent embryonic tissues as discontinuous surfaces and therefore ignore morphological information between sections. The purpose of this study was to develop a computerized routine for the three-dimensional surface modeling of craniofacial morphology in human embryos. Tissue outlines are digitized, thus converting contour information into x,y,z coordinate data. The three-dimensional reconstruction program BCSURF opens the data file and plots each tissue polygon. A center is determined for each contour, and this value is used to divide each polygon into four segments. Surface patches are generated by mapping each segment onto the corresponding segment of subsequent sections. A face table is constructed representing the surface patches and plane normals are generated for each patch. The normal and depth values are appended to the face table, and these measures determine the color intensity for each patch. Finally, patches are plotted providing a polygon mesh model, and each patch is filled with a dither pattern according to shading values. Three-dimensional reconstructions of the craniofacial region in Carnegie embryos (stages 15-17) are generated, and major morphological features are observed. Although bilevel shading capabilities cause discontinuous shading textures, this simple and inexpensive system can be easily upgraded for high-resolution graphics.