This project will test the hypothesis that studying anatomy from 3-dimensional (3-D) computer models enhances learning. Two-dimensional (cross sectional) photographic images of female anatomy from the National Library of Medicine's Visible Human Project will be used to construct 3-D surface contours using a computer program written by the Co-PI running on a Silicon Graphics Workstation. The resulting computer models of anatomical structures will be incorporated into computer aided instructional packages with two levels of flexibility and sophistication. Effectiveness of the different approaches will be tested by first year medical students taking the gross anatomy course who will be assigned to one of three groups using a matched randomized procedure based on initial survey of experience with WWW and Virtual Reality. The same text will be used in each learning package to present the conceptual and factual information, but different modes of visual illustration will be employed. One group will be exposed to traditional static, single-view 2-D labeled illustrations, although these will be derived from reconstructions of the Visible Human data. A control group will study the same text and images printed on paper. The third group will study text linked to 3-D models presented in Virtual Reality Markup Language (VRML) using Netscape and Cosmoplayer. This group will be able to interact with the 3-D models, manipulating their position to allow viewing from virtually any chosen perspective. Both computer packages will be implemented on standard personal computer platforms. To avoid contamination, effectiveness will be measured immediately after training, using a battery of tests including standard multiple choice questions probing understanding of concepts and spatial relationships; identification of structures in cross sectional images; and ability to match the position of a structure with its location within rectangles indicated in otherwise """"""""empty"""""""" cross-sectional diagrams. In a quasi-experimental phase, performance of each group on spatial and conceptual questions on the course examinations will be compared. Subjective responses to the experience will be solicited by questionnaire. The subject of these learning packages will be the female pelvis and perineum. These areas of anatomy are particularly difficult to demonstrate by dissection and increasingly important for nearly all physicians to master considering the movement toward generalist medicine, but they are also emotionally charged. This project will have an important impact on health care delivery by improving physicians' understanding and visualization of a difficult and important anatomical area. In addition, issues of instructional design and effectiveness are addressed, including the crucial question regarding cost- effectiveness of using 3-D surface models in education, and platform- independent delivery through the use of hypertext markup language and VRML. The impact of downloading time will be studied to evaluate use via Internet, local area networks, and use from CD-ROM.
