Anatomical Models Virtual Anatomy from the Visible Man:
Creating Tools for Medical Education
Helene Hoffman PhD, Ann Irwin MD, Ram Prayaga MS,
Mark Danks MA, & Margaret Murray MA

University of California, San Diego, School of Medicine
Learning Resources Center
La Jolla, California, 92093-0661, USA

Presented at the First Visible Human Project Conference Bethesda, MD October 7 & 8, 1996

Abstract: The Learning Resources Center of the UCSD School of Medicine is actively exploring the educational applications of VR. Our designs use VR as a vehicle for building powerful and compelling simulated environments, where students are free to interact without the usual constraints of the physical world. VR is also being used as an intuitive interface for accessing information, a framework onto which students can structure and organize the seemingly endless array of disparate facts. UCSD's project capitalizes on these unique aspects of VR by creating a hybrid learning environment which combines VR with MM curricular resources. Developing the first anatomy lesson has afforded us the opportunity to identify challenges associated with creating effective teaching paradigms for use within virtual environments. The initial prototyping stage has included the design and evaluation of generalizable visualization aids and developer tools which will facilitate subsequent expansion of our system. We also plan to extend the scope, depth of content, and utilization of this learning system within UCSD's medical curriculum. In addition, educational outcomes analyses and evaluations of the underlying VR and interface elements will take place during the next year.

Section Figures
IntroductionFigure 1: Models of the Abdomen
Project Description Figure 2: Intuitive 3D Workspace
Future Directions Figure 3: About the I-HUD
Figure 3.1: Table of Contents
Figure 3.2: Student Lesson Manual
Figure 3.3: Anatomic Position Reference Tool (GRACE)
Figure 3.4: Navigating with List or MAP Displays
References Figure 4: Multiple Resource Blocks
Figure 5: Managing Blocks with FACET

Introduction:
The Learning Resources Center (LRC) of the University of California, San Diego (UCSD), School of Medicine is the hub of the school's technology-based curricular education, research, and development efforts. Under a recent grant from the Defense Advanced Research Projects Agency (DARPA), we have begun to actively explore the educational applications of Virtual Reality (VR) and the synergism arising from combining VR with multimedia (MM) curricular resources (UCSD's "Virtual Reality-Multimedia Synthesis"[1-3]). To realize this and future research endeavors, an Applied Technologies Laboratory (LRC/ATL) was established. It was initially equipped with Silicon Graphics® workstations (Onyx RE2(TM), Indigo2 High Impact(TM);, and an Indy XZ(TM)) as well as a variety of three dimensional (3D) input devices, such as Ascension's Flock of Birds(TM); trackers. Together with the Instructional Production Group (LRC/IPG), whose expertise is in MM design and curricular development, the programmers and engineers of the LRC/ATL have formed a small but enthusiastic multidisciplinary development team. UCSD clinicians and researchers, representing expertise in anatomy, histology, pathology, radiology, medicine, surgery, etc., comprise the project's Faculty Advisory Board (FAB).

This team is currently creating and evaluating a multi-modal learning environment, where VR serves both as the lesson core and as the interface to diverse instructional and reference elements. Applications designed for a variety of target user groups and subject areas will be developed over time within this environmental framework. The specific topics will be selected from among those requiring mastery of elaborate 3D mental models and intellectual integration of spatial knowledge with large quantities of diverse pedagogic materials.


 Project Description:
Our prototype curricular application is Human Anatomy for preclinical medical students. The first lessons, focusing on the abdominal viscera, are now being implemented and are undergoing formative evaluations by members of the development team, the FAB, and selected students. The 3D polygonal models which constitute the nucleus of this application have been produced and supplied by Visible Productions, of Fort Collins, Colorado [4] [Figure 1].
Anatomical Models
Figure 1: Abdominal Models from VP
 These high-quality objects were obtained using a proprietary triangulation algorithm to connect contours traced from successive slices of the National Library of Medicine's Visible Human Project dataset[5].

Developmental efforts began about 18 months ago with a formal assessment of faculty and student needs for anatomy training at our school, and an analysis of the relevant literature[6-8]. We investigated teaching and learning issues during the preclinical years and the subsequent application of this knowledge during the clinical years. Once this analysis was completed, broad educational goals were specified, including the need to provide diverse learning experiences with exposure to anatomical variation and disease states, and to facilitate the integration of basic anatomical knowledge with clinical skills and reasoning. Target learning outcomes and challenging concept areas were also delineated, enabling the establishment of a finite set of measurable learning objectives, providing a basis for the creation of evaluation protocols, and identifying a focus for prototype lesson content.

Our VR-MM anatomy lessons are being created using a flexible and extensible object-oriented architecture developed by the LRC/ATL[9]. Over the past several months, implementation of the user interface has evolved dramatically, and many of the features specified in the initial design documents are now being realized.
An intuitive 3D workspace provides depth and location cues through the use of a tiled reference platform [Figure 2]. An interactive heads-up display (I-HUD) supplies the curricular framework, suggests actions/options, and organizes resources available to the student. Active learning within the lesson context and free exploration of all materials and models are encouraged. Users currently interact with lesson materials and navigate the 3D world using a mouse, keyboard, and hand-held motion tracker.
Anatomical Models
Figure 2: Intuitive 3D Environment
Zoom TOC Zoom Manual Zoom GRACE Zoom Loadlist/MAP

Figure 3: Interactive Heads-Up Display (I-HUD)

The I-HUD has multiple and diverse user-support functions [Figure 3]. It provides an interactive Table of Contents (TOC) for lesson access, and a procedural lesson manual with task-appropriate selectable actions. For example, suppose a student has launched the Hepatobiliary Anatomy lesson, and then chooses the "Landmarks and Relations" module from the TOC. If a specific configuration is predefined for the exercise, the system updates the displayed models accordingly. At this point, options appear in the I-HUD which let the student manipulate the transparency of the various models in order to see boundaries with respect to adjacent structures, thus facilitating the student in completing this exercise. Other task-specific options might include the presentation of prepared fly-throughs or preferred viewpoints. To help users with spatial orientation, the I-HUD includes GRACE, an intuitive, real-time indicator of anatomical position for the currently visible model(s). The I-HUD also provides two alternative methods for managing resources in the 3D workspace. The MAP display, a symbolic representation of the 3D reference platform, uses colored dots to indicate the relative locations of models and other resources from the user's viewpoint. As an alternative to the MAP, users may display a selectable list of currently active and available named resources.

Figure 4: Multiple Resource Blocks

The VR-MM lesson architecture consists of three components. Applet-driven 3D display objects, BLOCKS, are used to present models and other resources within the lesson environment. Specific BLOCKS have been developed for each type of curricular resource, including a MedPics©-like slide projector[10], a Visible Human cross-section viewer, a radiograph projector, and a Medline® database citation search window [Figure 4]. A 3D space manager, FACET, provides the environmental controls necessary to display and interact with the BLOCKS (e.g., moving, resizing, and iconicizing) [Figure 5]. To do this, an object-oriented programming paradigm is combined with an OSF/Motif-like approach to manage space and provide widgets[11]. The Lesson Environment Manager, LEM, tracks and manages all links between the instructor provided curricular framework, models, and other resources. The LEM uses record-based descriptive text files to organize the full spectrum of lesson materials and activities. These files provide TOC entries, instructions to the application on loading and placing resources, and instructions to the student including descriptions of interactive actions. Resources are described and loaded using World Wide Web Universal Resource Locator pathnames, which may be programmed absolutely or be the result of database queries to the LRC's Multimedia Catalog (MEDIACAT).

Figure 5: Managing Resource Blocks with FACET


Future Directions:
Our goals for the current virtual anatomy training environment are numerous. We intend to continue developing and refining the user interface and program architecture. We will expand the depth and breath of curriculum encompassed by these lessons, working closely with the FAB to maintain strong connections between the program and our school's curricular goals. In the next few months, selected UCSD students will begin to use our anatomy lessons as adjuncts to their lectures and labs. As the scope of the lesson content grows, the VR-based system will find a larger role in the overall educational process at UCSD. Concurrent with the expanding use of these anatomy lessons, educational outcomes studies evaluating the efficacy of our learning environment will be initiated. We are also interested in investigating the immersive and haptic requirements necessary for advancing our visualization/cognitive-based system and creating more complex surgical/procedural-based lessons.


This work was supported by a grant from DARPA (DAMD 17-94-J-4487/P5002).


References:

1. Hoffman, HM, M Murray, AE Irwin, & T McCracken: (1996), Developing a Virtual Reality Multimedia System for Anatomy Training; Health Care in the Information Age; IOS Press; pp. 204-209.

2. Hoffman, HM, AE Irwin, R Ligon, M Murray, & C Tohsaku: (1995),Virtual Reality-Multimedia Synthesis: Next Generation Learning Environments; Journal of Biocommunications; 22 (3); pp. 2-7.

3. Hoffman, HM: (1994),Virtual Reality and the Medical Curriculum: Integrating Extant and Emerging Technologies; Proceedings, Medicine Meets Virtual Reality 2: Interactive Technology and Heathcare; Morgan K (ed).

4. McCracken, T & TL Spurgeon: (1991), The Vesalius Project: interactive computers in anatomical instruction; Journal of Biocommunications; 18 (2); pp. 40-44.

5. Ackerman, MJ, VM Spitzer, AL Scherzinger, & DG Whitlock: (1995), The Visible Human data set: an image resource for anatomical visualization; Medinfo; 8 (2); pp. 1195-1198.

6. Rosse C: (1995), The potential of computerized representations of anatomy in the training of health care providers; Academic Medicine; 70 (6); pp. 499-505.

7. Schubert, R, KH Hohne, A Pommert, M Riemer, T Schiemann, & U Tiede: (1993), Spatial knowledge representation for visualization of human anatomy and function; In: Information Processing in Medical Imaging, IPMI '93 Proceedings of 13th International Conference on Information Processing in Medical Imaging; Barnett HH (ed); Springer-Verlag; pp. 168-181.

8. Erkonen, WE, MA Albanese, WL Smith, & NJ Pantazis: (1992), Effectiveness of teaching radiologic image interpretation in gross anatomy. A long-term follow-up; Invest Radiol; 27 (3); pp. 264-266.

9. Hoffman,HM, R Prayaga, M Danks, AE Irwin, & M Murray: (1997), A Flexible and Extensible Object-Oriented 3D Architecture, and its Application in the Development of Virtual Anatomy Lessons; Proceedings of Medicine Meets Virtual Reality Conference 5; Submitted.

10. Hoffman HM, AE Irwin, S Baird, CM Bloor, K Miyai, & MC Savoia: (1993), UCSD's MedPics: Implementation and Impact on the Curriculum; Proceedings of the Seventeenth Annual Symposium on Computer Applications in Medical Care; pp. 776-780.

11. Young, D: (1992), Object-Oriented Programming with C++ and OSF/Motif; Englewood Cliffs NJ, Prentice-Hall.