We are developing three products based on the Visible Human Project. The goal of these products is to facilitate the understanding of human anatomy for both students and clinicians.

Complete Visible Human Male, Version 2.0 and Female, Version 1.0

Laserdisc technology has significantly enhanced the teaching of anatomy in medical schools around the US. Multimedia packages taking advantage of the laserdisc's storage capacity, quick access times, and ability to deliver large, high-quality images have achieved significant penetration and dedicated usage. However, in order to be fully effective, the laserdisc must be controlled by an intuitive, functional software program delivering commands via serial port. Such a controller was developed for the Visible Male laserdisc at the Center For Human Simulation (CHS) at the University of Colorado, and reengineered for cross-platform delivery at Gold Standard Multimedia Inc. (GSM).

The controller brings two particularly beneficial functions to the user. First, the user can "play through" any identifiable anatomical structure axially, coronally, and sagittally, using information about the boundaries of each structure in all three planes. Second, the controller accesses the corresponding image from a CD-ROM collection whenever a laserdisc image has been settled upon, allowing the controller to act as an ultra-functional "index" to the digital images, which can of course be easily manipulated by common graphics applications.

The Complete Male also features a new laserdisc (Version 2.0) premastered at CHS. The laserdisc contains an award-winning six-minute video produced at NCAR, several animated renderings, and a preview of the Female images.

The Complete Female laserdisc has been premastered for the first time at CHS, and features three times as many axial cross sections as does the Male. A controller for this laserdisc is also available.

Segmented and Classified Visible Human Male Voxels and 3D Models

The primary difficulty in utilizing the original Visible Male data in educational programs has been the lack of "labeling" in the data. Specifically, we know, for any point in the dataset, what color (RGB) is present, but we do not know what anatomical entity is present. GSM has funded this project, referred to as "segmentation and classification," at CHS. Under the direction of Drs. Spitzer and Whitlock, the principal investigators for the Visible Human Project, each voxel in the Visible Male dataset is being classified as to its anatomical structure. With the additional data, several new educational possibilities present themselves. First, the user can query any point of a 2D image or 3D rendering to receive feedback on the structure of interest. Second, structures can be isolated from their surroundings individually, to highlight their shapes, or in selected groups, to emphasize anatomical relationships; this can also be done in two or three-dimensional presentations. Third, polygonal 3D models can be (non-trivially) derived from the segmented voxels in order to enable more real-time interaction with specified organs and systems. Photorealistic texture maps can also be derived for additional realism and impact. We plan to utilize these models and textures in future versions of the Complete Visible Human products described above.

The scheduled completion date for the voxel segmentation and classification is October 31, 1996. We will make the voxel data, and also the 3D models and textures, available for third parties to utilize in projects relating to virtual reality simulation and education.

Volume-Rendering Anatomy Atlas

Perhaps the most useful presentation of the Visible Human datasets will be through volume, or voxel-based, rendering. This process, which bypasses polygonal intermediates, provides the most detailed look at the data, specifically when oblique cutting planes are applied, or the study of internal structure and appearance of anatomical entities is necessary. Traditionally, volume rendering systems have been too computationally intensive for personal computers, and have been limited to expensive graphics workstations. However, with the acceleration of Pentium and PowerPC processor speeds, we feel that a volume-rendering product is achievable now. Development of this product is underway. We feel this product will constitute both a valuable educational tool and resource of significant importance to physicians and other clinicians.