Next: Full Text Index Contents: Conference
Page
VHIF: A Prototype File Format for Anatomical Images
Abstract
Availability of the Visible Human® dataset
provides the medical imaging community a rich collection of full body anatomical
images. However, its delivery highlights a number of issues related to
effective methods for sharing large anatomical image files. These issues
relate to the format in which the images are stored and accessed. Presently
Visible Human® images are stored as body cross section slices
in raw files , and delivered via ftp in a non anatomical format. This requires
users to have prior knowledge of the sliced ordered sequence if they wish
to retrieve a specific anatomical structure. To address this issue the
National Library of Medicine approved a second phase of the Visible Human®
program to address the anatomical labeling, store and retrieval of the
dataset. This extended phase included the 3DSystem project[1], which was
tasked to design and developed a prototype anatomical image database.
A critical part of the database design was the development of data
structures for importing, storing, and delivering useful anatomical image
data to a diverse population of users. Based on a review of existing users
of the Visible Human® dataset ,we proposed multiple image types
to support the imaging needs a range of potential users. For example, educational
users may expect rendered 3Dimages depicting selected anatomical structures,
to include in curricula development. Product developers may require access
to the segmented masks to render additional anatomical views not included
in the database. Whereas, researchers may need to access the original image
data. However, to be useful, each of the proposed image types must be network
accessible, and delivered in a functional electronic file format.
Keywords: 3D anatomical models, file format, anatomical image
database.
Next: Full Text Index Contents: Conference
Page
