ANATLINE GENERAL INFORMATION

Introduction

AnatLine is a prototype system consisting of an anatomical image database and an online browser. This prototype was developed by the Lister Hill National Center for Biomedical Communications at the National Library of Medicine to provide a client/server Internet interface to access gross anatomy images of the human body. During its beta phase, AnatLine stores a subset of the Visible Human Data Set. This version of AnatLine consists of image records of thoracic cross sectional planes, thoracic voxel structures, and rendered bitmap images. Access to the Anatomical Database during its beta test phase requires registration with the Lister Hill National Center for Biomedical Communications. AnatLines Anatomical Browser provides online viewing of low resolution surface rendered images of the human thoracic region.

Anatomical Database

AnatLine's database is a Java based object-oriented database which uses an ObjectStore persistent engine. AnatLine data structure is based on anatomical and spatial relationships of gross anatomy components and physiological systems. The underlying structure for these relationships is the Database UMLS Terms Authority.

Users query and retrieve image records via the Anatomical Database. Available image records consist of anatomical data and gross anatomy images. Image records are stored in NLM's prototype .vhi format and its internal data and bitmaps may be extracted or displayed using AnatLine support software.

Images

AnatLine, in its prototype phase, stores images processed from anatomical structures of the Visible Human male thorax. These images were scanned from the 70mm film cross section data set and digitized in a 4096x2700 pixel frame size. This digitized collection of cross section images was aligned and segmented into 434 thoracic structures by Engineering Animation Inc.

Image data retrieved from the database are returned as image records. Each image record contains bitmaps, image data, and anatomical data. Image records are enclosed within a prototype .vhi image format. This format contains a file specifier and internal data encoded to be machine readable by the user's client system. Presently four image file format types available; Cross Section files, Volume of Interest (VOI), Segmented Byte Masks, and Rendered Images. Image record data may be displayed using NLM's VHDisplay software, or image data may be extracted using NLM's VHParser.

AnatLine queries of stored anatomical structures return data which specify the image type and file size for available image records. Cross section records are stored with color bitmaps, and graphical overlay labels. Two types of voxel volumes are stored. First are Volume of Interest (VOI) records, which are full color z-plane stacks of the structures region of interest (ROI), and z-plane stacks of segmented masks. VOI files are limited in size to 1GB. Second are Segmented Byte masks, which contain z-plane stacks of the segmented masks without the color ROI. Byte mask files contain compressed images, therefore the file size increases when uncompressed. Rendered Images are 2D representations of anatomical 3D structures.

Query Functions

The Web interface provides a query engine with retrieval access to high-resolution anatomical images that range in size from 100KB for browser viewable rended images, to 1GB for anatomical structures in voxel file format. The query and retrieval client/server system is composed of applet GUIs, servlets, and RMI application modules which communicate with each other to allow users to query for specific anatomical structures, and retrieve image data as well as associated anatomical images.

Pull-down menus allow the user to select the gender and media type of the image. For the purposes of this prototype, only digital color images relating to the male specimen thorax are available.

A spell-assist GUI inteface helps users to formulate database queries. The spell asssist window contains a scrollable lexicon of unique human anatomical keywords. Users may enter the anatomical name of the query into the text field above the list, or may scroll the spell assist window and select from among the listed entries. As the user types each query character, the spell assist window scrolls to a recognized anatomical term, thus permitting the user to select the highlighted term. If searching for a term that contains two or more words, the spell assist window will repeat the scroll and selection process for each entry. Use of the spell assist word selection is optional, its purpose is to assist with terminology spelling. The spell assist will not restrict the user's entry, nor does it display a table of known images within the database.

Upon typing or selecting the desired query term (pressing the RETURN-key or selecting "Search") the system provides a listing of all UMLS anatomical terms which contain the entry. The user may select a more accurate term which represents the desired query and query the database.

Queries may be unreconized by AnatLine for several reasons: the entry may be mispelled or the entry may not exist as a preferred UMLS term. If an entire term is unrecognized by AnatLine, an error message box displays instructing the user to retype the query or select from the recognized terms from the spell assist.

If there is a partial match of the queried term, an Extended Search message box will display prompting the user to query on a partial match or cancel the query. Extended Searches map several terminology databases to AnatLine recognized terms. The user may then opt to query for the returned terms.

Query Results

Successful queries return results containing embedded hyperlinks to retrievable image types, and links related to anatomical objects for the queried structure (i.e. anatomical parent, substructures, physiological system and body region).

DIGITAL COLOR Images for the left lung :

Related Anatomical Data for the left lung :

Body Region

• thorax
• thoracic cavity

Substructures

• left upper lobe of lung
• left lower lobe of lung

To retrieve images, the user must select the Image Type within the Query Results table. The Image Retrieval Results table lists available image files for the structure and image type.

Download Procedures

Selected images may be downloaded individually as single files via HTTP or downloaded in batch-mode. Please note: To download .vhi image files, the user's browser must be configured to accept the .vhi file type. Browser configurations are provided on the FAQ page.

Single File Download

To download a single file, the user selects a file via the hyperlink within the File Name column of the Image Retrieval Results table. File download process begins, and the user must define the download path for local storage.

Batch-mode Download (Netscape Object Signing)

To download multiple files, the user may select the "download Multiple Files" option from the Image Retrieval Results table. This option writes files in batch-mode to the user's hard disk. The user has the option to select/deselect files for download. If no files are individually selected, all files will download.

Downloading multiple files uses Netscapes Object Signing technology developed by Netscape and VeriSign, Inc. Users who choose not to accept the Netscape VeriSign certificate, or who are not using Netscape must download files using the Single File Download option. Use of Object Signing technology displays a security pop-up window requesting the user to grant or deny the program's request to read or modify the user's hard disk. Clicking the "Grant" button displays, in the browser, an applet that consists of a text entry field, a table and buttons. The text field contains the path to the user's home directory. The user can manually change the default path.