AJR 2004; 183:1517-1521
© American Roentgen Ray Society
A Web-Based ACR Index for Radiological Diagnoses
Chun-Shan Yam1,
Jonathan Kruskal,
Arkadiusz Sitek and
Michael Larson
1 All authors: Department of Radiology, Beth Israel Deaconess Medical Center and
Harvard Medical School, WCC, Rm. 306, 1 Deaconess Rd., Boston, MA 02215.
Received January 16, 2004;
accepted after revision March 26, 2004.
Address correspondence to C.-S. Yam
(csyam{at}caregroup.harvard.edu).
Abstract
OBJECTIVE. We created an online computer system that allows
radiologists to access the American College of Radiology (ACR) index for
radiologic diagnoses from any Web browser.
CONCLUSION. A total of 11,260 indexes (10 anatomic, 722 pathologic,
90 subanatomic, and 10,438 subpathologic) were integrated into this online
system. Users can navigate the entire ACR hierarchic indexing structure with
easy-to-use cascading dropdown lists. This password-protected system also
allows free-text searches, ACR index searches, and bookmark functions. We have
used this system in our departmental intranet for more than 2 years.
Introduction
The first edition of the "Index for Roentgen Diagnoses" was
published in 1955 by the American College of Radiology (ACR), the second
edition in 1961, and the third edition in 1986. This index is also known as
the "Index for Radiological Diagnoses of the American College of
Radiology." It is the standard guideline for academic radiologists in
classifying teaching cases according to the underlying anatomy and pathology
[1]. All three editions have
been printed as paperback books available exclusively through the ACR. Because
nearly 10,000 indexes are listed in each of these documents, the search
process can be time-consuming and cumbersome. In 1992, a study was conducted
at another institution to simplify paper-based searches by replacing the
hard-copy ACR book with a stand-alone computer program
[2]. In that study, each of the
52 volunteers (21 radiologists, 21 residents, and 10 medical students) was
asked to code 30 teaching cases using the book and a computer program
(HyperCard, Apple Computer). Results indicated that using the computer program
to code cases yielded higher accuracy, confidence, and timeliness. In 1999,
the ACR released a fourth edition of the index as a single-user CD-ROM version
(for PC only). Today, most radiologists still carry the ACR book, part of the
book, or copies of the book for coding teaching cases. Although the ACR coding
system has been continuously improved since 1955, this invaluable teaching
tool is still unavailable on the internet
[1].
Inspired by ACR policy to encourage radiologists and teaching faculties to
use, modify, or create new ACR indexes to suit their teaching needs
[1], we developed a Web-based
ACR index system to standardize terminology among our diverse faculty and
student bodies. This system was developed on the basis of the original
database supplied in the ACR index CD-ROM (4th edition). With this system,
users can easily navigate, search, and bookmark their commonly used
indexes.
System Design
The primary function of this system is to allow users access to the ACR
index from any Web browser. As shown in
Figure 1, this online system
uses a standard database-driven Web development approach. This system contains
three primary components: a Web server, a data source, and scripting codes.
The Web server provides Internet connectivity for client users to access Web
pages hosted at a specific Web address called a URL (uniform resource
locator). The data source is the physical information stored in a database for
Web sharing. However, in standard Web pages written in HTML format, only
static information such as text and pictures can be used. To connect a
database for Web sharing, a data gateway, such as a scripting code, is often
required. The basic functions of a scripting code can be summarized as a
four-step process: collect the user-requested information from a Web page,
pass the information to a database engine, receive the results from the
engine, and convert the results to a standard HTML Web page. In our
implementation, we used the Microsoft Internet Information System (IIS 5.0,
Microsoft Corporation) as the Web server software with a standard desktop PC
(GX 150, Dell Inc.) running Windows 2000 Professional (Microsoft Corporation).
The cost of this PC system was $815 in October 2002. We used Active Server
Page (ASP 2.0, Microsoft Corporation) for scripting code and Open Database
Connectivity (ODBC for Microsoft Access Database) for the database engine. We
chose a PC system for Web development mainly because the ACR database is
already configured in the Microsoft Access 97 format, which is easily
integrated into Web applications using ASP, ODBC, and IIS. All these software
components are already native to Windows 2000 and supported by Microsoft.
Use of ASP and the Access database for radiology teaching applications has
been reported previously [3,
4]. Writing ASP is similar to
composing an HTML document because ASP and HTML are both plain-text documents
and do not require a compiler. Although ASP is not the only programming tool
that can be used for connecting databases to Web pages, it is widely used by
Web programmers. Because of its popularity, many resources are available
online (e.g.,
www.aspfree.com,
www.aspguru.com
and
www.microsoft.com).
However, using ASP and a PC system to implement database-driven Web
applications is not the only choice. Other systems such as Mac OS, Linux and
Unix, and other software (e.g., Coldfusion
(www.Macromedia.com),
ChiliSoft
(www.sun.com),
and Linux-ASP
(www.asp-linux.com)
can also be used.
Data Source
We purchased a licensed CD-ROM version of the ACR Diagnostic Index (4th
edition) from the ACR for departmental use. The current price of this CD-ROM
software is $85 for ACR members and $160 for nonmembers. It can also be
purchased online at the official ACR web site
(www.acr.org).
The CD-ROM contains a database file (raddiag.mdb) with all the ACR indexes
encoded. This database file is stored in Microsoft Access 97 format. The
database contains four data tables: Anatomical, Sub-anatomical, Pathological,
and Sub-pathological (Fig. 2A).
The Anatomical and Sub-anatomical tables are used to store the indexes for
different anatomic regions (Fig.
2B). Ten anatomic indexes are listed in the Anatomical table and
722 in Sub-anatomical. The Pathological and Subpathological tables store the
indexes for different pathologic findings and imaging techniques
(Fig. 2C). Ninety indexes are
listed in the Pathological table and 10,438 in the Sub-pathological table.
Currently, ACR uses a 4-digit scheme for coding anatomy and a 5-digit scheme
for pathology or imaging techniques. Detailed information on the ACR Index for
Radiological Diagnoses and its latest development is available to the public
at the official ACR Web site
(www.acr.org).
View larger version (71K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. —Data tables of American College of Radiology (ACR) index.
Screen capture of Anatomical and Sub-anatomical tables in Access 97. Three
data fields used are aid, sid, and description. Field aid stands for anatomic
id and stores primary body indexes (e.g., [0] = Breast). Field sid stands for
subanatomic id and stores secondary body indexes (e.g., [1] = Upper outer
quadrant).
|
|
View larger version (69K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2C. —Data tables of American College of Radiology (ACR) index.
Screen capture of Pathological and Sub-pathological tables in Access 97. In
these two tables, new data field pid is for pathological id representing
either primary pathology findings (e.g., [2] = Inflammation) or primary
imaging techniques (e.g., [1] = Normal, Technique). Field sid stands for
subpathology id representing either subpathologic findings (e.g., [0312] =
Breast/ Neoplasm/ Benign/ Papilloma) or subsidiary imaging techniques (e.g.,
[14] = Spot compression).
|
|
Web Design
As previously stated, 11,260 indexes are encoded in the ACR database. To
display all these indexes in static Web pages or as a single PDF file would
require hundreds of pages that the user must search through. Thus, when we
designed the Web user interface for this system, easy navigation of the entire
hierarchic ACR index was a priority. For this critical requirement, we used
simple cascading dropdown lists to display the ACR indexes and simple
mouse-click navigation to each dynamically linked index page. Other functions
such as free-text search, index search, and index bookmarking were also built
into the system to enhance the user interface. Each of these functions is
detailed in the following text.
Index Navigation
Figure 3A is a screen
capture of the ACR index system home page. This page has a simple dropdown
list showing the 10 primary anatomic regions. The number before each entry is
the first-digit anatomic index (e.g., [0] = Breast). A mouse click on any of
these 10 entries will take the user to the next anatomic subdigit index page.
For example, if the user clicks on the sixth entry ([5] Heart and Great
Vessels), another index page will appear as shown in
Figure 3B. In this page, a
dropdown list of all the seconddigit indexes cascades from the first-digit
index (index [5]). In this example, nine entries are listed (e.g., [53] =
Cardiac Valve, [54] = Coronary Vessel). Again, each entry in this page is
linked to the next available subdigit index page. For example, if the user
clicks on the Coronary Vessel entry (index [54]), the system will jump to
another index page showing all the subcoronary vessels
(Fig. 3C). Similarly, the
pathologic index can be navigated in the same manner. For example,
Figure 3D is a screen capture
of the third-digit index page under the category Heart/Cardiac
Failure/Coronary Atherosclerosis (e.g., [761] = Less than 50% narrowing of
lumen). Again, a single mouse click on any of these six entries will take the
user to the next level. In this particular example, the next level will be at
the fourth digit.
View larger version (36K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. —Sample pages of American College of Radiology index. Screen
capture of subanatomic pages show second-digit indexes (B) and
third-digit indexes (C). As shown in C, all third-digit coronary
vessels are displayed (e.g., [541] = Right Coronary Artery, [542] = Left
Coronary Artery, etc.). In this example, cascading list ends at third-digit
index page because no other subindexes exist for subcoronary vessels. However,
for other body parts anatomy index can be extended to four digits (e.g.,
[4331] = Skeletal System/Wrist and Hand/Carpal Bone/Scaphoid).D, Screen
capture of subpathologic page shows third-digit indexes.
|
|
View larger version (38K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3C. —Sample pages of American College of Radiology index. Screen
capture of subanatomic pages show second-digit indexes (B) and
third-digit indexes (C). As shown in C, all third-digit coronary
vessels are displayed (e.g., [541] = Right Coronary Artery, [542] = Left
Coronary Artery, etc.). In this example, cascading list ends at third-digit
index page because no other subindexes exist for subcoronary vessels. However,
for other body parts anatomy index can be extended to four digits (e.g.,
[4331] = Skeletal System/Wrist and Hand/Carpal Bone/Scaphoid).
|
|
Free-Text Search
Further enhancing the user interface is the free-text search function.
Figure 4A is a screen capture
of the Anatomy search page showing the results of the search for the word
"artery." In this example, eight indexes are found. Each index
displayed on this page is linked to its subdigit index page (i.e., if the user
clicks on the first entry, [541] = Right coronary artery, the system will jump
to an index page with the equivalent digit level
(Fig. 3C). The same mechanism
is used for pathology searches (Fig.
4B).
View larger version (51K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —Search and bookmark pages of American College of Radiology
Index for Radiological Diagnoses. Screen captures show free-text search pages
for anatomy indexes (A) and pathology indexes (B). C and
D, Screen captures show search page (C) and bookmark page
(D).
|
|
View larger version (39K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —Search and bookmark pages of American College of Radiology
Index for Radiological Diagnoses. Screen captures show free-text search pages
for anatomy indexes (A) and pathology indexes (B).
|
|
ACR Index Search
Similar to the free-text search, the ACR index search was designed for
searching specific ACR indexes. Figure
4C is a screen capture of the search results of an index
[541.763]. The results are displayed as cascading dropdown lists for both
anatomy and pathology indexes with links to subdigit index pages. The ACR
index can be stored for quick reference by clicking the Bookmark this Index
hyperlink located at the left bottom of this page
(Fig. 4C).
Index Bookmark
Figure 4D is a screen
capture of a sample bookmark page with 12 indexes stored. Each bookmark entry
is linked to its detailed index page. For example, if the user clicks on the
third entry [541.763], the results for both anatomy and pathology indexes for
this particular ACR index will be displayed
(Fig. 4C).
Performance
We have been using this Web-based ACR index for intranet access since March
2002. Today we are still using the original software and hardware
configurations for this system with no major modifications. The system has
been operating continuously 24 hr a day without interruption except for
necessary maintenance. Users are required to log on to the system using their
hospital personal identification keys.
Discussion
We described a Web-based computer system for radiologists to easily
navigate the ACR index for radiologic diagnoses. We used the official index
database supplied in the ACR index CD-ROM (4th edition). A total of 11,260
indexes (10 anatomic, 722 pathologic, 90 subanatomic, and 10,438
subpathologic) were integrated into this online system. Authenticated users
can browse the entire hierarchic ACR index system using cascading dropdown
lists and navigate between index pages using simple mouse clicks. On the basis
of feedback from faculty members, residents, and medical students, this system
was modified to also provide free-text searches, ACR index searches, and
bookmark functions to enhance the user interface. With this online access
capability, the ACR index has been easily used for standardized coding of
teaching cases across all sections and subspecialties in radiology. Without
such standardization, valuable teaching cases would be lost or misplaced. This
Web-based ACR index has been used successfully in our departmental intranet
for more than 2 years. All the ASP pages described in this article are
available as open source for academic institutions.
References
- The official Web Site of the American College of Radiology (ACR)
(http://www.acr.org).
Accessed January 12, 2004
- Renfrew DL, Bass SH, Albanese MA, Whitis BM. Hypercard coding
system for the ACR index for radiological diagnoses.
AJR 1992;158:669
–672[Abstract/Free Full Text]
- Sistrom C, Drane W. Networked ICD-9 coding system for a radiology
department. AJR2003; 176:335
–339
- Tran TH, Roach NA, O'Kane PL, Thune M. Creating a digital
radiographic teaching file and database using a PC and common software.
AJR 2000;175:325
–327[Abstract/Free Full Text]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
D. L. Weiss and C. P. Langlotz
Structured Reporting: Patient Care Enhancement or Productivity Nightmare?
Radiology,
December 1, 2008;
249(3):
739 - 747.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
