EPUBs are an experimental feature, and may not work in all readers.

Creating RSNA EJ, the online journal of the Radiological Society of North America (RSNA), was an exciting undertaking; it was an opportunity to experiment with the many capabilities of the World Wide Web to better communicate scientific and clinical research in radiology. After a year of publication, we believe we have shown that the Web offers advantages found in no other medium.

The Electronic Communications Committee and Publications Council of the RSNA had in 1995 approved the implementation of an online journal and provided a budget and staff support for a three-year pilot project for the first online-only journal published by the Society. The committees recognized that Internet communications were improving, that access to the Internet was growing, and that de facto standards for creating, distributing, and browsing material on the World Wide Web were emerging. Given those developments in cyberspace, and the advantages of electronic publishing, the time had come to launch RSNA EJ as an investment in the future of radiologic research and education. The cost of the journal would be reasonable, representing the salary of the managing editor, the only full-time staffer, time spent by other RSNA staff members on the journal, and fees paid to a computer consulting company that maintains the server and provides programming support for the journal. All costs would be borne by the Society until other sources of revenue—such as advertising and subscriptions—become available.

Why Publish on the Web?

The advantages of Web publication are many. An online journal can overcome some of the problems inherent to print publication, particularly space limitations imposed by the ever-increasing price of paper and postage, and the time constraints inherent in the production of articles for publication. Articles in an online journal theoretically have no length limitations — although as in all scientific journals, articles should not contain one more word or image than is necessary to convey the relevant information — and can include as many radiologic images as are appropriate. Also, an article can be published — posted on the Web — as soon as it is accepted and prepared for publication: It need not wait for the appearance of the next issue.

The specific improvements online publication achieves are

  1. use of multimedia and the Java programming language — publication not only of text and still images but of sound and motion;

  2. interactivity — with applications such as Java, cgi programs, and forms, continuing medical education (CME) credit can be offered through online tests, or images can be presented that can be manipulated by the user (see the article Radiographic Image Processing with Java in RSNA EJ; you will have to register first);

  3. speed — faster dissemination of scientific information by expediting peer review, editorial and production work, and distribution;

  4. incorporation of hypertext links — linking to other electronic information, such as links between reference citations and article abstracts and between references to material on the Web (e.g., a teaching file) and the remote sites themselves;

  5. enhanced discourse — expanded and facilitated scientific discourse about research, enabling letters to the editor and discussions concerning articles to be created and distributed online and linked to the articles in question; and

  6. improved access — ready access to information through electronic search mechanisms.

Online publication also has a few disadvantages. One notable disadvantage from a radiologic perspective is that the author or publisher of an article has little control over the quality of images because of the diversity of computer-monitor pairs used by readers. In a sense, every reader sees a slightly different image. In conventional paper publication, the choice by the publisher of paper, ink, and image manipulation dictates the quality and uniformity of the image displayed.

Also, readers without laptop computers do not have the ability to read articles on a commuter train or in bed. Readers who must access the journal

"Some submissions have been returned because they had none of the features that require electronic publication; the submissions could just as well have appeared in a print journal."

over a modem may be hindered by the slow acquisition of large image or movie files (although readers can begin reading the text while the images download). To help alleviate that problem, we ask authors to limit the size of their still images, which must be in GIF or JPEG format, to 100 kbytes at most and to make inline images as small as possible, with a link to a larger version if necessary. An alternative for the reader is to use an off line browser (eg, WebWhacker), which can access RSNA EJ while the reader works, or even sleeps, and download the entire site or any designated part of it to the reader's hard disk, to be perused at hard-disk speed whenever convenient.

How RSNA EJ Articles Differ from Print Articles

RSNA EJ articles are peer reviewed and cover all aspects of radiology: diagnostic, interventional, therapeutic, and nuclear medicine, in addition to computer applications. However, while many different types of articles are appropriate for RSNA EJ, the set can be succinctly differentiated as those which, by their electronic nature, cannot adequately be presented in a print journal. In other words, they offer one or more of the following features: Java, animation, sound, interactivity, outside links, or a quantity of images beyond the scope of a print journal (see Ultrasound of the Shoulder in RSNA EJ, which presents hundreds of images, in addition to several movie files and a CME test).

RSNA EJ also allows the user to read nonlinearly; in other words, the user might leave the article and the journal entirely to follow a link to a Web page halfway around the world or jump between a movie file, for example, and text within the article itself, or return to the table of contents. As a result, navigation aids are included and the journal is designed and structured to orient the user and ease the way.

Design of RSNA EJ

RSNA EJ opens on the registration page. Users are required to register so that the RSNA can gather some demographic data, such as the subspecialty, primary professional role, and RSNA membership status of the user. The user is asked to provide an ID and password, which then must be used to access the journal.

The opening page of the journal proper consists of three frames: the logo at the top, a menu on the left for the active volume (Volume 1 presently), and the main frame on the right into which the menu choices open. The frames help to orient the reader. The menu choices include four tables of contents — for articles, for abstracts of articles, for editorials, and for letters to the editor. In addition, there is a What's New button that opens a list of everything in the active volume of the journal in reverse order of publication date, allowing users to quickly see if anything new has been published since their last visit. Below those are the departmental buttons (you will have to be registered to access those pages):

When Volume 1 is complete, at the end of the calendar year, it will be archived and a button link to it will be added to the opening page of Volume 2. All articles ever published will be readily accessible at all times.

The reader accesses an article by clicking on the article's title in the table of contents. That brings up the article's opening page in the full browser window. A button at the top links back

"Statistical data on reviews can be gathered and analyzed easily."
to the table of contents, and another button links to the Letters to the Editor form. Navigation aids within an article are the responsibility of the author, although they may be added or changed if the managing editor feels the structure of the article is confusing. Readers wishing simply to browse the abstracts can click on the first abstract title in the abstract table of contents, bringing up that abstract, and then click on the Next Abstract button at the top to retrieve each abstract sequentially. There are also buttons at the top of each abstract linking to the relevant article, the Letters to the Editor form, and back to the abstract table of contents.

Processing Articles Authors are asked to fill out an online submission form. After filling out the form, the author receives an automatic acknowledgment and a manuscript number by e-mail. The manuscript number becomes the filename. The author sends the manuscript, plus any graphic files or tables, by anonymous binary file transfer protocol (FTP) to the RSNA server. All manuscripts must be submitted in HTML. The files are placed in the appropriate directory, and an e-mail message requesting a review and providing review instructions is sent by the managing editor to three reviewers with expertise in the topic of the article. Those reviewers enter their reviews and scores in an online form that allows statistical data on reviews to be gathered and analyzed easily. The editor selects the portions of the reviews to be conveyed to the author, and the managing editor includes those in an e-mail message requesting revision. The author resubmits the revised manuscript, which is then copyedited in an HTML editor. All HTML codes are verified as conforming to the latest HTML standard, and links are checked for viability. The author is then asked to retrieve the files by FTP, approve the copyediting, and respond to any queries. The author returns the files, and the managing editor makes any necessary changes before the final version is posted on the Web site. Once the article is posted, the author is not permitted to make any additional changes.

The system works reasonably well. Our authors have had no problems with the submission procedure or in revising their articles. Some submissions have been returned to the author because they had none of the features (i.e., multimedia, interactivity) that require electronic publication; the submissions could just as well have appeared in a print journal. A small number of reviewers have been unable to review manuscripts because they lack the technical resources. Very few, however, have had problems filling out the online review form. One stumbling block in the process has been that a number of reviewers have changed their e-mail addresses without informing us, resulting in reviewing delays.

CME Availability

RSNA EJ offers convenient CME credit for selected articles. Physicians may receive up to a maximum of three hours in Category 1 credit by scoring at least 80% on an online multiple-choice quiz provided by the author and peer reviewed along with the article. A cgi program accepts the input and informs the participant if he or she has passed. The program then allows those who have passed to print out a Certificate of Participation.

Other Considerations

RSNA EJ has been online for a year now and has received a favorable response from readers and from members of the RSNA, who are supporting this new venture in scientific publishing. As of the beginning of July 1997, the Journal had 2,772 registered users. We have installed a tracking program that acquires data on the number of visits to the journal by profession, specialty, age range, state or country, and society membership; it can also provide the top 20 documents visited and the top 20 users. In a recent month (April 1997), 1,244 users "visited" the site, which we define as staying at least 30 minutes. Of those visitors, 785 were under age 45, perhaps indicating that the younger generation is more familiar and comfortable with computers. The site was accessed 28,549 times during that same month. We do not now tell authors the number of times their article has been accessed, but that is a future possibility.

At present, there is no subscription charge for the journal, but that also remains a possibility once the journal becomes established, and may be necessary to keep the journal viable. We are open to advertising as a source of revenue and are actively pursuing companies that advertise in other RSNA publications. We have publicized the journal through print ads in RSNA publications, through demonstrations at our annual meeting and at meetings of other societies, and through announcements in other relevant Web publications and site listings.

The main drawback of the journal now is the lack of enough published articles to allow its indexing in Index Medicus and Medline. That has put a damper on the number of submissions, a kind of publishing Catch-22 that affects all new journals. Once we do reach the required number of articles for consideration (20-25), and are then indexed, we expect the number of submissions to increase substantially. We expect the journal to eventually be accepted by the wide radiologic community because it offers possibilities and dimensions that a print journal cannot match.

Laurens Ackerman is Section Director of Medical Informatics in the Department of Radiology at Rush Presbyterian-St. Luke's Medical Center in Chicago, Illinois, and spends most of his clinical time doing mammography. In his past life as Chairman of the RSNA Electronic Communications Committee and infoRAD, he promoted connectivity for the RSNA by connecting the meeting with high-speed networks, the Internet, and DICOM (digital imaging and communications in medicine).

He is the editor of RSNA EJ. His interests include hiking, biking, guitar, tennis, and Montana. He has been trying to build an airplane, but he finds his job too interesting. Dr. Ackerman may be reached at lackerma@rad.rpslmc.edu.

Al Simonaitis has been managing editor of RSNA EJ since its inception in 1996 and works at the headquarters of the RSNA in Oak Brook, Illinois. He was previously the managing editor of the Journal of Magnetic Resonance Imaging, a print journal, for five years, and before that a manuscript editor for Radiology, the flagship publication of the RSNA and the premier journal in the field of radiology. He is fascinated by the power of computers and the Internet and is happy to no longer be dealing with paper manuscripts and blue pencils. You can contact him by e-mail at simonait@ej.rsna.org.