Title page Introduction Internet Teaching file Sharing Conclusion References
Table 1 Table 2 Table 3 Table 4 Figure 1 Figure 2

Case-based Teaching File

Why Make an Electronic Case-based Teaching File?

A digital teaching file has both advantages and disadvantages relative to a film-based teaching file. The primary advantage is the potential for widespread network access, both from multiple clinical sites within a medical center and by residents and staff at different institutions. Multi-institutional efforts allow the most valuable teaching cases to be viewed by many individuals, while spreading out the effort of creating the teaching cases among a larger number of faculty. Such use requires ready network access using commonly available programs, as well as a method of ensuring the quality of the "digitally published" cases. Other advantages of an electronic teaching file include the capability for cine display, the ability to search based upon diagnosis or type of study, the ability to rapidly incorporate new information based upon user feedback and new publications, and the ease with which slides can be made using a film recorder.

Use of a digital teaching file has several disadvantages, however. Images (especially radiographs) may not be stored at their full resolution due to limitations of the storage and network speed; additionally, if the films were not acquired in electronic form, there may be slight degradation of the image during the process of scanning the image. Presentation of cases to a group of residents requires use of a large monitor or expensive projection equipment. Finally, unless creation of digital cases is as easy as the making of cases in a film-based teaching file, there will be resistance to the addition of new teaching material.

Case Creation and Presentation

Importing of images into an electronic teaching file may be accomplished directly utilizing a picture archiving and communication system (PACS), by manual transfer of images from a digital acquisition computer (e.g. CT, MRI or gamma camera computer), or by digitization of film copies of a radiographic study. Once the images are obtained, it should be easy to combine text and images into a teaching file. One medium-level approach is the use of a hypertext markup language (20). However, even with aid of editing software, creating cases can still be arduous with these medium-level tools. A higher-level approach provides "fill in the blank" templates for the user, with automatic formatting and creation of appropriate hypertext links. This is the approach taken by Wallis et al (11) in the creation of an electronic teaching file. A more general hypermedia authoring system was developed by Greenes et al (21), which provides a "building block" approach utilizing tools which allow entities such as hypertext, images, buttons, animators, and simulators to be combined into a dynamic book . Other tools intended for single-user multimedia development on personal computers include Hypercard, Supercard, Macromind Director, and interText; however these tools are less applicable to a collaborative network effort.

To be useful for teaching, the cases should be viewable both without the diagnoses (as "unknowns") and with the accompanying diagnoses and discussion (as "known" cases). Useful components of the presentation are shown in Table 2

New Strategies Made Possible by Hypermedia

Ideally, while stepping through a case, the viewer might be "charged" for each additional test or imaging procedure performed, with automatic updating of diagnostic certainties and feedback regarding the effectiveness of the diagnostic workup. In practice, however, this is difficult; precise information regarding the sensitivity and specificity of imaging procedures is not always available, and the optimal workup may depend on the level of certainty needed in specific clinical settings, as well as the ability to affect patient outcomes in each disease under consideration.

For the purposes of a teaching file, it may be best to be able to "share" text among cases, that provides a general discussion of diagnostic imaging in that clinical setting. For example, following the discussion specific to individual cases, all cases of ventilation- perfusion imaging might have a link to a general page offering information regarding the utility of such imaging in the diagnosis of pulmonary embolism. As new data became available, this single common page could be updated, rather than having redundant (and possibly outdated) information residing in each individual case discussion. A prototype system which illustrates the use of known and unknown pages, flexible case formatting, shared text, and links to similar cases is available (11). A system demonstrating the use of online diagnostic workup strategies has also been developed (21).

Indexing and Searching

The ability to search can be added more easily to an electronic teaching file than to a film- based one. A simple approach might be to allow text-word searches of the case. More complex searching requires prior classification of the cases using a standard scheme or structured vocabulary. The Index for Radiological Diagnoses, developed by the American College of Radiology (ACR), was developed for use with radiology teaching files (22). At its highest level, each diagnosis is classified from the lists by both Anatomic and Pathologic codes (see Table 3).

More specific sub-specification of both the anatomic and pathologic codes is available, and is particularly necessary for the last three pathologic categories. Indexing programs using this classification are available for personal computers from ACR, and World-Wide-Web links to Internet sites demonstrating searching and indexing are available. Although indexing using the full ACR code is desirable, it is likely that searching using the broad categories above coupled with the ability to search text-words in the diagnosis will suffice for most users. One area of deficiency of the ACR classification is that of exam type. While it is possible to code broad groups of studies using this scheme, specific examination types may not be listed. For example, Gallium-67 imaging, In-111 labeled white blood cell imaging, and Tc-99m HMPAO white cell imaging are all specified as "inflammation-avid imaging." Since searching by exam type is desirable for a nuclear medicine teaching file, a standardized list of exam types has been developed at Mallinckrodt Institute of Radiology, which is available on-line.

As usage of on-line systems grows, it will be beneficial to have cross linking among the ACR classifications and other standardized nomenclatures, such as the MESH terminology, SNOMED, ICD9, and other vocabularies. The Unified Medical Language System (UMLS) developed by the National Library of Medicine provides this mapping designating a concept and providing links to all variations of that concept, including know synonyms, in other taxonomies (23).

Another relatively unexplored area is that of classification by scintigraphic findings. It may be useful to be able to retrieve cases that have similar characteristics to a clinical case under discussion, such as increased gallium uptake in the lungs. To date, these needs have been most effectively addressed by books of gamuts (24) and differential diagnoses (25). Initial feature-based classification schemes have been proposed in other areas of radiology; an example is the Image/ ICON system of Swett et al. (26).

Reprinted by permission of the Society of Nuclear Medicine from: Parker JA, et al.: Collaboration using Internet for development of case-based teaching files. Journal of Nuclear Medicine 1996; 37:178-184.

Title page Introduction Internet Teaching file Sharing Conclusion References
Table 1 Table 2 Table 3 Table 4 Figure 1 Figure 2