Comparison of virtual and glass microscopy in dermatopathology testing

| December 9, 2008

I recently came across this abstract presented in October at the American Society of Dermatopathology meeting in San Francisco.  There was an accompanying news item on the story in a nursing news journal online of all places on the presentation and subsequent discussion as well.

Little is published on the use of virtual microscopy as an educational tool or as a testing tool, particularly among students or residents from a single institution, let alone several institutions. 

One item of interest in the accompanying news story is mention of 23% of residents between dermatology (who receive histopathology training in their course of their residency) and pathology having experience with virtual slides.  Certainly that number is at least 20% more than 5 years ago and will continue to increase with increased exposure with more medical school and resident program adoption.  Increased exposure online will increase this percentage and use as well.

Presumably these were dermatopathology cases and it is not clear what vendor(s) scanner or viewer(s) were used to share these cases.

The study provides one more objective and subjective piece about the use of whole slide images in education and for testing. 

If a manuscript is published about the study would be particularly interested in break down of number and opinions of dermatology residents compared with pathology residents.  The subjective complaints of the participants seem to be more related to PC and monitor issues rather than the tecnology itself.


L. Koch1; J. Woosley2; J. Lampros1; L. Delong3; S. Chen3

1Eastern Virginia Medical School, Norfolk, VA, USA 2University of North Carolina, Chapel Hill, NC, USA 3Emory University, Atlanta, GA, USA

Virtual microscopy is currently being used in medical schools, in resident education, in-training examinations, and certification examinations. There are, however, no randomized studies comparing the diagnostic accuracy and acceptability of virtual microscopy to traditional glass microscopy among resident physicians. The current study compared diagnostic accuracy using virtual and glass slide microscopy at the resident level. Dermatology and pathology residents at 14 institutions were given a randomized combination of 10 virtual and 10 glass slides of dermatopathology disorders and were asked to correctly select the diagnosis from a list of foils. They were also asked to report their subjective experiences with virtual microscopy. 132 residents performed equivalently (p=0.90) in making diagnoses with virtual microscopy (5.49, SD 1.76) compared to traditional glass microscopy (5.46, SD 2.07). The order of administration of virtual versus glass slides did not affect performance.

While performance was equal, traditional glass slides were preferred by 119 of the respondents. The majority of residents endorsed the use of virtual microscopy as a learning tool while fewer favored its application as a testing tool (76% vs. 43%, respectively). While residents may currently prefer glass slides, this study supports the use of virtual microscopy in testing.


Medical residents made similarly accurate diagnoses using virtual microscopy or traditional glass microscopy, but said they’d prefer glass slides when being tested, results of a randomized study of 132 residents found.

The dermatology and pathology residents at 14 institutions diagnosed neoplastic and inflammatory skin conditions after examining slides in virtual (computerized) and glass format. Each of 20 cases appeared in both virtual and glass slides, but the order in which the residents saw them varied among randomized subgroups of participants.

Residents correctly diagnosed a mean of 5.5 cases on glass slides and 5.6 cases on virtualslides, Dr. Laine H. Koch and associates reported at the annual meeting of the American Society of Dermatopathology.

The order in which slides were viewed did not change diagnostic accuracy between groups. The rate of correct diagnoses increased with each additional year of residency, but did not differ between same-level residents based on the use of virtual or glass microscopy, said Dr. Koch of Eastern Virginia Medical School, Norfolk.

When asked their opinions of the experience in an 11-item questionnaire, however, 79% of residents said they think virtual microscopy is useful for learning, but only 47% said it is useful for testing.

Approximately 40% complained of one or more problems during virtual microscopy. Complaints from 31% of residents pointed to fuzzy images on the computer screen, while 10% reported poor color on the screen, 9% said the image froze on the screen, and 8% complained of poor contrast. Four complaints each were reported by less than 5% of participants: difficulty starting the computer, inability to adjust the screen, computer failure, or power failure.

At one institution, all residents complained of problems with image clarity, while another institution produced no complaints about clarity, and responses varied from the other 12 institutions. It may be that fewer image problems occurred when all residents completed the exam in a computer lab, compared with using the technology on a computer of their own choice, Dr. Koch said.

Glass slides have been considered the highest standard for diagnosis and for medical education in histology and pathology. They are easy to prepare and are part of learning to use light microscopes, but can be expensive to purchase, maintain, store, and distribute for educational purposes. Virtual slides can be duplicated endlessly and inexpensively, stored and catalogued easily, and can be made available to many people in a variety of locations.

As a result, medical schools increasingly are incorporating virtual microscopy into curricula, and in some programs virtual microscopy has replaced traditional microscopy entirely, Dr. Koch said.

At the start of the study, 23% of residents reported some experience with virtual slides, usually from online or Internet sources. Diagnostic accuracy after examining virtual slides in the study did not differ significantly between residents who had or had not used virtual slides before.

Few studies have looked at the performance of students using virtual or glass microscopy. One study that randomized first-year medical students to teaching and testing using virtual or traditional microscopy found no significant differences in testscores (Clin. Anat. 2007;20:565-70).

Investigators in the current study prepared glass slides using standard hematoxylin and eosin staining techniques. Virtual slides were created using a slide-scanning machine and were distributed to the medical schools on compact discs. Dr. Koch and her associates have no association with the company that makes the scanning machine, but are considering negotiating an educational discount to obtain a one, she said.

One dermatology resident in the audience who participated in the study said the main problem with virtual microscopy was the slow speed of loading images. “Any time you moved around [in an image], it was like waiting for the dial-up connection to have your Internet,” she said. “It was really cumbersome. Once that’s straightened out, I think it will be super.”

Category: Education, Histology, Whole slide

Comments are closed.