Construction of a Virtual Microscope that can Accurately Assess Diagnoses Rendered Over the Web
Faculty AdvisorRothschild, Anthony
UMass Chan AffiliationsDepartment of Psychiatry
Copyright the Author(s). All rights reserved.
Health Services Research
MetadataShow full item record
AbstractBackground:A Virtual Microscope was designed to serve as an interactive microscope emulator that presents, via the Internet, a complete digitized pathology tissue section. It mimics the use of a microscope in both the stepwise increase in magnification and in lateral motion in the X and Y Cartesian directions. This permits a pathologist to navigate to any area on a slide, at any magnification, similar to a conventional microscope. Objective:The aim of this study was to construct a Virtual Microscope and assess the accuracy of diagnoses rendered over the Web. Methods:: 10 hemotoxalin & eosin stained glass slides comprised of cases from the general surgical pathology section of the University of Massachusetts Medical Center were randomly selected and digitized. They were then presented to 5 pathologists at various levels of training over the Internet. Participants were asked to examine each slide online and provide a diagnostic classification, comments, and confidence level using online feedback forms. Results:The overall concordance rate with the standard diagnosis was 44/50 (88%). This includes participants who submitted the standard diagnosis as part of a differential diagnosis. All participants were able to correctly diagnose the major pathologic process (i.e. benign vs. malignant) for each virtual slide 50/50 (100%). Factors which negatively influenced confidence were cited a total of 52 times. Poor image quality was cited in 20 out of 50 cases (40%), patient history was cited as necessary in 5 of 50 cases (10%), insufficient tissue was cited in 5 of 50 cases (10%), insufficient magnification was cited in 10 of 50 cases (20%), and more workup (i.e. immunohistochemical) was cited as necessary in 12 of 50 cases (24%). Image quality, image magnification, and a participant’s level of confidence in a specific diagnosis are not correlated with diagnostic accuracy. Level of training was associated with perceived negative factors regarding the virtual slide format, but was not associated with diagnostic accuracy. Conclusion:: Results from this study indicate that Virtual Microscopy can be used to make a correct diagnosis. Drawbacks include the need for high speed Internet connectivity and the availability of only one focal plane. However virtual microscopy may be considered a useful tool in medical school education, distance learning, and remote diagnosis.
Permanent Link to this Itemhttp://hdl.handle.net/20.500.14038/49372
The author participated in this study as a medical student in the Senior Scholars research program at the University of Massachusetts Medical School.