Static image analysis of skin specimens: the application of telepathology to frozen section evaluation

Although the ability to transmit high-resolution images of histopathological sections could have a profound impact on the practice of pathology, the application of video microscopy to the daily activities of surgical pathology has not been rigorously evaluated. In particular, certain aspects of video microscopy relating to frozen section evaluation have not been adequately assessed. We conducted a retrospective analysis of 48 excisional skin biopsy specimens encompassing a spectrum of benign and malignant lesions. To simulate an actual frozen section evaluation, only original frozen section slides were evaluated. Fields were selected and digitized (Roche Image Analysis System) by a pathology resident. Two sets of diagnoses were subsequently rendered by a surgical pathologist, the first set based on the digitized images and the second based on direct microscopic examination of the histological slides. The two sets of diagnoses were compared, and the concordance rates were as follows: malignant diagnoses, 100%; benign diagnoses, 100%; positive margins, 96%; negative margins, 99%. One (4%) of the 25 positive margins was indexed as negative by image analysis. Conversely, one (1%) of the 121 negative margins was indexed as positive by image analysis. In both of these cases, error was attributable to selection and digitization of an inappropriate field. We conclude that telepathology of static images is an accurate method of evaluating frozen sections of skin lesions. Potentially, this technology could be applied to the frozen section evaluation of other lesions as well. Static image analysis is, however, susceptible to errors induced by inappropriate field selection, emphasizing the need for trained and skillful personnel on both sides of the video camera.