Validation of Pathologist Use of Whole Slide Images for Remote Frozen Section Evaluation.
Joel P Ramey, KarMing Fung, Lewis A Hassell. Oklahoma University Health Sciences Center, Oklahoma City; Oklahoma City Veterans Affairs Medical Center, OK
Background: Telepathology has been used for remote evaluation of frozen sections (FS) with acceptable results for many years. Advances in whole slide imaging (WSI) and other technologies introduce new possibilities to apply to this challenge. Validation, regulatory and credentialing issues associated with WSI appear important. Previous studies have limited case types evaluated rather than looking at a broad spectrum of cases referred for FS evaluation.
Design: A series of 72 consecutive cases with FS requests were selected for study. FS slides were scanned using an Aperio Scanscope scanner at 20x magnification. All materials utilized in the FS consultation including gross images (if available) and descriptions, clinical information, cryostat sections and cytologic preparations (if available) were catalogued with the images. Scan times were recorded. WSI FS cases were presented to 8 pathologists for diagnosis on desktop or laptop computers using Webscope or Imagescope viewing software. Time to diagnosis and other parameters of evaluation were noted. The WSI FS results were compared to original reported and final diagnosis to derive kappa values. Cases with discrepancies were subjected to further consensus review of the original glass slides. Cases are available to pathologists at http://moon.ouhsc.edu/kfung/HR2010/Default.htm for additional data accrual and educational purposes.
Results: Average scan time was 2 minutes 14 seconds per slide. Computer hardware in use showed the following minimal specifications: monitor resolution 1024x768 pixels; processor speed 2.59gHz, memory (RAM) 2 gb. Evaluation times for WSI FS averaged 4 min 13 seconds/slide, of which an average of 55 seconds was spent at magnifications above 10x. Complete concordance between initial FS diagnosis (FSdx) and WSI FSdx was observed in 91% of cases. Minor discrepancies resulting in no change in management were observed in 7% of cases. Overall method kappa value was calculated to be 0.84. Type of FS request (margin, etc.) did not influence performance in this series.
Conclusions: WSI offer an acceptable solution for remote FS consultation on a wide variety of cases. Scanning and examination times should not pose a barrier to meeting service time demands intraoperatively. Existing standard hardware in use generally offers acceptable performance. Validation within institutions should include system processes or procedures, hardware and personnel considerations and be be thoroughly tested prior to adoption in routine use.
Category: Quality Assurance
Tuesday, March 1, 2011 11:00 AM
Platform Session: Section H 2, Tuesday Morning