Systematic Evaluation of Flatbed Scanners for Digital Gross Imaging
Winner - Pathology InformaticsLiron Pantanowitz ; Baystate Medical Center/Tufts University School of Medicine;
Content:
Gross imaging is an important element of anatomical pathology. Photography of gross surgical and autopsy specimens using digital cameras is being widely adopted. Acquiring digital images of gross pathology with a flatbed scanner (FBS) has been proposed as an alternative method to using a camera that is fast, reliable and simple. However, there has been no systematic evaluation of this technique. The aim of this study was to evaluate the FBS in gross imaging compared to a digital camera.
Technology:
HP Scanjet 8270 flatbed scanner (Hewlett-Packard, Palo Alto, CA); Sony digital camera DXC-390 (Sony Corporation, Tokyo, Japan); Dell Optiplex 745 desktop computer (Dell, Round Rock, TX); CoPath PicPlus (Cerner Corporation, Kansas City, MO).
Design:
A variety of fresh and formalin-fixed surgical specimens (n=25) including breast, viscera (colon, kidney, liver, thyroid, uteri, ovary, thymus), lymph nodes, bone, and soft tissue tumors were both photographed and scanned by six prosectors. Specimens were scanned with the FBS lid closed and open. All images were 24 bits/pixel and saved in JPEG format. For each specimen components of the photography (image acquisition time, focus, fitting the specimen in the image, ease of preparation and clean up) and image quality (focus, resolution, and presence of artifacts) were compared.
Results:
Photography with a digital camera was preferred by 56% of prosectors, 32% preferred the FBS, and 12% found both techniques comparable. These findings were unrelated to specimen type. Image acquisition with the FBS took 4 seconds and required no focus adjustment. Photography with the camera took longer, and prosectors experienced difficulty focusing the camera in almost half of the cases (48%). The FBS was able to focus at different depths in the field (3-D structure) in all specimens. Shadows were more commonly seen with the camera. Scanned images with the lid open produced a uniform black background. In only 1 case did a total colectomy specimen not fit onto the FBS plate. Image quality produced by the FBS was rated superior in 44% of cases. Enlargement of images up to 200% was better for those acquired with the FBS. Fluid (blood, serous, formalin) on the scanner plate in some cases obscured specimen detail, especially in cross sections of solid organs. With compressible specimens the lid of the FBS flattened certain areas.
Conclusion:
We recommend the use of a FBS as an inexpensive, complementary imaging tool for acquiring high-quality digital images of gross specimens. The FBS provides rapid image acquisition of all specimen types and is operator independent. Dry specimens without blood (e.g. fixed tissue) that can easily fit on to the scanner are ideal for digital imaging with a FBS. To avoid tissue being compressed, gross specimens can be scanned with the lid open.
