[1352] Tumor Pathology Report in 3D Dynamic Image Format

B Castaneda, D Hosey, J Strang, DJ Rubens, KJ Parker, J Yao, Z Qu. University of Rochester, Rochester, NY

Background: Tumor pathology reports are a critical component in the clinical management of cancer patients. Currently, pathology reports of malignant neoplasm remain in text format with the occasional presence of selected 2D still images for marketing purposes. Images are considered the most comprehensive and intuitive form of information for morphological abnormalities. In this abstract, we propose and report a 3D image pathology reporting module using prostate as a model that allows direct and visual assessment of several critical tumor attributes such as size, volume, shape, location, extent and margin status.
Design: Intact prostate glands with carcinoma were received after radical prostatectomy. A landmark device, which consisted of two sets of four (3 mm diameter) mating metal prongs, was inserted into the specimen through the apex and base to provide fiducial markers for 3D reconstruction. Subsequently, a whole-mount protocol was followed and complete serial sections (4mm apart) were mounted on 70 x 50 mm slides and stained by routine H&E for microscopic evaluation. The carcinoma was outlined in the slides. Image segmentation techniques were used to extract the contours of the gland and the tumor. These contours were interpolated to create a 3D model which was dynamically visualized using a web interface.
Results: Figure 1 illustrates the implemented web interface showing a 3D model of the prostate gland (cyan) and the carcinoma (red). Users can access this interface to view a colored 3-D model of a tumor from several angles to reveal the tumor size, shape and location in the specimen.


Conclusions: We propose a 3D pathology tumor report by combining well established pathology specimen whole-mount methods, image processing techniques, and 3D modeling software tools. This reporting method allows for a more comprehensive appreciation and assessment of several critical cancer attributes. Furthermore, this 3D reporting methodology can be applied to other organ systems (such as liver) and used to validate radiology imaging modalities.
Category: Informatics

Wednesday, March 11, 2009 9:30 AM

Poster Session V # 215, Wednesday Morning