[2156] SLIM as an Optical Tool To Support Pathologists in Prostate Diagnosis

Shamira Sridharan, Ryan Tapping, Andre Kajdacsy-Balla, Krishnarao Tangella. Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaigny, Urbana, IL; University of Illinois at Chicago, Chicago, IL; Christie Clinic and University of Illinois at Urbana-Champaign, Urbana, IL

Background: Light microscopy is the backbone of anatomic pathology. Of the methods that could help to pathologists to obtain information beyond that obtained through usual staining methods. Spatial Light Interference Microscopy (SLIM) is an optical imaging technique that measures the phase and thus refractive index of unstained tissue. SLIM has been previously used to differentiate between benign and malignant prostate glands. In order to better evaluate SLIM as a diagnostic and prognostic tool we now attempt to use it to help recognize Gleason patterns.
Design: Using this phase map we calculated the anisotropy factor g which is the average cosine of the scattering angle associated with a single scattering event. g is related to both the variance of phase in a given tissue region and the mean value of the gradient of the phase. Two pathologists marked areas of prostate tissue microarrays where they agree on Gleason score. Anisotropy was calculated at 10X magnification.
Results: In the images obtained using SLIM, individual stromal strands can be identified. Excellent separation could be found between benign appearing and malignant areas, with a degree of accuracy of less than 1% overlap. Gleason 5 areas did not overlap with the other patterns. However, when we measured the anisotropy values for a single layer of stroma surrounding individual grade 3 and grade 4 glands in 55 tissue microarray cores with grade 3 and 33 cores with grade 4 tumor we found approximately 40% overlap. Even though the overlap precludes clinical usefulness, the mean value for anisotropy in grade 3 stroma was 0.9953, with 0.0053 SD; and for grade 4 stroma, the mean value was 0.9903 with 0.0129 SD; p-value= 0.004.
Conclusions: While SLIM is useful for differential diagnosis between benign and malignant, and also to classify Gleason grade 5 areas, additional investigation is needed in order to get a better separation between Gleason grades 3 and 4. We believe further analysis with images at 40X magnification will improve this distinction. This work will serve as background preparation in order to later address the possibility that SLIM could be used as a prognostic tool. Interestingly, the differences among Gleason grades are detectable by tissue stromal anisotropy, and not the optical properties of the epithelial component.
Category: Techniques

Monday, March 19, 2012 1:00 PM

Poster Session II # 313, Monday Afternoon


Close Window