SLIM: Tool To Aid Pathologists in Separation of Prostate Cancer Gleason Grade 3 and 4
Shamira Sridharan, Jonny Liang, Abhijit Marar, Le Wang, Karan Shah, Ryan Tapping, Virgilia Macias, Krishnarao Tangella, Andre Kajdacsy-Balla, Gabriel Popescu. University of Illinois at Urbana-Champaign, Urbana, IL; University of Illinois at Chicago, Chicago, IL; Christie Clinic, Urbana, IL
Background: Spatial Light Interference Microscopy (SLIM) is an add-on module to the commercial phase microscope which enables us to obtain a refractive index map of an unstained, transparent sample of constant thickness. In the past SLIM has been used to diagnose prostate cancer, and now we propose to use it to help in separation of Gleason grade 3 vs. Gleason 4.
Design: Prostate tissue microarray (TMA) with cores of different Gleason grades was obtained from the CPCTR collection. Twenty-two unstained cores with Gleason grade 3 and 21 unstained cores with Gleason grade 4 prostate cancer were imaged using SLIM at 40X/0.75NA. We measured the solidity of 6-8 glands per core (patient), which is the ratio of the area occupied by the glands to the area of an elliptical fit for the gland. We also measured the anisotropy of stroma adjoining each of these glands, which is the square of the ratio of the phase gradient to the phase variance.
Results: Our results indicate that Gleason grade 3 glands have higher glandular solidity and lower anisotropy in stroma adjoining the gland compared to glands of Gleason grade 4. By creating a scatter plot of glandular solidity vs. stroma anisotropy for each gland in an image, we can separate Gleason grade 3 from Gleason grade 4 and therefore determine Gleason scores in a given image. Thirty- four cores out of 43 were correctly classified by SLIM (79%).
Conclusions: SLIM can be a valuable tool in aiding pathologists in Gleason grading of prostate cancer. Further work needs to be done to automate the process.
Wednesday, March 6, 2013 9:30 AM
Poster Session V # 294, Wednesday Morning