Automatic Prostate Cancer Grading System Based on 3-D Histo-Pathological Images

Iri tliis paper we have bre1)7y described a triochitre visioti systcrii for grading the prostate cancer based otr 3-0 histopathological ititages. Tlre 3-D inrages arc obtaitlcd frorti corfocal ~iiicroscope as a stack of optical sectiorrs. Various heuristics atid cotivetttiotial ittiage ettharicctttetit teclrtiiqites are used to ctihatice tlie irltage features. Dfiretit itrtcrgr segrrietitation rricthods iriclr~dir~g high-level trtethods are etrrployed to isolatcd the cells it1 a volutrictric ittiage. We cancer tumor based on the three dimensional image analyses. The three dimensional images of the thick tissue specimens are obtained using Confocal Laser beam Scanning Microscope (CLSM). CLSM produces a stack of 2D optical sections that gives the three dimensional information of the tissue specimen. Such a stack of optical sections suffers from uneven illumination, non-isotropic voxels due to relatively poor axial resolution, instrumentation noise, etc.. have presented various nletliods for quatltitatioti of rlic tissue arcliitecture and cell distributiorl patterti. Tllis s)lstertr 2* Image Enhancement and car1 be used as a supportirtg tool for- progtiosis and Reduction diagrtosis of the prostate carrcer patierits.