figureposted on 19.04.2019, 00:42 by VISILABVISILAB
Nuclei segmentation is a very useful tasks of image processing for cancer grading and for prediction of treatment effectiveness in digital pathology.
This dataset contains 20 samples of digital images of 1000x1000 taken at 40x and their segmented nuclei masks. The masks contain around 14,600 nuclear boundary annotations in total. The mask format are also images of 1000x1000. 6 types of carcinomas are included, those are: breast, kidney, prostate, bladder, gastro-intestinal and colon.This dataset has been manually annotated by pathologist from the AIPDATH project.