Computer Vision News - March 2019

12 Challenge: BreastPathQ Computer Vision News The SPIE (the international society for optics and photonics), American Association of Physicists in Medicine (AAPM), and the National Cancer Institute (NCI) developed a Grand Challenge for the quantitative task of determining cancer cellularity from pathology hematoxylin and eosin (H&E) slide patches of breast cancer tumors. Cancer cellularity is the percent area (0-100%) of the tumor bed that is comprised of invasive or in situ tumor cells and is one part of the MD Anderson residual cancer burden assessment approach with applicability, for example, in assessing neoadjuvant treatment of breast cancer. This intro and almost all the article are courtesy of Nicholas Petrick, Deputy Director Division of Imaging, Diagnostics and Software Reliability at the Center for Devices and Radiological Health, U.S. Food and Drug Administration. Nicholas led this challenge effort together with Kenny Cha (also from the FDA), Shazia Akbar and Anne Martel (both from Sunnybrook Research Institute) among others. Challenge Challenge synopsis The challenge included a training (2579 512x512 patches from 46 patients), validation (185 512x512 patches from 4 patients) and test phase (1121 512x512 patches from 18 patients) using digitized H&E stained breast cancer slides. The data was collected at the Sunnybrook Research Institute, Toronto as part of a research project funded by Canadian Cancer Society and IRB approval was obtained to allow the anonymized data to be shared. The training and validation phases included feedback to the participants on their algorithm’s performance and had visible leaderboards allowing participants to compare their performance with other algorithms. The test phase was blinded without any performance feedback available to participants until after the challenge closed. The performance metric selected was predication probability (p k ), which assesses the ordering of the patches between the method and the reference but not the absolute cellularity scores for each patch and ranges between 0.0 and 1.0. The reference standard (“truth”) was a cellularity scores from one pathologist for the training/validation patches and cellularity scores from two pathologists for the test patches. A total of 100 qualified algorithms were submitted from 37 different groups. The winning algorithm achieved an p k =0.94 with algorithm performance ranging from p k =[0.21, 0.94]. Nicholas Petrick