LYSTO: The Lymphocyte Assessment Hackathon and Benchmark Dataset

• DOI: 10.48550/arxiv.2301.06304
• Publication Date: 2023-01-01
• Authors: Yiping Jiao; Jeroen van der Laak; Shadi Albarqouni; Zhang Li; Tao Tan; Abhir Bhalerao; Jiabo Ma; Jiamei Sun; Johnathon Pocock; Josien P. W. Pluim; Navid Alemi Koohb...; Raja Muhammad Saa...; Shan E Ahmed Raza; Sibo Liu; Simon Graham; Suzanne C. Wetstein; Syed Ali Khurram; Thomas J. Watson; Nasir Rajpoot; Mitko Veta; Francesco Ciompi
• Keywords: Benchmark (surveying); Digital Pathology

We introduce LYSTO, the Lymphocyte Assessment Hackathon, which was held in conjunction with MICCAI 2019 Conference Shenzen (China). The competition required participants to automatically assess number of lymphocytes, particular T-cells, histopathological images colon, breast, and prostate cancer stained CD3 CD8 immunohistochemistry. Differently from other challenges setup medical image analysis, LYSTO were solely given a few hours address this problem. In paper, we describe goal multi-phase organization hackathon; proposed methods on-site results. Additionally, present post-competition results where show how presented perform on an independent set lung slides, not part initial competition, as well comparison lymphocyte assessment between panel pathologists. that some capable achieve pathologist-level performance at assessment. After hackathon, left lightweight plug-and-play benchmark dataset grand-challenge website, together automatic evaluation platform. has supported research oncology. will be long-lasting educational challenge for deep learning digital pathology, it is available https://lysto.grand-challenge.org/.