PAIP2019 is the liver cancer segmentation challeng, held as a part of MICCAI2019.
The objective of the challenge is to assess advanced computer algorithms for an automated evaluation of liver cancer in whole-slide images (WSIs). It includes two tasks: 1) Task 1: Liver Cancer Segmentation, and 2) Task 2: Viable Tumor Burden Estimation.
This repository contains the deep learing-based methods that we developed for the challenge.
The methods were ranked 5th and 3rd for Task 1 and 2, respectively.
: This file is for training classification model : This file is for generating patches for training : this file is for prediction : This file is for training segmentation model : This file is test result after removing samll componenets by using connected components