KiTS19_ACE

KiTS 2019 challenge in MICCAI 2019
Team name : ACE (Asan Coreline Ensemble)
http://results.kits-challenge.org/miccai2019/manuscripts/sungchul7039_3.pdf

Training

• TO DO

Prediction

All checkpoints are located in checkpoint/. Checkpoints used in challenges will be updated.

• For searching ROI of kidney
  python evaluation.py --mode 1 --testset /path/testset
• For predicting kidney and tumor
  Select a mode using prediction. Before predicting kidney and tumor, RUN the mode 1 first.
  2_1 : coreline's model 2_2 : model with dice loss, normalization with tumor's mean and std and using ONLY ONE kidney in CT.
  2_3 : model with dice loss, minmax scaling and using ALL kidney in CT.
  2_4 : model with focaldice loss, minmax scaling and using ALL kidney in CT.
  2_5 : model with dice loss, normalization with tumor's mean and std and using ALL kidney in CT.
  python evaluation.py --mode 2_3 --testset /path/testset