/MNet_DeepCDR

Code for TMI 2018 "Joint Optic Disc and Cup Segmentation Based on Multi-label Deep Network and Polar Transformation"

Primary LanguageJupyter Notebook

MNet_CDR_Seg

Code for TMI 2018 "Joint Optic Disc and Cup Segmentation Based on Multi-label Deep Network and Polar Transformation"

Project homepageļ¼šhttp://hzfu.github.io/proj_glaucoma_fundus.html

  1. The code is based on: Keras 2.0 + Tensorflow
  2. The input image is the disc center patch with 800 x 800 size.
  3. The output is raw segmentation result without ellipse fitting.
  4. The pre-train model 'Model_MNet_ORIGA_pretrain.h5' is trained on ORIGA full dataset.
  5. Please cite the following paper:

[1] Huazhu Fu, Jun Cheng, Yanwu Xu, Damon Wing Kee Wong, Jiang Liu, and Xiaochun Cao, "Joint Optic Disc and Cup Segmentation Based on Multi-label Deep Network and Polar Transformation", IEEE Transactions on Medical Imaging (TMI), 2018. DOI: 10.1109/TMI.2018.2791488 (ArXiv version: https://arxiv.org/abs/1801.00926)

Update log:

  • 18.02.26: Add CDR calculation code (based on Matlab)
  • 18.02.24: Release the code