HealthcareCase Study

Chinese University of Hong Kong: Deep Learning Wins 2015 Gland Segmentation Challenge

Chinese University of Hong Kong: Deep Learning Wins 2015 Gland Segmentation Challenge

·11 min read
The Chinese University of Hong KongHealthcare Providers · Academic Medical ResearchResearch institution · Department of Computer Science and Engineering

Key People & Companies

The Chinese University of Hong Kong

DCAN

The Chinese University of Hong Kong · Internal

GlaS Challenge

Pheng-Ann Heng

Professor at The Chinese University of Hong Kong

Hao Chen

PhD Student at The Chinese University of Hong Kong

u-net

University of Freiburg · Framework

Nasir M. Rajpoot

University of Warwick

GlaS Challenge Contest

+ 17 more entities in the full study

Key Results

  • DCAN won the 2015 MICCAI Gland Segmentation Challenge with a rank sum of 17 out of 13 competitive teams. As the team stated, 'Our method won the 2015 MICCAI Gland Segmentation Challenge out of 13 competitive teams, surpassing all the other methods by a significant margin.'
  • Achieved F1 score of 0.912 on Test Part A (benign-dominated) and 0.716 on Test Part B (malignant-dominated), demonstrating robust performance across tissue types.
  • Object-level Dice coefficient reached 0.897 on Test Part A and 0.781 on Test Part B, with object-level Hausdorff distance of 45.418 and 160.347 respectively, indicating high boundary accuracy.
  • + 4 more results inside

Our method won the 2015 MICCAI Gland Segmentation Challenge out of 13 competitive teams, surpassing all the other methods by a significant margin.

Hao Chen, Xiaojuan Qi, Lequan Yu, Pheng-Ann Heng, Research Team at Department of Computer Science and Engineering, The Chinese University of Hong Kong

Sign in to unlock the full case study

Get the complete analysis, all results, expert quotes, and entity profiles — free.