Hypergraph Learning: Methods, Tools and Applications in Medical Image Analysis

MICCAI 2019 Tutorial


Sunday, 16 September 2018

Time: 15:00 - 19:00

Location information: Room Picasso.


  • 15:00-15:15 Introduction & Motivation
  • 15:15-16:30 Techniques for Interpreting Deep Models
  • 16:30-17:00 Coffee Break
  • 17:00-18:00 Applications of Interpretability
  • 18:00-19:00 Case Study: Interpretable Machine Learning in Histopathology




In recent years, the MICCAI community has witnessed a booming increase of machine learning technique for medical image analysis. Among these, how to build data correlation and conduct classification is one important task. Hypergraph, as an effective data modelling and learning method, has shown its effectiveness in many tasks, such as Alzheimer’s Disease diagnosis, brain structure segmentation, medical image retrieval and brain network modelling. In this tutorial, we will introduce hypergraph learning methods, applications and tools in medical image analysis field. The topics covered in this tutorial include hypergraph preliminaries and hypergraph generation methods, recent progress in hypergraph learning, the applications of hypergraph learning in medical image segmentation, retrieval and diagnosis, and the toolbox of hypergraph learning which can be used in medical image analysis.

Outline of the tutorial

  1. Preliminaries of hypergraph and hypergraph generation methods
    • the basic knowledge about hypergraph
    • the comparison between hypergraph and graph
    • hypergraph generation methods
    • typical hypergraph modelling methods
  2. Recent hypergraph learning methods
    • traditional transductive learning on hypergraph
    • traditional inductive learning on hypergraph
    • the dynamic hypergraph structure learning method
    • recent hypergraph neural network methods
  3. Applications in medical image analysis
    • mild cognitive impairment diagnosis
    • medical image retrieval
    • brain structure segmentation
  4. The toolbox for hypergraph learning

About Organizers

  • Yue Gao is currently a Tenured Associate Professor in School of Software, Tsinghua University, Beijing, China. He received his B.E. degree from Harbin Institute of Technology, Master Degree and Ph.D. degree from Tsinghua University, respectively. He has been working in School of Computing, National University of Singapore and School of Medicine, University of North Carolina at Chapel Hill from 2012 to 2016. His research falls in the field of medical image analysis and computer vision. His recent research work contains 3D vision, machine learning and computer-aided diagnosis. He has published 100+ papers in premier journals and conferences like TIP, Human Brain Mapping, MICCAI, CVPR, IJCAI, AAAI, ECCV and ACM Multimedia. He also serves as editorial board member for IEEE Transactions on Signal and Information Processing over Networks and Journal of Visual Communication and Image Representation. He has been served as associate editor for Neurocomputing.
  • Rongrong Ji is currently a Minjiang Chair Professor at the Department of Cognitive Science, School of Information Science and Engineering, Xiamen University. Ji’s research falls in the field of computer vision, multimedia, and machine learning. He has published 100+ papers in tier-1 journal and conferences like IJCV, TIP, CVPR, ICCV, ECCV, IJCAI, AAAI, ACM Multimedia etc, with over 4500 citations in the past 5 years. He is associate editor of Neurocomputing, Multimedia Tools and Applications, The Visual Computer, PLOS ONE, Frontiers of Computer Science etc., guest editor of ACM Transactions on Intelligent Systems and Technology, IEEE Multimedia Magazine, Signal Processing, Neurocomputing etc.
  • Zizhao Zhang is a Ph.D candidate in Tsinghua University, respectively. She is working on hypergraph learning and has developed the tools for hypergraph learning. She has published several papers on top tier conferences and journals, such AAAI, IJCAI, CVPR and TIP, on the topic of hypergraph learning. She will demonstrate the usage of these tools during the tutorial.