Robust and conditional deep learning based medical image registration

Speaker: Prof. Albert CHUNG
         Department of Computer Science
         University of Exeter

Title:  "Robust and conditional deep learning based medical image
         registration"

Date:   Monday, 6 February 2023

Time:   3:00pm - 4:00pm HKT

Zoom link:
https://hkust.zoom.us/j/465698645?pwd=aVRaNWs2RHNFcXpnWGlkR05wTTk3UT09

Meeting ID: 465 698 645
Passcode: 20222023


Abstract:

Medical imaging, such as magnetic resonance imaging and computed
tomography, produces high-quality, high-resolution images that reveal the
anatomical and functional structures of the human body. For example,
neuroimages can delineate details of brain tissues and structures and
contribute to the diagnosis of brain diseases and the detection of
biomarkers for early treatment. To analyze images, image registration
needs to be performed to highlight anatomical changes over time, transfer
segmentation labels for interpretation, or aid image-guided surgery. The
registration of images with large anatomical deformation, e.g., tumors, is
a challenging task due to the implicit assumption of small deformation
settings in conventional methods. In this talk, I will give an overview of
our recently developed image registration methods, which rely on the deep
Laplacian pyramid networks, a controllable regularization module, and a
multi-resolution alignment strategy. The new methods can mitigate the
problems arising from images with large anatomical deformation and offer
accurate image registration results. At the end of the talk, I will
introduce some possible research directions for medical image analysis.


Biography:

Albert Chung is currently a Chair in AI at the Department of Computer
Science, University of Exeter, and Professor Emeritus and Adjunct Professor
at the Department of Computer Science and Engineering, HKUST. He received
his BEng (First-Class Honors) in Computer Engineering from the University
of Hong Kong in 1995. He joined HKUST in 1996 and obtained his MPhil in
Computer Science in 1998. He was awarded the Croucher Foundation
scholarship in 1998 to read DPhil in Engineering Science at the Medical
Vision Laboratory, University of Oxford. He graduated in 2001 with the
BMVA Best Thesis Award in the UK. He was a visiting scientist at the AI
Laboratory, Massachusetts Institute of Technology, in 2001-02 and had been
with HKUST since then until 2021. He served as the CPEG Director
(2010-12), Associate Head of CSE (2016-19) and Associate Dean of
Engineering at SENG HKUST (Jan 2020 - Dec 2021).

His research interests include medical image analysis, image processing
and computer vision. His research focus is to create and apply
computational techniques in medical image analysis, and the ultimate goal
is to improve the quality of healthcare services. He received two Best
Paper Awards in MIAR'16 and DLMIA'15. He is active in his research
community. He co-chaired and hosted the international conference on
Information Processing in Medical Imaging (IPMI) in 2019 at HKUST and
served as the Chair of the IPMI Conference Board in 2019-21. He was the
Tutorial Chair for the IEEE International Symposium on Biomedical Imaging
in 2019. He is serving as one of the Editorial Board members for the
Medical Image Analysis journal and the Journal of Computerized Medical
Imaging and Graphics.