Alumni Association
Professor Daniel Rueckert (Imperial College London)
05 April 2013
JSPS Joint Research
Project
Project Title:
Development of novel image processing method based on large-scale medical image
database
Japanese Lead Scientist:
Professor Kensaku Mori, Information and
Communications Headquarters, Nagoya University
UK Counterpart: Professor
Daniel Rueckert, Department of Computing, Imperial College London
Project Duration: April
2011 to March 2013
The purpose of this
international collaboration project is to develop image analysis system for
processing chest and abdominal CT images. The goal is to implement a system that
can recognize the locations of organs, blood vessels and lymph nodes in 3D CT
images. The collaboration combines the expertise of Professsor Mori’s lab in
developing CAD systems with the expertise of Professor Rueckert’s group in
image registration and multi-atlas segmentation techniques. The collaboration started
as a result of several meetings at the MICCAI conferences in 2009 in London,
UK, and 2010 in Beijing, China. Throughout the bilateral research project for two
years from April 2011, the Japanese research team has been awarded funding
totaling about five million yen by JSPS. The UK research team has been awarded
additional funding by the Royal Society as part of an international joint
project. The JSPS funding has been used for several visits of PhD students,
Post-Docs and the PI in London, UK. During these visits expertise, know-how and
software has been exchanged. In addition, the Japanese team has given several
lectures and courses at Imperial College London. The collaboration has led to
several joint publications, including a conference paper at MICCAI 2012 which
has won the best paper award [1]. In particular, the collaboration has led to
very successful patch-based segmentation algorithms that allow the dynamic
construction of multi-organ probabilistic atlases that are customized to
individual subjects.
[1] R. Wolz, C. Chu, K.
Misawa, K. Mori and D. Rueckert. Multi-organ Abdominal CT Segmentation using
hierarchically weighted Subject-specific Atlases. MICCAI 2012.
