| We are pleased to invite you to the following seminar talk jointly organized by Centre for Information Mining and Extraction (CHIME), School of Computing, National University of Singapore and Pattern Recognition and Machine Intelligence Association (PREMIA): |
| Energy Minimization Methods for Segmentation and Registration in Cardiac MRI Sequences |
Prof José M. F. Moura
Departments of Electrical and Computer Engineering and BioMedical Engineering
NMR Center for Biological Research
Carnegie Mellon University
Pittsburgh PA, USA
|
| |
| Date: Tuesday, 8 March 2005 |
| Time: 1.30 pm - 2.30 pm |
| Venue: LT31, Block S16 Level 3
(For those traveling by car, please park at Car Park 10 which is opposite Block S16 across Lower Kent Ridge Road, car park payment by cash card. Please see campus map at http://www.nus.edu.sg/campusmap/ )
|
| |
| Admission: Admission is free |
Registration: Please register online via http://www.premia-sg.org/index.php?option=com_events&event_id=7&task=register
(Registration is closed now). |
Synopsis:
We present energy minimization methods for automatic image segmentation and registration in cardiac MR image sequences that overcome low contrast, the difficulties with segmenting the papillary muscles, and turbulent blood flow. The energy functional combines stochastic region-based and edge-based information with shape priors of the heart and local properties of the contour. The minimization algorithm–the stochastic adaptive contour segmentation (STACS)–solves, by the level set method, the Euler-Lagrange equation that describes the contour evolution. STACS includes an annealing schedule that balances dynamically the weight of the different terms in the energy functional. Particularly attractive features of STACS are: (1) ability to segment images with low texture contrast by modeling stochastically the image textures; (2) robustness to initial contour and noise because of the utilization of both edge and region-based information; and (3) ability to segment the heart from the chest wall and the undesired papillary muscles due to inclusion of heart shape priors. Application of STACS to a set of 48 real cardiac MR images shows that it successfully segments the heart from its surroundings such as the chest wall and the heart structures (the left and right ventricles and the epicardium.) We compare STACS' automatically generated contours with manually-traced contours, or the ``gold standard," using both area and edge similarity measures. This assessment demonstrates very good and consistent segmentation performance of STACS. |
Biography:
José M. F. Moura is a Professor of Electrical and Computer Engineering and BioMedical Engineering at Carnegie Mellon University. He got his MSc, EE, and D.Sc. in EECS from MIT, and his licenciatura in Electrical Engineering from Instituto Superior Técnico (IST) (Lisbon, Portugal). He was on the faculty of IST (1975-1984) and a visiting professor at MIT (1984-86 and 1999-00). His research interests are in statistical signal and image processing and statistical communications. He has over 300 publications in Journals and Conferences.
He is the President Elect for the period 2006-2007 of the IEEE Signal Processing Society (SPS) and serves on the Editorial Boards of the IEEE Proceedings and the ACM Transactions on Sensor Networks. He has served IEEE and SPS in several capacities including: Editor in Chief (EiC) for the Transactions on Signal Processing and Interim EiC for the IEEE Signal Processing Letters; Chair of the IEEE Transactions Committee that joins all EiCs of the over 80 IEEE Journals; Vice-President Publications of SPS; Vice-President Publications of the IEEE Sensors Council. He is a corresponding member of the National Academy of Sciences of Portugal and a Fellow of the IEEE. He received in 2000 the IEEE Millenium Medal and in 2004 the SPS Meritorious Service Award.
|
