Recovery of Dynamic PET Regions via Simultaneous Segmentation and Deconvolution
Abstract
We derive a new model for simultaneous segmentation and deconvolution of dynamic PET images. By incorporating the PSF of the imaging system into our segmentation model, we simultaneously estimate region boundaries, and correct tissue activities for the partial volume effect. We show improved segmentation results, and outperform two state-of-the-art dynamic PET segmentation methods.
Top- Smith, Ben
- Saad, Ahmed
- Hamarneh, Ghassan
- Möller, Torsten
Supplemental Material
Shortfacts
Category |
Paper in Conference Proceedings or in Workshop Proceedings (Other) |
Event Title |
MICCAI 2008 Workshop on Analysis of Functional Medical Images |
Divisions |
Visualization and Data Analysis |
Subjects |
Computergraphik |
Event Location |
New York, USA |
Event Type |
Workshop |
Event Dates |
10 Sep 2008 |
Date |
September 2008 |
Export |