Robust segmentation and accurate target definition for positron emission tomography images using affinity propagation
ISBI 2013 – 2013 IEEE 10th International Symposium on Biomedical Imaging: From Nano to Macro, pp. 1461–1464
Abstract
An abstract is not available in the lab archive. View the paper at its source.
This 2013 conference paper proposes an intensity-based method for selecting a threshold in PET images, using a similarity metric and Affinity Propagation clustering. The authors tested the method on PET images of rabbits infected with tuberculosis.
Summary
The paper addresses threshold selection for segmenting positron emission tomography (PET) images with distributed uptake. The authors propose calculating a similarity metric between gray-level histogram data points, then applying Affinity Propagation (AP), a clustering method, to group image intensities and select a threshold. They tested the approach on PET images from rabbits infected with tuberculosis. The abstract describes the results as “promising”; it does not establish clinical validation or readiness for human use.
- Brent Foster, Ulas Bagci, Brian Luna, Bappaditya Dey, William Bishai, Sanjay Jain, Ziyue Xu, Daniel J. Mollura
- 2013
- ISBI 2013 – 2013 IEEE 10th International Symposium on Biomedical Imaging: From Nano to Macro, pp. 1461–1464
- DOI: 10.1109/ISBI.2013.6556810
Cite this work
Brent Foster; Ulas Bagci; Brian Luna; Bappaditya Dey; William Bishai; Sanjay Jain; Ziyue Xu; Daniel J. Mollura (2013). Robust segmentation and accurate target definition for positron emission tomography images using affinity propagation. ISBI 2013 – 2013 IEEE 10th International Symposium on Biomedical Imaging: From Nano to Macro, pp. 1461–1464 https://doi.org/10.1109/ISBI.2013.6556810
BibTeX