Journal article
Detection of shielded radionuclides from weak and poorly resolved spectra using group positive RIVAL
Radiation measurements, Vol.48(1), pp.18-28
01/2013
DOI: 10.1016/j.radmeas.2012.11.002
Abstract
This paper is concerned with the identification of nuclides from weak and poorly resolved spectra in the presence of unknown radiation shielding materials such as carbon, water, concrete and lead. Since a shield will attenuate lower energies more so than higher ones, isotope sub-spectra must be introduced into models and into detection algorithms. We propose a new algorithm for detection, called group positive RIVAL, that encourages the selection of groups of sub-spectra rather than the selection of individual sub-spectra that may be from the same parent isotope. Indeed, the proposed algorithm incorporates group positive LASSO, and, as such, we supply the consistency results of group positive LASSO and adaptive group positive LASSO. In an example employing various shielding materials and material thicknesses, group positive RIVAL is shown to perform well in all scenarios with the exception of ones in which the shielding material is lead. ► Identification of nuclides from weak and poorly resolved spectra. ► Shielding materials such as carbon, water, concrete, and lead are considered. ► Isotope spectra are decomposed into their sub-spectra. ► A variable selection algorithm is proposed that encourages group selection. ► Simulations demonstrate the proposed method's performance when nuclides have been shielded.
Details
- Title: Subtitle
- Detection of shielded radionuclides from weak and poorly resolved spectra using group positive RIVAL
- Creators
- Paul Kump - Dept. of Electrical and Computer Engineering, University of Iowa, 4316SC, Iowa City, IA 52242, USAEr-Wei Bai - Dept. of Electrical and Computer Engineering, University of Iowa, 4316SC, Iowa City, IA 52242, USAKung-Sik Chan - Dept. of Statistical and Actuarial Science, University of Iowa, Iowa City, IA 52242, USAWilliam Eichinger - Dept. of Civil and Environmental Engineering, University of Iowa, Iowa City, IA 52242, USA
- Resource Type
- Journal article
- Publication Details
- Radiation measurements, Vol.48(1), pp.18-28
- Publisher
- Elsevier Ltd
- DOI
- 10.1016/j.radmeas.2012.11.002
- ISSN
- 1350-4487
- eISSN
- 1879-0925
- Language
- English
- Date published
- 01/2013
- Academic Unit
- Statistics and Actuarial Science; Civil and Environmental Engineering; Electrical and Computer Engineering; Radiology
- Record Identifier
- 9983985940402771
Metrics
7 Record Views