一种基于嵌套多元化学计量学的校准策略，用于利用能量色散X射线荧光光谱法（EDXRF）和散射光谱法对软组织中的痕量生物金属进行直接分析。
A nested multivariate chemometrics based calibration strategy for direct trace biometal analysis in soft tissue utilizing Energy Dispersive X-Ray Fluorescence (EDXRF) and scattering spectrometry.
作者信息
Okonda J J, Angeyo K H, Mangala J M, Kisia S M
机构信息
Department of Physics, University of Nairobi, P.O Box 30197-00100, Nairobi, Kenya.
Department of Physics, University of Nairobi, P.O Box 30197-00100, Nairobi, Kenya.
出版信息
Appl Radiat Isot. 2017 Nov;129:49-56. doi: 10.1016/j.apradiso.2017.08.008. Epub 2017 Aug 8.
Compton scatter-modulated fluorescence and multivariate chemometric (artificial neural network (ANN) and principal component regression (PCR)) calibration strategy was explored for direct rapid trace biometals (Mn, Fe, Cu, Zn, Se) analysis in "complex" matrices (model soft tissues). This involved spectral feature selection (multiple fluorescence signatures) normalized to or in conjunction with Compton scatter. ANN model resulted in more accurate trace biometal determination (R>0.9) compared to PCR. Hybrid nested (ANN and PCR) approach led to optimized accurate biometals' concentrations in Oyster tissue (≤ ± 10%).
探索了康普顿散射调制荧光和多变量化学计量学（人工神经网络（ANN）和主成分回归（PCR））校准策略，用于直接快速分析“复杂”基质（模型软组织）中的痕量生物金属（锰、铁、铜、锌、硒）。这涉及到与康普顿散射归一化或结合的光谱特征选择（多个荧光特征）。与PCR相比，ANN模型在痕量生物金属测定方面更准确（R>0.9）。混合嵌套（ANN和PCR）方法使牡蛎组织中生物金属浓度的优化准确度达到≤±10%。
Zotero 插件