Early Rapid diagnosis of Alzheimer’s disease based on fusion of near- and mid-infrared spectral features combined with PLS-DA
Early Rapid diagnosis of Alzheimer’s disease based on fusion of near- and mid-infrared spectral features combined with PLS-DA
复制标题
DOI：
10.1016/j.ijleo.2021.166485
复制
发表时间：
2021-03
期刊：
影响因子：
3.1
通讯作者：
Chuanmei Yang;Peiwen Guang;Li Li-Li;Han Song;Furong Huang;Yuanpeng Li;Lihu Wang;Junhui Hu
中科院分区：
文献类型：
--
作者：
Chuanmei Yang;Peiwen Guang;Li Li-Li;Han Song;Furong Huang;Yuanpeng Li;Lihu Wang;Junhui Hu
为探讨光谱特征融合技术诊断阿尔茨海默病（AD）的可行性，采集了265份健康人和阿尔茨海默病患者血清的近红外（NIR）和中红外（MIR）光谱。使用偏最小二乘判别分析（PLS-DA）的PC的最佳数量来构建AD的诊断模型。基于NIR光谱的PLS-DA模型的准确性、灵敏度和特异性分别为80.6%、93.3%和93.7%，而基于MIR光谱的PLS-DA模型的对应物分别为98.5%、97.5%和100%。对于基于近红外-中红外光谱特征融合的PLS-DA模型，对应率分别为100%、97.5%和100%。综上所述，将联合收割机光谱特征融合与偏最小二乘-判别分析相结合，建立AD的诊断模型，对AD的准确、快速诊断是有效可行的。
To investigate the feasibility of using spectral feature fusion technology to diagnose Alzheimer’s disease (AD), near-infrared (NIR) and mid-infrared (MIR) spectra were collected on 265 serum samples from healthy individuals and Alzheimer’s disease patients. The optimal number of PCs is used to construct diagnostic models for AD using partial least-squares discriminant analysis (PLS-DA). The accuracy, sensitivity, and specificity of the NIR spectra-based PLS-DA model are 80.6 %, 93.3 %, and 93.7 %, respectively, whereas the counterparts of the MIR spectra-based PLS-DA model are 98.5 %, 97.5 %, and 100 %, respectively. For the PLS-DA model based on NIR-MIR spectral feature fusion, the counterparts are 100 %, 97.5 %, and 100 %, respectively. In summary, it may be effective and feasible to combine spectral feature fusion and PLS-DA to establish a diagnostic model for an accurate and rapid diagnosis of AD.