摘要
衰减全反射傅里叶变换红外光谱技术(ATR-FTIR)可实现微量样本的快速可靠的在体红外光谱检测。本研究采用自行搭建的空芯光纤(HOF)ATR-FTIR探针技术,结合主成分分析(PCA)及Fisher判别(FDA)算法,对不同病变阶段(健康,病变8周、3个月、7个月)比格犬的骨关节炎(OA)离体样本进行原位鉴别分析。对初始样本识别正确率为100%;各组分别选取一独立样本作为预测组,预测组样本识别正确率为100%;所有交叉验证的识别率均超过95%。本方法能客观地准确鉴别不同病变阶段的OA样本,可作为不同OA病变阶段的诊断参考。结合主观的OA评分结果,HOF-ATR-FTIR技术在OA的在体原位临床诊断中具有重要的应用价值,有望为OA提供更加客观精确的光谱分级和分期结果。
In situ infrared spectral detection can be fast and reliably achieved by attenuated total reflection Fourier transform infrared(ATR-FTIR) spectroscopy technique. In this study, home-made hollow optical fiber(HOF-) ATR-FTIR technique was combined with principal component analysis(PCA) and Fisher discriminant(FDA) algorithm to identify in situ osteoarthritis(OA) canine samples at different pathological stages(health, lesions of 8 weeks, 3 months and 7 months) in vitro. The initial samples were correctly recognized with 100% accuracy. And all of the prediction groups were 100% correctly identified when an independent sample was selected from each group for prediction, respectively. In cross-validation, the recognition rate was more than 95%. This method can objectively, quantitatively and accurately identify OA samples at different stages of OA, and can be used as a diagnostic basis at different OA stages. The HOF-ATR-FTIR technique has significant application prospect in in vivo and in situ clinical diagnosis when combined with subjective OA grading results, and is expected to provide more objective and accurate OA grading and staging.
作者
赵远
朱勇康
陆燕飞
尚林伟
符娟娟
马丹英
王潇
尹建华
ZHAO Yuan;ZHU Yong-Kang;LU Yan-Fei;SHANG Lin-Wei;FU Juan-Juan;MA Dan-Ying;WANG Xiao;YIN Jian-Hua(Department of Biomedical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《分析化学》
SCIE
EI
CAS
CSCD
北大核心
2019年第12期1981-1986,共6页
Chinese Journal of Analytical Chemistry
基金
国家自然科学基金项目(No.61378087)
江苏省研究生科研与实践创新计划项目(No.KYCX18_0321)资助~~
关键词
骨关节炎
空芯光纤-衰减全反射傅里叶变换红外光谱
主成分分析
FISHER判别
Osteoarthritis
Hollow optical fiber attenuated total reflection Fourier transform infrared spectroscopy
Principal component analysis
Fisher discriminant
