近红外光谱技术应用于癌症检测的研究进展
Advances of Near-Infrared Spectroscopy in Cancer Diagnosis
朗韦玮 1李嘉晨 2董林毅 1韩丽1
作者信息
- 1. 天津医科大学药学院,天津 300070
- 2. 天津医科大学药学院,天津 300070;天津市第五中心医院生态城医院,天津 300467
- 折叠
摘要
近红外(Near-infrared,NIR)光谱主要反映了含氢基团的振动信息,可用于获取由疾病引起的生物样品中蛋白质、脂质等成分的含量和结构变化信息.近年来,近红外光谱技术由于其快速、无损和非侵入性等优势,在疾病诊断领域发挥了重要作用.在癌症检测方面,恶性肿瘤细胞与正常细胞的化学结构和成分存在明显差异,并且癌症引起的异常代谢功能会导致生物组织和体液成分发生变化,从而影响光谱吸收,这使得近红外光谱有望成为探测生物体内癌变信息的有效工具.然而,近红外光谱的谱峰重叠严重,因此通常需要借助化学计量学方法对光谱进行解析,以有效提取癌症诊断信息.本文总结了近年来近红外光谱技术应用于癌症检测的研究进展,包括使用光谱预处理方法用于获取高分辨率光谱信号和建立模型方法用于癌症诊断,以及近红外高光谱成像结合深度学习方法的医学诊疗应用趋势.通过对光谱进行解析,可以探测生物体系中分子的定量信息和结构变化,特别是利用近红外光谱对水的敏感性,分析了水在癌变过程中的含量和结构变化及其在癌症检测中的作用机理.
Abstract
Near-infrared(NIR)spectroscopy reflects the vibrational information of hydrogen-containing groups,allowing the detection of changes in components such as proteins and lipids in biological samples caused by diseases.Due to the advantages such as rapidity,non-destructiveness and non-invasiveness,NIR spectroscopy has been widely used in disease diagnosis.Malignant tumor cells exhibit significant differences in structure and composition compared to normal cells,and abnormal metabolic functions cause changes in the composition of biological tissues and fluids,leading to the variation in spectra.This makes NIR spectroscopy a promising tool for detecting cancer.However,severe overlapping of spectral peaks in the spectra requires the help of chemometrics methods to analyze the spectra and extract diagnostic information.This review summarizes the recent advances in NIR spectroscopy technology in cancer detection,including the methods of spectral preprocessing,diagnosis modeling and the development in NIR hyperspectral imaging combined with deep learning methods.The quantitative information and structural changes of molecules in biological systems can be extracted.Furthermore,the sensitivity of NIR spectra to water is utilized to analyze the content and structural changes of water induced by cancer and its role in cancer detection.
关键词
近红外光谱/癌症诊断/化学计量学/定量分析/结构分析/评述Key words
Near-infrared spectra/Cancer diagnosis/Chemometrics/Quantitative analysis/Structural analysis/Review引用本文复制引用
基金项目
天津市教委科研计划项目(2022ZD053)
出版年
2024
国家科技期刊平台
NSTL
万方数据