振动光谱技术结合信息融合思想提高宫颈癌早期诊断准确率的方法研究

目前我国宫颈癌负担仍然很重，通过筛查尽早发现癌前病变及早期癌，可有效降低癌症相关死亡率。现有的筛查方法均存在一定弊端，找到一种快速且高准确率的替代方法具有重要现实意义和经济价值。本项目在前期工作的基础上：1）进一步收集大量的宫颈脱落细胞临床标本，分别获取对应的拉曼光谱和近红外光谱数据；2）提出了一种拉曼与近红外光谱融合的方法，以提高宫颈癌的早期诊断准确率；3）针对生物样品的拉曼光谱强荧光背景干扰问题，提出了将拟合的荧光背景信号与拉曼光谱相融合的方法，避免了将荧光背景信号简单的扣除而造成对诊断有用信息丢失的问题。本项目的研究方法，可应用于人体其他疾病的筛查诊断，具有一定的普适性。

At present, the burden of cervical cancer in China is still very heavy. Early detection of precancerous lesions and early cancer through screening can effectively reduce cancer-related mortality. There are some disadvantages in the existing screening methods. It is of great practical significance and economic value to find a fast and high-accuracy alternative method. Based on the previous work, this project: 1) further collect a large number of clinical samples of cervical exfoliated cell specimens, and obtain the corresponding Raman spectra and near-infrared spectra respectively; 2) a method of combining Raman and near-infrared spectroscopy was proposed to improve the accuracy of early diagnosis of cervical cancer; 3) aiming at the problem of strong fluorescence background interference in the Raman spectroscopy of biological samples, a method of fusing the fitted fluorescence background signal with Raman spectroscopy was proposed to avoid the loss of useful diagnostic information caused by simple deduction of the fluorescence background signal. The research method of this project can be applied to the screening and diagnosis of other human diseases and has certain universality.