Study On Terahertz Time-Domain Spectroscopy Feature Extraction And Lesion Recognition Methods For Serum Sialic Acid

Terahertz Time Domain Spectroscopy(THz-TDS)is sensitive and non-destructive to the response of biomolecules,which has a broad application prospect in the field of biomedical research,especially in the field of tumor detection.Relying on the National Natural Science Foundation of China,the study carry out research on the detection of abnormal sialic acid in human serum based on terahertz time-domain spectroscopy technology.The detection and analysis methods of terahertz time-domain spectroscopy were studied from the tumor marker sialic acid,sialic protein and blood cells.Especially,the research focused on techniques such as spectral response mechanism analysis,feature extraction and abnormal discrimination of tumor lesion marker,which provide support to the medically assisted diagnosis.The main contributions and innovations of this dissertation are as follows:(1)The experimental study the responses of the main THz tumor marker for sialic acid.To better explain the detection of the abnormality information and the sialic acid content of human serum tumor markers using terahertz time domain spectroscopy,this study carried out terahertz wave absorption characteristics of sialic acid crystals,simulation methods of single-molecule and unit cell configurations,and the response mechanism and forward analysis of sialic acid aqueous solutions.This chapter also completed the comparison and experimental verification of the sialic acid single molecule and the unit cell terahertz response.At the same time,the hydration of aqueous sialic acid was explored,and the effect of sialic acid concentration on the absorption spectrum of terahertz waves was analyzed.Synthesizing the above results,this section proposes a terahertz time-domain spectral response and verification method for tumor markers,which provides a reference for the use of terahertz time-domain spectroscopy to detect tumor markers in the biomedical field.(2)Using terahertz time-domain spectroscopy to identify sialic glycoproteins.If a protein is sialylated(sialic acid becomes a protein group),the protein can be used as a marker for tumor lesions.This section revolves around the issue of specific identification of sialylation of proteins using terahertz time-domain spectroscopy.The glycoprotein has no significant absorption peak by the influence of the concentration.We propose a method for extracting terahertz time-domain spectral features of sialic glycoprotein based on composite multi-scale entropy.Which has a good advantage in nonlinear spectral analysis.At the same time,the sensitivity of terahertz parameters such as absorption coefficient,refractive index and permittivity to sialylation of glycoproteins was studied,and the influence of terahertz parameters on identification was explored.(3)Using terahertz time domain spectroscopy and pattern recognition algorithms to recognize abnormal serum sialic acid and neoplastic lesions.Abnormal levels of free sialic acid and sialic glycoprotein in serum are a hallmark of human tumors.The chapter focuses on several issues related to the detection of abnormal serum sialic acid and the identification of tumor lesions based on terahertz time-domain spectroscopy.Aiming at the sensitivity selection of terahertz parameters in human serum recognition,the method of mean intersection over union was proposed to quantitatively evaluate and select sensitivity parameters;Aiming at the problem of feature selection,a fusion method based on frequency features,principal component features and sequence information entropy features was proposed.We have established a disease serum discrimination model,which improves the sensitivity and specificity of disease recognition.The above method laid the foundation for the future detection of serum sialic acid abnormalities.All in all,the article focuses on several key issues in serum sialic acid detection and tumor lesion identification methods.Aiming at terahertz response comparison and experimental verification of sialic acid,the spectral response and verification method for tumor markers;Focusing on the problem of specifically identifying sialic glycoprotein,a method for extracting terahertz time-domain spectral features of sialic glycoprotein based on composite multi-scale entropy was proposed;Aiming at the sensitivity selection of terahertz parameters,the method of mean intersection over union was proposed to quantitatively evaluate and select sensitivity parameters;Aiming at the problem of feature selection,a fusion method based on frequency features,principal component features and sequence information entropy features was proposed.The related progress and results provide a reference for terahertz time-domain spectroscopy in tumor marker recognition and lesion detection.The pathological analysis supported by this project were provided by the Cancer Research Institute of the Second Affiliated Hospital of Zhejiang University Medical College.