Research On Biological Tissue Spectral Characteristic Based On Wavelet Denoise

Optical-biology technology is an important research direction in life science. Biological tissue as a part of organism carries a lot of information of life. The combination of spectroscopy research on human tissue and medical detection, cure shows the tremendous value and application prospects.Combining the research background and current situation in and abroad, the thesis selected human blood as typical biological tissues. In this thesis we mainly research on the absorption spectrum and fluorescence spectrum of blood. Apply the wavelet analysis method to denoising and analysis of the blood spectral signal. We study and explore the relationship of blood spectral characteristics and the biochemical index. The content of this thesis mainly concentrated in the following areas:(1) We design the wavelet treatment of blood spectral signal. In this method we decompose the blood spectral signal, further decompose in layers, then using threshold to get reconstructed signal.(2) Apply the wavelet method on the decomposing and reconstruction of absorption spectrum of human serum and whole blood. Reconstructed spectral signals obviously reduce signal disturbance.(3) By comparing the spectral signal of normal serum and high-fat serum we study the differences between these two. We can determine whether content of fat in serum is abnormal through detection of absorption spectrum of this spectrum segment. This provides a new method for medical detection.(4) Apply the wavelet methods on decomposition and reconstruction of human serum fluorescence spectrum. Study the differences of fluorescence spectrum between normal serum, high-fat serum, and high-glucose serum..(5) In addition of the wavelet method, this thesis also proposes a method with simple principle to denoise, and finish the corresponded program. Using absorption spectrum signal as an example, the results of the new method and previous method are basically the same.The research shows that applying the wavelet analysis method to the study of characteristics of blood spectrum can effectively filter the disturbance of spectral signal, and show a lot of new information. The decomposed signals also contain a lot of features information. Study the relationship between the feature of reconstruction signal or decomposition signals and the correspondent biochemical index can provide the new detecting and management for medical diagnosis.