DocumentCode
2142093
Title
Study on Hyperlipidemia Serum Ultraviolet Visible Absorption Spectra Based on Wavelet Transform
Author
Weihua Zhu ; Zhao, Zhimin ; Xin Guo ; Hong, Xiaoqin
Author_Institution
Coll. of Sci., Nanjing Univ. of Aeronaut. & Astronaut., Nanjing, China
fYear
2009
fDate
17-19 Oct. 2009
Firstpage
1
Lastpage
5
Abstract
The wavelet transform technology was used for the analysis of serum ultraviolet-visible (UV) absorption spectra. The experimental results show that the absorption spectra are of serious noise in the wave band from 200 nm to 300 nm, which covers some useful spectral characteristics. Serum absorption spectra were broken down into four layers by db4 wavelet. The shapes of the low-frequency signals are similar to the original one. The high-frequency signals include more noise and minutiae features. The effective separation of the noise was achieved by using different threshold for the decomposition signals of different layers. The signals of reconstruction have been obviously improved in the noise serious wave band from 200 nm to 300 nm, resulting in some useful information. There is one more peak in the absorption spectrum of hyperlipidemia serum near 289 nm when compared with the absorption spectrum of the normal serum. Therefore, the spectral characteristics of the serum between 260 nm and 300 nm can be used to estimate whether the blood lipid level is normal or abnormal, providing a new method for analyzing the blood of disease diagnosis.
Keywords
blood; diseases; medical diagnostic computing; medical signal processing; patient diagnosis; spectral analysis; ultraviolet spectra; wavelet transforms; blood lipid level; decomposition signals; disease diagnosis; hyperlipidemia serum ultraviolet visible absorption spectra; serum absorption spectra; spectral characteristics; wavelet transform; Blood; Diseases; Electromagnetic wave absorption; Lipidomics; Low-frequency noise; Mirrors; Noise shaping; Optical filters; Wavelet analysis; Wavelet transforms;
fLanguage
English
Publisher
ieee
Conference_Titel
Image and Signal Processing, 2009. CISP '09. 2nd International Congress on
Conference_Location
Tianjin
Print_ISBN
978-1-4244-4129-7
Electronic_ISBN
978-1-4244-4131-0
Type
conf
DOI
10.1109/CISP.2009.5303600
Filename
5303600
Link To Document