Paper Title :Nonlinear Least Squares Regression Analysis And Diagnosis For Optical And Laser Measurement Systems
Author :Zhengmao Ye, Habib Mohamadian
Article Citation :Zhengmao Ye ,Habib Mohamadian ,
(2015 ) " Nonlinear Least Squares Regression Analysis And Diagnosis For Optical And Laser Measurement Systems " ,
International Journal of Electrical, Electronics and Data Communication (IJEEDC) ,
pp. 1-6,
Volume-3, Issue-7
Abstract : Laser Doppler and Raman spectroscopy are widely adopted for biomedical diagnosis. Both Doppler ultrasound
and Raman spectra are used to display normal and abnormal signature waveforms that are unique in collected samples, in
order to recognize normal and abnormal regions in the spectral display respectively. Visualization of quantitative Doppler
ultrasound data and Raman spectra provides a straightforward means to distinguish between normal and abnormal samples,
however, mixing artifacts due to motion, instrumentation and background are in fact inevitable. The fluorescence artifact is
the dominating factor whose formation mechanism varies. In general, fluorescence waveforms could be formulated as
typical Gaussian, Sigmoidal or Lorentzian distribution, while single spectrum is seldom enough to represent the fluorescence
waveforms completely to further extract intrinsic spectra with unique signatures. In this case, combination of Gaussian,
Sigmoidal and Lorentzian waveforms could be the solution to model the fluorescence artifact. To identify a set of modeling
parameters, a simple but powerful nonlinear least squares regression technique is introduced in this article. Case studies on
both normal and abnormal sample data are conducted.
Index Terms—Nonlinear Least Squares Regression, Raman Spectroscopy, Doppler Ultrasound
Type : Research paper
Published : Volume-3, Issue-7
DOI - 10.18479/ijeedc/2015/v3i7/48250
Copyright: © Institute of Research and Journals
|
 |
| |
 |
PDF |
| |
Viewed - 128 |
| |
Published on 2015-07-04 |
|