Venue
Speakers
Abstract
Use of Wavelet Transform has become popular in recent years in a variety of signal processing applications such as noise reduction, data compression, time-frequency analysis and image analysis. Prof. Ramanathan will present representative applications of wavelet transform to process spin noise and NMR spectra and illustrate the advantages that result from the adoption of the method. Spin noise is naturally present in samples and can be recorded without using r.f. excitation [1]. Several conventional signal enhancement techniques used in NMR would fail to produce the desired S/N enhancement in the case of spin noise spectra. On the other hand, their observation says that the use of Discrete Wavelet Transform (DWT) is one of the simple and convenient ways of increasing the signal of the correlated nuclear spin noise in comparison to the uncorrelated random noise, thus significantly reducing the time required to obtain a spin noise spectrum [2]. In the talk, optimum combinations of wavelets and thresholding methods will be presented along with application to both 1H and 13C spin noise spectra.
In Fourier-transform NMR, a strong free induction decay (FID) signal tends to result in large baseline artifacts, including baseline offset and baseline distortion that are sometimes much larger than small peaks, making data interpretation difficult. As a facile and efficient means of baseline correction of the NMR spectrum, they have explored the utility of the DWT conjugated with spline functions. At the same time, the algorithm allows the possibility of filtration of the signal leading to the removal of the high frequency components, thus achieving base line correction and S/N enhancement simultaneously. The method of Wavelet Transforms has also been applied to suppression of t1 noise in 2-dimensional NMR spectroscopy and the results have been demonstrated for the case of NOESY spectra where spectral artifacts seriously affect the quality of the results obtained.