Small peaks are amplified when stochastic noise is applied

Skip to Navigation


  • Published: Oct 15, 2017
  • Author: Ryan De Vooght-Johnson
  • Channels: Laboratory Informatics / Chemometrics & Informatics
thumbnail image: Small peaks are amplified when stochastic noise is applied

Stochastic resonance methods for chromatography need improvement

Stochastic resonance is a method for amplifying weak signals by adding a limited amount of intrinsic or external noise to the initial output. It can be used for the amplification of weak chromatographic peaks that are not clearly visible, triggering an output signal that was not previously present. A 2003 paper (Pan et al.) introduced a new stochastic resonance algorithm (SRA) for trace chromatographic peaks. However, in practice the algorithm did not always work well, sometimes giving rise to split or distorted peaks or failing to make the peaks clearer. Various modifications have been proposed, but they have tended to be complicated and not at all easy for analysts to apply in practice.

Nanjing researchers have developed a new modified SRA, referred to as DC-SRA, which is suitable for the amplification of hidden weak peaks. The new method is reasonably simple, and equations are given for modifying the output using known parameters. In the new method, a direct current (DC) signal is applied to the original output, which is then processed by stochastic resonance to give the modified signal (O1). The direct current signal is then separately processed by stochastic resonance to give another signal (O2). The final output signal is given by the difference of these two outputs, O1 - O2. The authors used Matlab 7 software for the calculations.

New stochastic resonance algorithm applied to real chromatograms

The modified algorithm was first applied to the examination of traces of menthol in plasma from dogs. GC-MS/MS was employed, using a Varian Saturn CP-3800/2200 ion trap system. Menthol was detected using electron ionisation (EI), following the transition m/z 123 to 81. At a level of 20 ng/mL, the menthol peak was barely visible. Using the original SRA, the peak was amplified, but the baseline became raised and tilted. Application of the modified algorithm, DC-SRA, gave similar amplification with a much flatter baseline. The new algorithm gave good linearity between concentration and peak strength for menthol, showing that it could be used for quantitative analysis.

In another example, the new algorithm was applied to the detection of three alkaloids, colchicine, ephedrine and scopolamine (hyoscine), in human plasma. UHPLC and time-of-flight mass spectrometry (TOF MS) were carried out using a Waters Acquity UHPLC system along with a Waters SYNAPT spectrometer in positive electrospray (ESI) mode. The three peaks were well separated, but very weak at the levels used (1 ng/mL for ephedrine and 0.1 ng/mL for colchine and scopolamine). Using a process of interpolation and re-scaling followed by SRA (original method) gave strong enhancement for the peaks of colchine and scopolamine, but only a slight amplification to that of ephedrine. On the other hand, treating the data using interpolation and re-scaling followed by the new algorithm (DC-SRA) gave enhancement of all three alkaloid peaks.

New DC-SRA method shown to enhance weak peaks

The new DC-SRA method effectively enhanced weak peaks, both for menthol and the three alkaloids, while still being relatively straightforward to use. The method gave good linearity with menthol, suggesting that it can be successfully used for quantitative work. It would be interesting to look at its application to a wider range of analytes. The use of stochastic resonance may seem like ‘black magic’ to some analysts, but such methods have real benefits when working at the limits of the available instrumentation.

Related Links

Journal of Separation Science, Early View paper. Li et al. The stochastic resonance algorithm with the direct current signal as external force and its application to the detection of weak chromatographic peaks.

Chemometrics and Intelligent Laboratory Systems, 2003, 66, 41-49. Pan et al. A new stochastic resonance algorithm to improve the detection limits for trace analysis.

Wikipedia, Stochastic resonance

Article by Ryan De Vooght-Johnson

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.

Follow us on Twitter!

Social Links

Share This Links

Bookmark and Share


Suppliers Selection
Societies Selection

Banner Ad

Click here to see
all job opportunities

Most Viewed

Copyright Information

Interested in spectroscopy? Visit our sister site

Copyright © 2018 John Wiley & Sons, Inc. All Rights Reserved