Data processing: measuring – calibrating – or testing – Calibration or correction system – Zeroing
Patent
1997-12-08
1999-08-31
Barlow, John
Data processing: measuring, calibrating, or testing
Calibration or correction system
Zeroing
702 22, 701 59, 706 15, 364 4803, 36452801, G06F 1542, G01N 2726
Patent
active
059466405
DESCRIPTION:
BRIEF SUMMARY
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a method of and apparatus for analysing the composition of substances and, more particularly, to a method and apparatus which correct for measurement drift, which occurs over time, in analytical apparatus or instruments, such as mass spectrometers.
2. State of the Art
There is a continuing need for rapid and precise analyses of the chemical or biochemical composition of biological and microbiological systems, both within biotechnology and for the purposes of identifying potentially pathogenic organisms. Pyrolysis mass spectrometry is an instrument-based technique which is currently used to perform such analyses. Pyrolysis is the thermal degradation of complex material in an inert atmosphere or a vacuum, which causes molecules to cleave at their weakest points to produce smaller, volatile fragments. For example, one version of this technique is known as Curie-point pyrolysis which involves drying a sample of the complex material to be analysed onto a metal and then heating the metal to its Curie point over a short space of time (typically 0.5 seconds). The degradation products are then separated and counted by a mass spectrometer in order to produce a pyrolysis mass spectrum which can be used as a `chemical profile` or fingerprint of the complex material undergoing analysis. Pyrolysis mass spectrometry involves minimum sample preparation and enables materials to be analysed directly, i.e. without using a reagent. It is also rapid, quantitative, relatively cost efficient and can be conveniently automated. Pyrolysis mass spectrometry is a well established method within biology and microbiology for the differentiation and identification of groups of bacteria, fungi and yeasts, and has also been applied to the authentication of foodstuffs, because it is a highly discriminatory method of analysis which may be used on any organic material.
In recent years, pyrolysis mass spectrometry techniques have been expanded in order to perform quantitative analysis of the chemical constituents of microbial and other samples, by means of supervised learning methods using artificial neural networks. Multivariate linear regression techniques such as partial least squares (PLS) regression and principal components regression (PCR) have been found to provide an effective method of following the production of indole in a number of strains of E. coli grown on media incorporating amounts of tryptophan, quantifying the chemical or biochemical constituents of complex biochemical binary mixtures of proteins and nucleic acids in glycogen, and measuring the concentrations of tertiary mixtures of cells of the bacteria Bacillus subtilis, Escherichia coli and Staphylococcus aureus. Such techniques, also known as chemometric techniques, have also been used for biotechnological purposes in order to perform quantitative analysis of recombinant cytochrome b.sub.5 expression in E. coli, and in order to effect rapid screening of high-level production of desired substances in fermentor broths. Pyrolysis mass spectrometry techniques combined with artificial neural networks have also been exploited for rapid and accurate assessment of the presence of lower-grade seed oils as adulterants in extra virgin olive oils, and for effecting rapid identification of strains of Eubacterium, Mycobacterium, Propionibacterium spp., and Streptomyces. The use of supervised learning techniques to identify and analyse samples from their pyrolysis mass spectra is advantageous because it eliminates the necessity for interpretation of complex principal components analyses and cononical variates analyses plots employed in the past: the identities of the components in a sample are binary-encoded at the output layer of the neural network such that the results may be easily read.
However, pyrolysis mass spectrometry techniques using neural network analysis are generally limited to short-term identification of components wherein all microorganisms are analysed in a single batch. This is due to th
REFERENCES:
patent: 5218529 (1993-06-01), Meyer et al.
patent: 5267151 (1993-11-01), Ham et al.
patent: 5349541 (1994-09-01), Alexandro, Jr. et al.
patent: 5446681 (1995-08-01), Gethner et al.
patent: 5545895 (1996-08-01), Wright et al.
patent: 5554273 (1996-09-01), Demmin et al.
Ali Ipakchi et al., "Neural Network Applications to Measurement Calibration Verification in Power Plants"; Apr. 29. 1996; vol. 53-11,pp. 959-963.
P. Olmos et al.; "Drift Problems in an Automatic Analysis of Gamma-Ray Spectra Using Associative Memory Algorithms"; IEEE transactions on Nuclear Science, vol. 41, No. 3, Jun. 1994.
Fabrio A.M. Davide et al; "Self-organizing multi-sensor systems for odour classification;internal categorization, adaptation and drift rejection"; 1994; in Sensors and Actuators B, 18-19,pp. 244-258.
Hanns-Erik Endres et al.; "Improvement in signal evaluation methods for semiconductor gas sensors"; in Sensors and Actuators B 26-27,pp. 267-270; 1995.
Mary Lou Padgett et al.; "Neural Network Robustness"; Proceedings of the 1991 Summer Computer Simulation Conference; Jul. 22-24, 1991,pp. 330-334.
Goodacre Royston
Kell Douglas Bruce
Barlow John
Gallagher Thomas A.
Gordon David P.
Jacobson David S.
University of Wales Aberystwyth
LandOfFree
Composition analysis does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Composition analysis, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Composition analysis will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2428792