NMR method for detecting and quantifying individual analytes in liquid analyte mixtures

10782256 ยท 2020-09-22

Assignee

Inventors

Cpc classification

International classification

Abstract

A method for detection and quantification of individual analytes in liquid analyte mixtures, by means of NMR spectrometry, in which method a sample of the analyte mixture has a high-frequency pulse applied to it in an NMR spectrometer, and the resulting NMR spectrum is evaluated, is to be developed further in such a manner that it allows a precise analysis of liquid analyte mixtures even when using low-field NMR spectrometers. For this purpose, the sample of the analyte mixture is placed into a low-field NMR spectrometer, and at least one high-frequency pulse that excites only one specific analyte is applied to it.

Claims

1. A method for detection and quantification of individual analytes in liquid analyte mixtures, by means of NMR spectrometry, in which method a sample of the analyte mixture has a high-frequency pulse applied to it in an NMR spectrometer, and the resulting NMR spectrum is evaluated, wherein the sample of the analyte mixture is placed into a low-field NMR spectrometer, and at least one high-frequency pulse that excites only one specific analyte is applied to it.

2. The method according to claim 1, wherein the sample of the analyte mixture has multiple high-frequency pulses applied to it, one after the other, each pulse exciting only one specific analyte, wherein the resulting NMR spectrum is evaluated after every high-frequency pulse application.

3. The method according to claim 1, wherein the magnetic field of the low-field NMR spectrometer has a magnetic intensity between 0.1 and 3.0 T.

4. The method according to claim 1, wherein a high-frequency pulse that is specific to a respective analyte and excites only this analyte is determined and archived for a plurality of analytes.

5. The method according to claim 4, wherein the respective high-frequency pulse, which excites only one specific analyte, is determined in numerically iterative manner.

Description

(1) In the following, the invention is explained in greater detail as an example, using the drawing. This shows, in

(2) FIG. 1 a non-analyzable total NMR spectrum of a bio-fluid,

(3) FIG. 2 a total NMR spectrum of an analyte mixture consisting only of two analytes (alanine and lactate),

(4) FIG. 3 the time-dependent amplitude progression of the high-frequency pulse relative to the total NMR spectrum according to FIG. 2,

(5) FIG. 4 the time-dependent phase progression of the high-frequency pulse relative to the total NMR spectrum according to FIG. 2,

(6) FIG. 5 the NMR spectrum of the pure analyte alanine,

(7) FIG. 6 the time-dependent amplitude progression of a high-frequency pulse for alanine at the beginning of the iteration,

(8) FIG. 7 the related phase progression of the high-frequency pulse according to FIG. 6,

(9) FIG. 8 the total NMR spectrum according to FIG. 2 on a smaller scale,

(10) FIG. 9 the alanine-specific NMR spectrum on the basis of excitation with a high-frequency pulse according to FIGS. 6 and 7,

(11) FIG. 10 the time-dependent amplitude progression of a high-frequency pulse for alanine in the middle of the iteration,

(12) FIG. 11 the related phase progression of the high-frequency pulse according to FIG. 10,

(13) FIG. 12 the total NMR spectrum according to FIG. 2 on a smaller scale,

(14) FIG. 13 the alanine-specific NMR spectrum on the basis of excitation with a high-frequency pulse according to FIGS. 10 and 11,

(15) FIG. 14 the time-dependent amplitude progression of a high-frequency pulse for alanine at the end of the iteration,

(16) FIG. 15 the related phase progression of the high-frequency pulse according to FIG. 14,

(17) FIG. 16 the total NMR spectrum according to FIG. 2 on a smaller scale, and in

(18) FIG. 17 the alanine-specific NMR spectrum on the basis of excitation with a high-frequency pulse according to FIGS. 14 and 15.

(19) In FIG. 1, a total NMR spectrum of a bio-fluid is shown as an example; it was obtained in a low-field NMR spectrometer. It is not difficult to recognize that such an NMR spectrum cannot be analyzed.

(20) In order to nevertheless be able to use low-field NMR spectrometers for analysis of liquid analyte mixtures, the method according to the invention, as described above, is used.

(21) FIG. 2 shows a total NMR spectrum of an analyte mixture, which consists of only two analytes, namely lactate and alanine, in order to allow an easier explanation. This spectrum, shown in FIG. 2, could therefore already be reasonably evaluated fundamentally. Hereinafter, however, it is assumed, for an explanation of the method according to the invention, that the spectrum shown in FIG. 2 cannot be analyzed.

(22) The spectrum shown in FIG. 2 comes about in that a liquid sample disposed in a low-field NMR spectrometer and consisting of only lactate and alanine has a high-frequency pulse applied to it, which is shown in FIGS. 3 and 4. For the intensity in FIG. 2, an arbitrary unit is selected; this is usual since the precise intensity (of a voltage) is very greatly dependent on the hardware (field intensity of the magnet, pre-amplifier) that is used for detection. Therefore, every NMR spectroscopist will measure a different voltage for one and the same pulse, depending on which device he/she is using.

(23) Normally, the absolute frequencies in the NMR lie in the range of 5 MHz to approximately 1 GHz (radio waves to lower microwave range). In contrast, the differences in the frequencies between different NMR signals lie in the range of kHz. The frequency scales would therefore contain very large numbers, which differ only slightly and would be difficult to read. For this reason, a carrier frequency in the MHz range is subtracted from all the frequencies, and only the changes relative to this carrier frequency are represented. The carrier frequency is placed in the center of the spectrum, so that positive and negative frequencies occur for the small changes between the signals. Contrary to convention, the frequency scale runs from right to left, i.e. the positive frequencies are situated on the left side of the X axis. This has historical reasons and is used in this manner up to the present.

(24) FIG. 5 shows the desired NMR spectrum for alanine. If, according to the invention, the liquid sample consisting only of lactate and alanine has a high-frequency pulse specific for alanine, i.e. a high-frequency pulse that excites only alanine, applied to it a low-field NMR spectrometer, then the NMR spectrum shown in FIG. 5 would occur if alanine is actually contained in the sample in a corresponding amount. In this regard, according to the invention, a high-frequency pulse specific for alanine is understood to mean, here, that the pulse is able to excite an entire series of frequency bands, specifically at all positions at which signals occur for alanine in the NMR spectrum.

(25) In order to determine an analyte-specific NMR high-frequency pulse, in this case an alanine-specific pulse, the method of procedure is as follows:

(26) An NMR experiment is fundamentally conducted as follows, in this regard: A high-frequency pulse typically having a length of 10 s is applied to a coil that is wound around a glass tube that contains the analyte mixture (also called sample), using a high-frequency generator. As the result of the action of the pulse, a magnetic moment that changes over time is generated in the analyte mixture. By means of a suitable technical apparatus, the coil is changed from transmission mode to reception mode. The magnetic moment of the sample, which changes over time, then induces a voltage in the coil, which is measured for approximately 500 ms as a function of time and digitalized using an analog-digital converter. After Fourier transformation of the digitalized time signal, a representation of the shape amplitude occurs as a function of the frequency; this is referred to as the NMR spectrum.

(27) Without further provisions, the NMR spectrum contains the NMR signals of all the analytes or substances of the analyte mixture, in other words lactate and alanine according to FIG. 2.

(28) What is sought is a high-frequency pulse that delivers only the NMR spectrum of one (freely selectable) substance of the mixture in an NMR experiment that is conducted on a mixture of multiple substances. For this purpose, the method of optimal control is used. This iterative numerical method optimizes a functional. This is understood to be an expression that is given a mathematical function and delivers a number in return. The goal of the algorithm is to iteratively maximize the functional. The function given to the functional is a quality function that compares the actual state of the pulse that has just been achieved (in other words the spectrum that the pulse actually delivers at the point in time of the calculation) with the reference state (in other words the spectrum that the pulse should ideally deliver). The functional delivers particularly large values if the actual state and the reference state agree. This would be the desired state in which the pulse generates precisely the NMR spectrum that is desired.

(29) The reference state is established in such a manner that only the signals of the desired substance appear in the resulting NMR spectrum. This is always easily possible, because the NMR spectra of all substances of the mixture are known. For calculation of the actual state, the pulse is replaced with a histogram of about 200 rectangular pulses, which are very short (approximately 50 s). In this case, the NMR spectrum that such a rectangular pulse generates can be calculated relatively easily. Such a calculation is known, for example from the journal Progress in NMR Spectroscopy, Vol. 16, pp. 163-192, 1983.

(30) For calculation of the NMR spectrum that is generated by the total high-frequency pulse, all the rectangular pulses of the histogram must be considered successively. Calculation of the quality function then indicates how far the actual state is still removed from the reference state. If the quality function is not yet satisfactory, the amplitudes of the individual rectangular pulses of the histogram are changed, and a new iteration begins. The way in which the amplitudes of the 200 rectangular pulses must be changed is established by means of an analytical updating rule given within the scope of the algorithm. This updating rule ensures that the quality function and thereby the function grows in strictly monotonous manner with every iteration. The calculation is stopped after a number of iterations that must be established in advance by means of empirical tests, since there is no criterion that allows establishing a maximal number of iterations.

(31) In FIGS. 5 to 17, determination of a high-frequency pulse that excites only alanine, in other words an alanine-specific pulse is shown as an example.

(32) At the beginning of the iteration, an alanine-specific high-frequency pulse is used, for example according to FIGS. 6 and 7, which is still approximated relatively close to the high-frequency pulse for the total spectrum according to FIGS. 3 and 4. From this, an alanine-specific NMR spectrum that is still very poor, according to FIG. 9, is obtained, but this already differs from the total NMR spectrum according to Figure (or FIG. 2).

(33) Subsequently, the alanine-specific high-frequency pulse is modified more and more; in FIGS. 10 and 11, an alanine-specific high-frequency pulse is shown in the middle of the iteration, as an example. This results in an alanine-specific NMR spectrum according to FIG. 13, which already clearly differs from the total NMR spectrum according to FIG. 12 (or FIG. 2), and is already developing in the direction of the desired alanine NMR spectrum according to FIG. 5, so to speak.

(34) At the end of the iteration, an alanine-specific high-frequency pulse according to FIGS. 14 and 15 is obtained. This results in an alanine-specific NMR spectrum according to FIG. 17, which clearly differs from the total NMR spectrum according to FIG. 16 (or FIG. 2), and corresponds, to the greatest possible extent, to the desired alanine-specific NMR spectrum according to FIG. 5. Thereby a high-frequency pulse that is specific for alanine (FIGS. 14 and 15) is found, which is archived or stored in memory in the low-field NMR spectrometer. This high-frequency pulse, which is specific for alanine, is therefore a high-frequency pulse that is able to excite an entire series of frequency bands, specifically precisely at all positions at which signals for alanine appear in the NMR spectrum.

(35) If a liquid sample is now inserted into a low-field NMR spectrometer, and if this sample is to be examined with regard to the presence and the amount of alanine, then alanine-specific high-frequency pulse stored in memory is used, and this pulse is applied to the sample. If alanine is present in the sample, an NMR spectrum that is typical for alanine occurs. If, in contrast, no alanine is contained in the sample, then nothing can be detected in the NMR spectrum other than noise signals, i.e. alanine cannot be detected.

(36) In the same manner, specific high-frequency pulses are determined correspondingly for other analytes or substances. During investigation of a sample, in other words an analyte mixture, for different analytes, different analyte-specific high-frequency pulses are subsequently applied to the sample, wherein in each instance, an evaluation takes place in the meantime, i.e. the respective NMR spectrum is evaluated, and it is determined whether and in what concentration the respective analyte is present.