CN108520088A - Emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology and analysis system - Google Patents

Emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology and analysis system Download PDF

Info

Publication number
CN108520088A
CN108520088A CN201810163795.6A CN201810163795A CN108520088A CN 108520088 A CN108520088 A CN 108520088A CN 201810163795 A CN201810163795 A CN 201810163795A CN 108520088 A CN108520088 A CN 108520088A
Authority
CN
China
Prior art keywords
relaxation
sample
model
analog
data
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201810163795.6A
Other languages
Chinese (zh)
Other versions
CN108520088B (en
Inventor
汪红志
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
East China Normal University
Original Assignee
East China Normal University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by East China Normal University filed Critical East China Normal University
Priority to CN201810163795.6A priority Critical patent/CN108520088B/en
Publication of CN108520088A publication Critical patent/CN108520088A/en
Application granted granted Critical
Publication of CN108520088B publication Critical patent/CN108520088B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N24/00—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects
    • G01N24/08—Investigating or analyzing materials by the use of nuclear magnetic resonance, electron paramagnetic resonance or other spin effects by using nuclear magnetic resonance

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Analytical Chemistry (AREA)
  • Immunology (AREA)
  • Pathology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computer Hardware Design (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • General Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

The present invention proposes a kind of emulation nuclear magnetic resonance relaxation spectrum analysis method and its system based on numerical simulation technology, the method includes：Build analog sample information model；Build relaxation device hardware parameter；Structure measures sequence physical model；By the analog sample information model constructed in abovementioned steps, relaxation analysis related hardware parameter, measure sequence physical models coupling, it builds NMR relaxation data and acquires Physics-mathematics model, it carries out analog data acquisition algorithm and program is realized, complete the analog acquisition of initial data；The data that abovementioned steps are acquired carry out relaxation spectra inversion, obtain the distribution map in relaxation time.The present invention can be composed extensively using NMR relaxation in the mating virtual experimental training platform of application technology, can also be applied to the self-study technical principle that related relaxation spectrum analysis technical field applies personnel, be can be also used for instrument development and detection calibration；Sample test spectrogram reference；The data source etc. of relaxation spectra inversion research.

Description

Emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology and analysis System
Technical field
The present invention relates to nuclear magnetic resonance technique fields, and in particular to a kind of emulation nuclear magnetic resonance based on numerical simulation technology Relaxation spectral analysis method and analysis system.
Background technology
What nuclear magnetic resonance had formed that Liquid NMR, solid-state nuclear magnetic resonance, Magnetic resonance imaging stand like the legs of a tripod at present answers Use situation.But nuclear magnetic resonance also steps into stone in recent years other than having above-mentioned important application in chemical field, medical domain The multiple fields such as oily industry, rubber industry, food, agricultural, weaving, construction material, porous media play more and more important Effect, gradually forms new nuclear magnetic resonance spectroscopy test application field.
Nuclear magnetic resonance application field is very extensive, and among constantly expanding.Since nmr analysis has Have speed is fast, accuracy is high, one-shot measurement can get multiple parameters, to sample nondestructive consumption, sample preparation it is simple, to operating personnel Health and environment without influence many advantages, such as, therefore it is many originally use other traditional detection methods application all visiting at present Suo Caiyong nuclear magnetic resonance techniques carry out.
The cardinal principle of nuclear magnetic resonance spectroscopy test is to pass through physical parameter (the mainly relaxation time and expansion to sample Dissipate coefficient) measurement, then be analyzed and characterized sample other performance parameters.Relaxation time includes the knot for the core molecule that is magnetic with sample The parameters such as structure, size, uniformity degree, the state that be combined with each other are associated, by detecting the measurement in relaxation time, can analyze inspection Survey the associated application messages such as different sample compositions, state distribution, molecule combination, molecular hinge.Diffusion coefficient and sample The dynamic change that product include the core molecule that is magnetic is related, reflects the molecular motion information of organization internal, by measuring, Ke Yifen Analyse the associated application messages such as flowing, infiltration, the dispersal direction of detection molecules.Magnetic core is not contained for some itself, But sample interior contains the sample (being referred to as porous media, such as construction material, soil, rock core etc.) of different aperture distribution, can To pass through external perfusion magnetic core sample, according to relaxation time difference or expansion caused by hole environmental difference residing for magnetic core sample Scattered error it is different come indirect detection sample gap situation, to analyze its performance.
Since the relaxation time is directly related with physical environment (such as temperature, pressure) of sample, by the change of physical environment, Can with the molecular dynamics variation characteristic of study sample, therefore in the application based on the relaxation time often in conjunction with alternating temperature or transformation come It carries out.This respect forms the third field of nuclear magnetic resonance application, can be collectively referred to as the application based on relaxation time difference or variation, Also there is referred to as NMR relaxation spectral technology.Since its final result is the distribution in relaxation time, relative to spectroscopic technique Show that the frequency distribution of signal is different, therefore also referred to as time domain nuclear magnetic resonance technique.
In the prior art, many groups for carrying out the research of relaxation inversion algorithm hold for the accuracy and noise of testing algorithm Degree of bearing also carries out the inverting experiment of virtual data, often builds one or more relaxation peak, then inverse by Laplce Transformation (i.e. forward modeling process) directly obtains initial data according to different sequence inverting core, after being then superimposed different degrees of noise Inverting is carried out again.The effect of inversion algorithm is evaluated by comparing the application condition of structure spectrum and final inverting spectrum.Its process As shown in Fig. 1.But the technology whole process does not embody sample, hardware, sequential parameter and the mistake of data acquisition Journey, without the introducing of instrument concept and the theory and technology of sequence measuring, only only algorithm and data processing concept.
NMR relaxation spectral analysis technology has been more and more widely used in many fields, such as food agricultural, stone Oiling work, geological prospecting, rubber material, biological medicine etc.；More and more scientific research personnel need to compose using NMR relaxation Analyzer carries out related scientific research and test job.But NMR relaxation spectral analysis technology is related to numerous nuclear magnetic resonance skills Art basis, especially time series technique, in addition the property complicated and changeable of sample to be tested so that the use of instrument and test data result Explain that many problems occurs in use.These are all derived from the practitioner in each field for NMR relaxation spectral analysis technology Basic principle do not know about, for relaxation inverting the characteristics of is not grasped clearly related.In order to keep related practitioner better Fast understanding grasps the principle of NMR relaxation spectral analysis technology, and acquisition parameter and inverted parameters are preferably arranged, to fast Fast efficiently and accurately show that analysis result, the present invention can meet this requirement.This emulation experiment instrument can simulate generation simultaneously The acquired original data of different signal-to-noise ratio can give the supply in researcher's Develop Data source of related field, and different invertings The contrast test of the noise degrees of tolerance of algorithm.In addition, the limitation based on physical device hardware index, such as diffusion coefficient (D) In measurement, need bigger gradient magnetic that can just measure smaller diffusion coefficient, but due to cannot achieve too high ladder at present Spend eddy current effect problem caused by magnetic field and big gradient magnetic so that the measurement of smaller diffusion coefficient is actually difficult to realize, Emulation experiment instrument based on the present invention can overcome hardware limitation, be carried out to the different required hardware indexs of diffusion measurement Pre assessment.
Invention content
The present invention proposes a kind of emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology, the method Including：
Build analog sample information model；
Build relaxation device hardware parameter；
Build different measurement sequence physical models；
By the analog sample information model constructed in abovementioned steps, relaxation device hardware parameter, measure sequence object Models coupling is managed, structure NMR relaxation data acquire Physics-mathematics model, carry out analog data acquisition algorithm and program is real It is existing, complete the analog acquisition of initial data；
The data that abovementioned steps are acquired carry out relaxation spectra inversion, obtain the distribution map in relaxation time.Obtain relaxation The wave spectrogram of time realizes that the present invention is based on the emulation nuclear magnetic resonance relaxation spectrum analysis methods of numerical simulation technology.
Wherein, analog sample information model is built according to the characteristic of different application field sample to be analysed；The simulation sample Product information model includes principle universal sample information model, typical fields sample message model etc..
Wherein, the principle universal sample information model includes the combination of one or more standard sample components.Preferably, The principle universal sample information model includes the oil samples of fixed T1/T2/D or the relaxation agent solution example that T1/T2/D can be changed.
Further, the present invention also builds typical fields sample message model, and e.g., the gap in porous media material is big The small relational model with T2 relaxation, rubber material crosslink density and relaxation time model, moisture phase and relaxation time in food Relationship etc..Such as, the temperature of associated sample and relaxation and diffusion time relational model can also be built to simulate Caloric test, pressure Power is tested with relaxation and diffusion time relational model to simulate transformation.
Wherein, the relaxation analysis related hardware parameter includes main field B0, main field inhomogeneities deltaB0, radio frequency Field B1, gradient fields G etc..Further, in building the step of relaxation analyzes related hardware parameter, in addition to building basic hardware Except parameter, the error parameter that some are affected to result existing for physical device, such as vortex (different width can also be built Degree and the electric current of damped expoential are superimposed upon on gradient current), the system dead time, Electronics noice etc..
Wherein, the measurement sequence include IR sequences, SR sequences, CPMG sequence, PGSE sequences, PGSE-CPMG sequences, And/or IR-CPMG sequences etc..Sequence construct in the method for the present invention is then the control by providing different data acquisition schemes Parameter is realized.Aforementioned measurement sequence is based on this field basic theories.And actual instrumentation in the prior art is to use sequence Generator (hardware) generates a series of control signal driving related hardwares and completes related work.In addition to aforementioned common relaxation measures Sequence, the present invention can also voluntarily build new relaxation and measure sequence method and model.
Wherein, relaxation spectra inversion, point in obtained relaxation time are carried out based on the initial data that abovementioned steps are acquired Butut includes one-dimensional spectrogram or two-dimentional spectrogram.
Specifically, the present invention is based on the emulation nuclear magnetic resonance relaxation spectrum analysis methods of numerical simulation technology, including step：
In step 1, according to the analog sample information model of the characteristic of different application field sample to be analysed structure.Simulate sample Product structure can be carried out by basic principle sample and typical fields with sample two major classes.
In step 2, the relevant hardware parameter of relaxation analyzer, including main field B0, main field inhomogeneities are built DeltaB0, radiofrequency field B1, gradient fields G etc..
The physical model of (different pulse trains), including IR (reversion in step 3, are built under different analysis measurement methods Restore pulse train), SR (saturation recovery pulse train), CPMG sequence, PGSE sequences, PGSE-CPMG sequences and IR-CPMG Sequence etc. is not limited to these types of sequence.When particular sequence physical model may include pulse type, pulse number, burst length Between, impulse amplitude, interval time, gradient application time and interval time, gradient amplitude, gradient step value, stepping number, data The chronological order etc. of acquisition time, sampling interval, sampling number, reverse time interval and number and each event.In the present invention Sequence construct is realized by providing the control parameter of different data acquisition schemes.Based on the structure of the object of the invention to Sequence (including parameter) pattern and technical solution of dummy acquisition are to belong to the scope of the present invention.
In step 4, by sample message, hardware parameter and the sequential parameter knot constructed by above-mentioned steps 1, step 2, step 3 It being combined, structure NMR relaxation data acquire Physics-mathematics model, carry out analog data acquisition algorithm and program is realized, Complete the analog acquisition of initial data.
In the step 4, the NMR relaxation data acquisition Physics-mathematics model is joined based on sample message, sequence Number, hardware parameter obtain it is each acquire relevant item with data, and to each continuous item carry out it is even multiplied arrive each sample constituent signals, Constituent signals are even added again, are then superimposed the flow of random noise again；With reference to figure 3.The progress analog data acquisition algorithm and Program realization refers to obtaining each and data based on sample message, sequential parameter, hardware parameter to acquire relevant mathematic(al) representation , and to each related mathematic(al) representation item carry out it is even multiplied arrive each sample constituent signals mathematic(al) representation, then constituent signals are connected The whole mathematic(al) representation of random noise is superimposed after adding again.Above-mentioned all mathematic(al) representations by computer interface programming software into Line program writes realization of debugging function.
In step 4, based on the analog sample information model, the measurement sequence physical model and relaxation analysis Related hardware parameter, is combined, and is calculated according to associated expression, and the T1 relaxation of each component of the sample is obtained Item, T2 relaxation terms, T2* relaxation decays item, diffusive attenuation item, inversion recovery item, magnetization vector overturning item, rotating coordinate system are orthogonal Data acquire the factors such as item, i.e. different because of sequence type difference according to different sequences, by any one of aforementioned items or arbitrarily The multinomial factor and sample scale factor carries out the even multiplied sample component signal to each component, and add up multiple aforementioned samples Product constituent signals, it is cumulative to obtain ideal signal, it is then superimposed with analogue noise, the initial data of analog acquisition is obtained, that is, completes The analog acquisition process of initial data.The initial data of acquisition is shown and is stored.
Present invention innovation proposes the above method and step, is analyzed with the NMR relaxation actually used in the prior art Instrument compares, with substantive significant difference and with unexpected advantageous effect.
True hardware component used in actual analysis instrument in the prior art, the instruction pair generated according to sequencer Authentic sample carries out signal acquisition, but which type of the signal specifically acquired is, which type of rule signal follows, and is not existing The technology that hardware in technology is implemented or solved.The present invention proposes signal and is virtually generated by mathematical model.The present invention It also proposed and correspond to different measurement sequences, the initial data virtually generated is different.
Specifically, the present invention proposes, e.g., the analog data acquisition for the CPMG sequence that T2 is measured；For another example, the IR sequences that T1 is measured The analog data acquisition of row；For another example, the analog data acquisition for the SR sequences that T1 is measured, step is identical as IR sequences, and difference exists It is changed in the relaxation term of third step：1-exp(-TI/T1)；For another example, the analog data acquisition for the PGSE sequences that D is measured；For another example, The IR-CPMG sequences that T1-T2 is measured, you can the analog data acquisition models coupling of above-mentioned IR and CPMG sequence get up；Again Such as, the PGSE-CPMG sequences that D-T2 is measured, you can play the analog data acquisition models coupling of above-mentioned PGSE and CPMG sequence Come.
Step 5, the initial data based on analog acquisition in abovementioned steps 4 carries out relaxation spectra inversion, obtains the relaxation time Distribution map.The present invention can obtain one-dimensional spectrogram or two-dimentional spectrogram.The present invention can both simulate one-dimensional spectrum, and (T1 is composed, T2 spectrums, D spectrums Deng), it can also simulating two-dimensional spectrum (T1-T2, D-T2) etc..Inversion algorithm in the step 5 can be used effective in this field Or common a variety of methods can be achieved.Preferably, wherein any one, such as TSVD, SQLR, nerve can be used in the present invention Network etc..
The emulation nuclear magnetic resonance relaxation spectrum analysis system based on numerical simulation technology that the invention also provides a kind of, to reality The existing emulation nuclear magnetic resonance relaxation spectrum analysis method of the present invention based on numerical simulation technology.The system comprises：Simulation Sample message model construction module, measures sequence (under i.e. different analysis measurement methods at relaxation device hardware parameter structure module Different pulse trains) physical model structure module, NMR relaxation data acquisition Physics-mathematics model structure module and Relaxation spectra inversion module；Wherein, analog sample information model builds module, for building analog sample information model；Relaxation point Phase separation closes hardware parameter and builds module, for building relaxation analyzer related hardware parameter；Analysis measurement method (pulse train) Physical model build module, for building different measurement sequence physical models；NMR relaxation data acquire physical mathematics Model construction module is used for the analog acquisition of initial data；And relaxation spectra inversion module, it is used for relaxation spectra inversion, is relaxed The distribution map of Henan time.
The present invention contains the involved whole elements (sample, hardware, software, sequence) of relaxation analyzer instrument work Relevant parameter, collected original number consistent (including signal amplitude, refresh time) with the signal performance of actual instrumentation acquisition Relaxation spectrum is obtained according to inverting is carried out, has the basic conception of instrument test sample.The present invention is different from the prior art, institute of the present invention The data being related to are to carry out dummy acquisition according to mathematical model to obtain, rather than collected by hardware.As shown in Figure 2.
The data acquisition of one peacekeeping two dimension relaxation analytical technology of analog simulation NMR relaxation spectrum of the present invention and relaxation Compose refutation process (being realized using TSVD algorithms) and result.There is the present invention function and effect to include：The T1 relaxation times can be achieved, T2 relaxation times, digital simulation acquisition and inverting realization of the one-dimensional relaxation analytical technologies of D (diffusion coefficient) etc.；D-T2 can be achieved, The data acquisition of the two dimension relaxation spectrum such as T1-T2 and two dimensional inversion function；Show a peacekeeping two dimension relaxation spectral sequence (including IR, SR, CPMG, PGSE, PGSE-CPMG, IR-CPMG) sequence chart；It can be according to than two kinds of samples of column selection, (one kind is T1, T2 and diffusion The all changeless oil sample of parameter, another kind are relaxation agent solution (this project five water sulphur of selection that T1, T2 and D change with concentration Sour copper)) mixture as detection sample；Data acquire and inverted parameters are all variable.
The purposes for the emulation nuclear magnetic resonance relaxation spectrum analysis system based on numerical simulation technology that the invention also provides a kind of, The Application of analysis system can also be applied in the mating virtual experimental training platform that NMR relaxation composes application technology Related relaxation spectrum analysis technical field is using in the self-study technical principle of personnel.It can also be applied to instrument development and test school The purposes such as accurate, the reference of sample test spectrogram, relaxation spectra inversion research data source.The invention also provides one kind based on the present invention Content can develop NMR relaxation spectral analysis technology simulation analysis instrument, which can be used as general nuclear magnetic resonance The mating virtual experimental training platform of relaxation spectrum application technology is used as related relaxation spectrum analysis technical field using personnel In self-study technical principle.Analysis method and system of the present invention can be a software for simulation analysis instrument (in addition to running software Computer carrier except, be free of any hardware).The present invention may be implemented and a relaxation spectrum analyzer (containing hardware and software) Identical data acquisition and processing (DAP) function.
One of present invention application, can be used for the experimental teaching and practical training of the related field talent.In present invention application, no Hardware (hardware separate unit price is between 20 ten thousand to 200 ten thousand) is needed, without being bound by hardware, realizes the experiment behaviour of inexpensive mass Make and train, fully grasps the affecting laws of data acquisition parameters and inverted parameters to the sample relaxation spectrum of different characteristics, to It is theoretical really to grasp relaxation spectral analysis technology, improves analytical technology analysis and operational capacity, lays the foundation for follow-up work.
The advantages that NMR relaxation analytical technology is lossless, quick, green due to its, keeps its applications in various fields more next It is more extensive.It is total that the graduate of many other professions (such as food, agricultural, oil, geology) also more and more touches nuclear-magnetism Relaxation of shaking analyzer, due to specialty background difference, during instrument use, there is a large amount of problems.For example do not know parameter Setting principle and skill, the influence of the pre-treatment difference of sample to result is huge, the influence of inversion algorithm and parameter to result Etc., the fixed operation step and parameter setting only provided according to equipment supplier, after mechanical completion sample and inverting It obtains a result, gap is big between result often occur and predicting, leads to not explain experimental phenomena, and equipment service efficiency is low, resistance Hinder NMR relaxation analytical technology in the development speed and application effect in each field.It has its source in equipment user of service and Related scientific research personnel do not know about the basis of NMR relaxation spectral analysis technology, and operative skill is unskilled.Using true The effective practical training of equipment progress is unrealistic, and hardware separate unit price will not generally be purchased more between 20 ten thousand to 200 ten thousand Complete equipment.Therefore, this simulation analysis instrument can make relevant speciality without being bound by hardware, realize the experiment behaviour of inexpensive mass Make and train, fully grasps the affecting laws of data acquisition parameters and inverted parameters to the sample relaxation spectrum of different characteristics, to It is theoretical really to grasp relaxation spectral analysis technology, improves analytical technology analysis and operational capacity, lays the foundation for follow-up work.
The another application of the present invention, can be used for instrument development and detection calibration；Actual analysis instrument is by numerous factors It influences, such as Electronics noice, vortex, system dead time etc., whether hard by equipment leads to not to evaluate final relaxation spectrum The influence of part or the actual response characteristic of sample.Using this virtual system, standard sample can be shown to ideal hardware condition Under relaxation spectrum as testing standard, to evaluate or formulate hardware index.Actual analysis instrument is influenced by numerous factors, Such as Electronics noice, vortex, system dead time etc., lead to not to evaluate final relaxation spectrum whether by device hardware shadow It rings or the actual response characteristic of sample.For example it in D and D-T2 is measured, is required to use prodigious diffusion gradient G. Gradient fields will necessarily form the amplitude whirlpool different with decaying in switching on ferromagnetic material (such as magnet, probe shield box) Stream.These vortex can be able to correction major part by being vortexed pre-emphasis unit, but it is nothing still to have the influence being vortexed on a small quantity Method is adjusted what correction was fallen by visually observing for engineer, but can influence in relaxation spectrum result.Using virtual system of the invention Standard sample can be obtained under ideal hardware condition the relaxation spectrum (such as without vortex) as testing standard, to evaluation or system by system Determine hardware index；The vortex for thinking to be arranged different amplitudes and decaying is can also be, paired observation is vortexed the influence degree to relaxation spectrum.
The another application of the present invention, is used as the result reference of sample analysis；When actual sample is tested, need by many samples Product processing procedure, including the full oil of high pressure, high pressure are satisfied water, centrifugation, displacement, high temperature etc., the Usefulness Pair in these processing procedures in As a result influence is very big.In order to evaluate sample test processing procedure influence, the system emulation of this Virtual Analysis is ideal place Therefore reason is as a result, can be used for the desired result reference of sample analysis, thus to instruct pretreatment process and technique.
The another application of the present invention, emulation data source is provided for inversion algorithm researcher；Relaxation spectra inversion algorithm is this The research hotspot in field, there is presently no one kind for all samples in terms of accuracy, resolution ratio and noise tolerance all Very efficient algorithm, therefore there are many research groups all to carry out correlative study.Inversion algorithm belongs to pure mathematical operation side Method research, the data source used both are from the relaxation spectrum voluntarily constructed after forward modeling plus make an uproar to obtain early periods, the data with Hardware parameter and sequence are completely irrelevant, can not embody the deviation effects of instrument hardware；The later stage of algorithm can be from actual instrumentation Data are obtained as data source, however to obtain all types of data sources, need different types of equipment, such as PGSE sequences D measure and the D-T2 of PGSE-CPMG sequences is measured, need device configuration to have powerful gradient power amplifier and a gradient coil, one As equipment do not have, so there are certain data source difficulty.Emulation mode and system of the present invention include Virtual Analysis Instrument etc. can calculate researcher for inverting and provide required any data source.
Advantage and advantageous effect of the present invention further include：Without huge hardware it is mating under, can get and NMR relaxation The similar experiment effect of analyzer avoids high request of the hardware imaging teaching appliance to hardware.It can be used as the measurement effect of actual sample The reference of fruit compares and parameter setting understands.Advantageous effect of the present invention further includes that can be used for intuitive rapidly study to understand core The principle of magnetic resonance relaxation analytical technology；The early stage that can be used for relaxation spectrum skilled addressee learns by oneself principle and operation, is familiar with ginseng Several setting rules；The signal and relaxation spectrum result that can be used for actual test sample compare reference.
Description of the drawings
Fig. 1 is the flow diagram for obtaining inverting relaxation spectrum in the prior art.
Fig. 2 is the flow diagram that inverting relaxation spectrum is obtained in the present invention.
Fig. 3 A indicate that the workflow schematic diagram of the method for the present invention, Fig. 3 B indicate that workflow and analysis result of the present invention are aobvious The schematic diagram for the specific implementation shown.
Fig. 4 is in the embodiment of the present invention, and IR sequences survey the schematic diagram of T1：Fig. 4 (a) shows IR sequence charts, and Fig. 4 (b) is more FID signal under secondary TI, Fig. 4 (c) are the first point signal amplitude under multiple TI, and Fig. 4 (d) is distributed for T1 relaxation.
Fig. 5 is in the embodiment of the present invention, and CPMG sequence surveys the schematic diagram of T2：Fig. 5 (a) shows CPMG sequence figure, Fig. 5 (b) For echo, Fig. 5 (c) is echo peak dot, and Fig. 5 (d) is distributed for T2 relaxation.
Fig. 6 is the schematic diagram of SR sequence measurings T1 in the embodiment of the present invention:Fig. 6 (a) shows that SR sequence charts, Fig. 6 (b) are FID signal under multiple TI, Fig. 6 (c) are the first point signal amplitude under multiple TI, and Fig. 6 (d) is distributed for T1 relaxation.
Fig. 7 is the schematic diagram of PGSE sequence measurings D in the embodiment of the present invention：Fig. 7 (a) shows PGSE sequence charts, Fig. 7 (b) For the echo-signal under multiple G, Fig. 7 (c) is the echo-signal peak dot amplitude under multiple G, and Fig. 7 (d) is the distribution of D.
Fig. 8 is the schematic diagram of PGSE-CPMG sequence measurings D-T2 in the embodiment of the present invention：Fig. 8 (a) shows PGSE-CPMG Sequence chart, Fig. 8 (b) are the echo-signal under multiple G, and Fig. 8 (c) is the echo signal peak dot amplitude under multiple G, and Fig. 8 (d) is D-T2 relaxation is distributed.
Fig. 9 is the schematic diagram of IR-CPMG sequence measurings T1-T2 in the embodiment of the present invention：Fig. 9 (a) shows PGSE-CPMG Sequence chart, Fig. 9 (b) are the echo-signal under multiple G, and Fig. 9 (c) is the echo signal peak dot amplitude under multiple G, and Fig. 9 (d) is D-T2 relaxation is distributed.
Figure 10 is the two-dimentional D-T2 spectrograms that the mesh formats of PGSE-CPMG sequence measurings D-T2 are shown.
Figure 11 is the flow diagram of the Raw Data Simulation acquisition of step 4 in the method for the present invention.
Figure 12 is the structural schematic diagram of analysis system of the present invention.
Specific implementation mode
In conjunction with following specific examples and attached drawing, the present invention is described in further detail.The process of the implementation present invention, Condition, experimental method etc. are among the general principles and common general knowledge in the art, this hair in addition to the following content specially referred to It is bright that content is not particularly limited.
As shown in Fig. 2, Fig. 3 A, 3B, the present invention proposes a kind of emulation NMR relaxation based on numerical simulation technology Spectral analysis method.The method includes：Build analog sample information model；It builds relaxation and analyzes related hardware parameter；Structure is not With measurement sequence physical model；The analog sample information model constructed in abovementioned steps, relaxation are analyzed into related hardware Parameter measures sequence physical models coupling, and structure NMR relaxation data acquire Physics-mathematics model, carry out analogue data and adopt Set algorithm and program are realized, the analog acquisition of initial data is completed；It is anti-to carry out relaxation spectrum for the data that previous step is acquired It drills, obtains the distribution map in relaxation time.Involved data are to carry out dummy acquisition according to mathematical model to obtain in the method for the present invention It arrives, rather than is collected by hardware.
Specifically, such as the workflow of Fig. 3 A, Fig. 3 B the method for the present invention indicated, the present invention is based on numerical simulation technologies Emulate nuclear magnetic resonance relaxation spectrum analysis method, including step：
Step 1, the analog sample information model built according to the characteristic of different application field sample to be analysed, sample structure It can be according to basic principle sample and typical fields sample two major classes.
The analog sample information includes T1, T2, PD, D, constituent scale factor of different samples etc.；Or with The corresponding voidage of above-mentioned parameter, the degree of cross linking or hydrone phase etc..
Step 2, the relevant hardware parameter of relaxation analyzer is built.
Hardware parameter includes main field B0, radiofrequency field B1, gradient fields G, main field inhomogeneities deltaB0, Frf etc..
Step 3, the physical model of different analysis measurement methods (pulse train) is built.The sequential parameter include TR, TE, TI, num_G, num_Echo, SW, TD, tau, P1, t, Delta, delta etc..
Step 4, sample message, hardware parameter and the sequential parameter constructed by above-mentioned steps 1,2,3 are combined together, structure NMR relaxation data acquisition Physics-mathematics model is built, analog data acquisition algorithm is carried out and program is realized, complete original number According to analog acquisition.
Step 5, the initial data based on step 4 analog acquisition carries out relaxation spectra inversion, obtains the distribution in relaxation time Figure.
Further, step 1, analog sample information model is built according to the characteristic of different application field sample to be analysed.
The present invention provides several instance model examples, but is not limited to these types, further includes other to realize mesh of the present invention Instance model and its structure, be based on the present invention inventive concept, belong to the scope of the present invention.
Principle universal sample information model includes the combination of one or more standard samples.Preferably, the principle is general Sample message model includes the oil samples of fixed T1/T2/D and the relaxation agent relaxation agent solution example that T1/T2/D can be changed.
For example, the universal sample model of principle could be provided as the combination of one or more standard specimens, an example is two kinds The combination of standard specimen, one of which are the oil samples of fixed T1/T2/D, another kind be various concentration relaxation agent (such as CuSO4 or MnCl2 etc. is not limited to both relaxation agents) solution example (various concentration has different T1/T2/D).
The present invention also constructs sample message model, e.g., constructs the pore size and T2 relaxation in porous media material Relational model, rubber material crosslink density and relaxation time model, moisture phase and relaxation time relationship etc. in food.This hair The bright temperature that can also build associated sample and relaxation and diffusion time relational model are simulating Caloric test, pressure and relaxation It is tested with diffusion time relational model to simulate transformation.
Typical fields sample message model further includes the sample message model built according to different application domain feature.Example Such as, the present invention provides three field instance model examples：
The aperture of porous media material (such as nano material, construction material, dirt band, coal mining, geological core) The T2/T1 information associations of size and information probes (such as water or oil) structure in hole, there is T2=C.r.T2 is the sample in different aperture Product relaxation time, r are pore size (being indicated with radius), and C is transformation model coefficient, with internal surface of hole relaxation rate, the shape in hole Correlation, different samples to be tested take different C values.For general rock core, C value 10us/nm, i.e. water in 1nm grades of radius gap The sample relaxation time is 10ns.The water sample relaxation time is 10ms in the gap of 1000nm.
The T1/T2 information associations etc. of the degree of cross linking structure of High-molecular rubber material.
Hydrone state (Free water, medium water, in conjunction with water etc.) structure T1/T2/D information in the aqueous systems such as food is closed Connection；Free water is according to T2=1200ms, T1=1600ms, D=2.6*e-9m^2/S, the Gaussian Profile of halfwidth 0.5；Medium water According to T2=50ms, T1=70ms, D=0.3*e-9m^2/S, the Gaussian Profile of halfwidth 0.2；In conjunction with water according to T2=2ms, T1=2.4ms, D=0.1*e-9m^2/S, the Gaussian Profile of halfwidth 0.2.
In step 2, the relevant hardware parameter of relaxation analyzer, including main field B0, main field inhomogeneities are built DeltaB0, radiofrequency field B1, gradient fields G etc..
In step 3, the physical model of different analysis measurement methods (pulse train), including IR (inversion recovery pulses are built Sequence), SR (saturation recovery pulse train), CPMG sequence, PGSE sequences, PGSE-CPMG sequences and IR-CPMG sequences etc., It is not limited to these types of sequence；Particular sequence model includes pulse type, pulse number, time in burst length, impulse amplitude, interval Between time, gradient application time and interval time, gradient amplitude, gradient step value, stepping number, data acquisition time, sampling Every, the chronological order etc. of sampling number, reverse time interval and number and each event.
For example, IR sequences, build following parameter to realize respectively：Max_TI (maximum reversing time), TI-num are (when reversion Between number), (reversing time stationing mode can be that logspan linearly layouts and linearly layouts etc. two with linespan to TI-span Kind of mode), D0 (repetition time), SW (sampling bandwidth), TD (sampling number), P90 (90 degree of pulse widths), P180 (180 degree arteries and veins Rush width)；These parameters are all related to the acquisition of IR sequence datas, subsequently can all use these parameters in the acquisition of comprehensive data.
For example, CPMG sequence, the parameter built is needed to be：TE (echo time), Echo-num (echo number), D0 (weights The multiple time), SW (sampling bandwidth), TD (sampling number), P90 (90 degree of pulse widths), P180 (180 degree pulse width).
For example, SR sequences, identical as IR sequence construct parameters；Difference lies in subsequent data collecting model is variant.
For example, PGSE sequences, need to build parameter be：Max-G (greatest gradient field intensity), G-Num (gradient steppings Number), Delta (spacing gradient time), delta (gradient duration), TE (echo time), Echo-num (echo number), D0 (repetition time), SW (sampling bandwidth), TD (sampling number), P90 (90 degree of pulse widths), P180 (180 degree pulse width).
For example, IR-SPMG sequences, the parameter built is needed to be：TE (echo time), Echo-num (echo number), Max-TI (maximum reversing time), TI-num (reversing time number), TI-span (reversing time stationing mode, Ke Yishi Logspan linearly layouts the two ways such as linearly layout with linespan), D0 (repetition time), SW (sampling bandwidth), TD (adopt Number of samples), P90 (90 degree of pulse widths), P180 (180 degree pulse width).
In step 4, by sample message, hardware parameter and the sequential parameter knot constructed by above-mentioned steps 1, step 2, step 3 It being combined, structure NMR relaxation data acquire Physics-mathematics model, carry out analog data acquisition algorithm and program is realized, Complete the analog acquisition of initial data.
As shown in figure 11, in step 4, in conjunction with the analog sample information, the parameter for measuring sequence and described hard Part parameter obtains the T1 relaxation terms of each component of sample, T2 relaxation terms, T2* relaxation decays item, diffusive attenuation item, inverts Restore item, magnetization vector overturning item, rotating coordinate system orthogonal data and acquire item, by aforementioned one or more and sample component ratio The example factor carries out even multiplied to one sample component signal, and multiple sample component signals even add, cumulative sample sets are obtained Sub-signal forms the ideal signal of a sample to get to the cumulative of all components signal of a sample.Then, it is superimposed Upper analogue noise obtains the initial data of analog acquisition, that is, completes the analog acquisition process of initial data.The initial data of acquisition It is further able to show and store.Wherein, even multiplying will each factor progress product budget successively；Even plus i.e. by each group sub-signal according to Secondary addition.
Wherein, a sample component signal is accordingly formed for each component of sample.Sample component (sample composition Ingredient) refer to each sample include different constituent, different constituents has different physical messages.It is different The different constituents that sample includes have different physical messages.For example, including water, oily two kinds of differences in an actual sample Constituent, constituent water, constituent oil are respectively provided with the physical parameters such as different T1, T2, diffusion coefficient, ratio.
Wherein, sample component signal refers to the sample component that each constituent for including accordingly is formed in sample Signal.It is mathematic(al) representation in implementation process, result of calculation is an one-dimension array data.Can include in a sample 1-n constituent is correspondingly respectively formed 1-n sample component signal, e.g., sample component information 1, sample component information 2 ..., sample component signal n, wherein n >=1, n take positive integer.
For example, a sample includes 2 components, then sample component signal 1, sample component signal 2, then, the two are formed The company of progress adds, and forms the sample signal, that is, ideal signal that adds up.
For example, sample includes n kind components, then calculate separately to obtain the sample component signal 1 of each constituent, Sample component signal 2 ..., sample component signal n, then, carry out n sample component signal company add, formed add up after Sample signal, that is, ideal signal.
Wherein, ideal signal refers to that the ideal signal of dummy acquisition is to be joined by sample, sequential parameter, hardware in the present invention The signal obtained by mathematical operation after the structure model such as number due to not accounting for the noise of practical electronics circuit, therefore is reason Think signal；After the superposition for carrying out simulation random noise, the signal of actual instrumentation acquisition can be more really expressed.The random of superposition is made an uproar Sound amplitude is adjustable, then can reach the inverting spectrum algorithm effect and spectrogram difference for simulating different signal-noise ratio signals.
Wherein, the T1 relaxation terms (1), T2 relaxation terms (2), T2* relaxation decays item (3) of each sample component, diffusion decline Deduction item (4), inversion recovery item (5), magnetization vector overturning item (6), rotating coordinate system data acquisition item (7) are to be based on the mould Quasi- sample message, relaxation measure the parameter of sequence and related hardware parameter is calculated according to associated expression and obtained.Its In, the associated expression is the prior art generally known in the art.
Wherein, for different sequences, T1 relaxation terms, T2 relaxation terms, T2* relaxation decays item, diffusive attenuation item, anti-may be selected The different multinomial combinations for turning to restore item, magnetization vector overturning item, rotating coordinate system data acquisition item even multiply.Then, then It is multiplied with sample scale factor.As shown in figure 11.
Such as IR sequences, T1 relaxation terms (1), T2* relaxation decays item (3), diffusive attenuation item (4), inversion recovery item are selected (5), rotating coordinate system data acquisition item (7) even multiplies.
Such as CPMG sequence, selection：T1 relaxation terms (1), T2 relaxation terms (2), diffusive attenuation item (4), inversion recovery item (5), Rotating coordinate system data acquisition item (7) even multiplies.
Such as PGSE sequences, selection：T1 relaxation terms (1), diffusive attenuation item (4), inversion recovery item (5), magnetization vector overturning Item (6), rotating coordinate system data acquire Xiang Liancheng.
Such as IR-CPMG sequences, selection：T1 relaxation terms (1), T2 relaxation terms (2), T2* relaxation decays item (3), diffusive attenuation Item (4), inversion recovery item (5), rotating coordinate system data acquisition item (7) even multiply.
Such as PGSE-CPMG sequences, selection：T1 relaxation terms (1), T2 relaxation terms (2), diffusive attenuation item (4), inversion recovery item (5), magnetization vector overturning item (6), rotating coordinate system data acquisition item (7) even multiply.
For thering is num_echo, num_G, num_TI etc. to have the parameter of number of variations, then need through a series of step values Company adds up after multiplying.
Wherein, sample component scale factor refers to building the basic ginseng of directly include one in analog sample information model Amount.It need not calculate.For example, containing component water and oil in certain a sample, wherein the content 20% of component water, oily content 80%.Sample component scale factor can influence respective signal amplitude.
Based on ideal signal obtained above, then it is superimposed with simulation random noise, that is, completes analog data acquisition.Wherein, The method for being superimposed with simulation random noise can refer to the prior art.
It is based on nuclear magnetic resonance relaxation measure theory to measure sequence IR, SR, CPMG, PGSE, IR-CPMG, PGSE-CPMG.
In the present invention, for different measurement sequences, the initial data virtually generated is also different.And further, base One-dimensional spectrum (T1 is composed, T2 spectrums, D spectrums etc.) can have both been simulated in initial data, it also can simulating two-dimensional spectrum (T1-T2, D-T2) etc..
Specifically, such as the analog data acquisition of the T2 CPMG sequences measured：
CPMG sequence surveys T2, as shown in figure 5, a, b, c, d in Fig. 5 show CPMG sequence figure, echo, echo respectively Peak dot and the distribution of T2 relaxation.
The first step：Data sampling model, i.e., the time array of one one-dimensional discrete, including sampling number, is adopted are built first Belt transect is wide, the echo time, including the multiple parameters such as sampling time t, such as t=(0:1/SW:TE-1/SW)+(i-1)*TE；For sequence Determining sampling rule, including sampling bandwidth SW, echo time TE, sampling number TD, i are the iteration serial number of number of echoes； Analog sampled data algorithm examples are：
Second step builds the amplitude information of two kinds of sample signals, including sample size, RF excited effect, sample respectively The factors such as the intrinsic T2 relaxation effects of T1 relaxation effects, sample, the practical T2* relaxation effects of sample, final amplitude be every factor it Product；T1 relaxation effects are:1-exp(-TR/T1).
It is directed to sample ratio P, TR, TE, T1, T2, B1, tau, deltaB0, the mould of the aforementioned structure such as echo number Shape parameter；
Third walks, and builds the frequency information of two kinds of sample signals respectively, including rotation when off resonance effect, data acquire The factors such as effect, final frequency are the product of every factor；There is real and imaginary parts sampling in rotating acquisition, therefore data are plural number According to；
It is directed to TE, t, T2, the model parameter of the aforementioned structure such as B0, Frf；
4th step is multiplied to the amplitude effect of each sample signal with frequency effect, obtains the single of each sample Echo-signal；
Each sample signal of 4th step is added by the 5th step, then after being superimposed corresponding analogue noise, is returned as the 1st The echo-signal array of wave；
6th step, repeats above-mentioned one to five step, and analog acquisition obtains the signal array of n echo；And by n array shaping It is one-dimension array to get to a CPMG sequence；
7th step repeats n-1 times if the accumulative frequency n of CPMG sequence is not 1 to above-mentioned one to six step；One N cpmg echo array is obtained, the CPMG initial data after being overlapped averagely as final analog acquisition.
Wherein P, T1, T2 are the sample message of structure；SW, TD, TE, Num-echo, TR, TE, tau, t are the sequence of structure Model parameter；B0, B1, Frf, deltaB0 are constructed in hardware parameter.
Specifically, such as the analog data acquisition of the T1 IR sequences measured：
IR sequences survey T1, as shown in figure 4, a, b, c, d in Fig. 4 show that the FID under IR sequence charts, multiple TI believes respectively Number, the first point signal amplitude under multiple TI and T1 relaxation distribution.
The first step builds echo data sampling model, i.e., the time array of one one-dimensional discrete, including sampled point first It counts, sampling bandwidth, the echo time, including the multiple parameters such as sampling time t, such as t=(0:1/SW:TE-1/SW)+(i-1)*TE； For the sampling rule that sequence determines, including sampling bandwidth SW, echo time TE, sampling number TD, i are the iteration of number of echoes Serial number；Since IR sequences only take first echo peak dot to inverting, i could be provided as 1.But in order to which data acquire Identical as true instrument performance, the value of i still can be identical as CPMG sequence.
Second step builds the cloth point model of a TI reversing time, such as log cloth point models, TI=logspace (minTI, MaxTI, num_TI)；Or it linearly layouts：TI=linespace (MinTI, MaxTI, Num_TI)；
Third walks, and the amplitude information of two kinds of sample signals is built respectively, including sample size, RF excited effect, sample The factors such as the intrinsic T2 relaxation effects of T1 relaxation effects, sample, the practical T2* relaxation effects of sample, final amplitude be every factor it Product；Sample T1 relaxation effect items herein are：1-2exp(-TI/T1)；TI is the value of first TI cloth points；
It is directed to sample ratio P, TR, TE, T1, T2, B1, tau, deltaB0, the mould of the aforementioned structure such as echo number Shape parameter；
4th step builds the frequency information of two kinds of sample signals respectively, including rotation when off resonance effect, data acquire The factors such as effect, final frequency are the product of every factor；There is real and imaginary parts sampling in rotating acquisition, therefore data are plural number According to；
It is directed to TE, t, T2, the model parameter of the aforementioned structure such as B0, Frf；
5th step is multiplied to the amplitude effect of each sample signal with frequency effect, obtains the single of each sample Echo-signal；
Each sample signal of 5th step is added by the 6th step, then after being superimposed corresponding analogue noise, is returned as the 1st The echo-signal array of wave；
7th step, repeats above-mentioned one to six step, and analog acquisition obtains the signal array of n echo；And by n array shaping It is one-dimension array to get to a CPMG sequence；
8th step repeats n-1 times if the accumulative frequency n of CPMG sequence is not 1 to above-mentioned one to seven step；One N cpmg echo array is obtained, the echo after being overlapped averagely as first TI value.
9th step repeats above-mentioned one to eight step, uses different TI values every time, you can obtain num_TI echo；Often A echo is an one-dimension array, therefore a two-dimensional array can be obtained, and dimension is respectively num_TI：num_echo；
Tenth step：By often going for the two-dimensional array of the 9th step, according to regular time position (array serial number) or pass through First max methods is found, it is first Echo Rating to obtain maximum value；This num_TI Echo Rating is formed one one Dimension group, the as initial data of IR sequence acquisitions are used for follow-up T1 relaxation spectra inversions；
Wherein P, T1, T2 are the sample message of structure；SW, TD, TE, I, Num-echo, TR, TE, tau, t are the sequence of structure Row model parameter；B0, B1, Frf, deltaB0 are constructed in hardware parameter.
Specifically, such as the analog data acquisition of the T1 SR sequences measured：Its step is identical as IR sequences, and difference is third The relaxation term of step is changed to：1-exp(-TI/T1).
SR sequence measuring T1, as shown in fig. 6, a, b, c, d in Fig. 6 show that the FID under SR sequence charts, multiple TI believes respectively Number, the first point signal amplitude under multiple TI and T1 relaxation distribution.
Specifically, such as the analog data acquisition of the D PGSE sequences measured：
PGSE sequence measuring D, as shown in fig. 7, a, b, c, d in Fig. 7 show returning under PGSE sequence charts, multiple G respectively The distribution of echo-signal peak dot amplitude and D under wave signal, multiple G.
The first step builds echo data sampling model first, essentially identical with the first step of CPMG sequence.Difference is i =1, because only needing to acquire an echo；
Second step builds the Step Model of diffusion gradient, G=linespace (0, maxG, num_G)；Max_G is maximum Gradient fields, num_G are gradient number of steps；
Third walks, and builds the amplitude information of two kinds of sample signals respectively, including sample size, RF excited effect, diffusion Factors, the final amplitudes such as effect, sample T1 relaxation effects, the intrinsic T2 relaxation effects of sample, the practical T2* relaxation effects of sample are each The product of item factor；Remaining effect is identical as CPMG sequence；Sample diffusion effect item is：1-2exp(-b*D)；B be diffusion because Son, expression formula are b=(gama*G*delta) .^2*Delta；D is sample diffusion coefficient；
It is directed to sample ratio, P, D, TR, TE, T1, T2, B1, tau, deltaB0, the aforementioned structure such as echo number Model parameter；
4th step builds the frequency information of two kinds of sample signals respectively, including rotation when off resonance effect, data acquire The factors such as effect, final frequency are the product of every factor；There is real and imaginary parts sampling in rotating acquisition, therefore data are plural number According to；
It is directed to TE, t, T2, the model parameter of the aforementioned structure such as B0, Frf；
5th step is multiplied to the amplitude effect of each sample signal with frequency effect, obtains the single of each sample Echo-signal；
Each sample signal of 5th step is added by the 6th step, then after being superimposed corresponding analogue noise, is returned as the 1st The echo-signal array of wave, as a SE echo-signal；
The accumulative frequency n of 7th step, SE sequences is not 1, then being repeated n-1 times to above-mentioned one to seven step；One is obtained n A SE echoes group, the echo-signal after being overlapped averagely as first G value.
8th step repeats above-mentioned one to seven step, uses different G values every time, you can obtain num_G SE echo；Each Echo is an one-dimension array, therefore a two-dimensional array can be obtained, and dimension is respectively num_G：TD；
9th step：By often going for the two-dimensional array of the 8th step, according to regular time position (array serial number) or pass through First max methods is found, it is echo peak shop to obtain maximum value；This num_G echo peak dot is formed into a dimension Group, the as initial data of PGSE sequence acquisitions are used for follow-up D relaxation spectra inversions；
Wherein P, T1, T2, D are the sample message of structure；SW, TD, TE, I, Num-G, TR, TE, tau, t are the sequence of structure Row model parameter；B0, B1, Frf, deltaB0, maxG are constructed in hardware parameter.
Specifically, the IR-CPMG sequences that T1-T2 is measured, you can by the analog data acquisition of above-mentioned IR and CPMG sequence Models coupling gets up.IR-CPMG sequence measuring T1-T2, as shown in figure 9, a, b, c, d in Fig. 9 show PGSE- respectively CPMG sequence figure, the echo-signal under multiple G, the echo signal peak dot amplitude under multiple G and the distribution of D-T2 relaxation.
Specifically, the PGSE-CPMG sequences that D-T2 is measured, you can adopt the analogue data of above-mentioned PGSE and CPMG sequence Collection models coupling gets up.PGSE-CPMG sequence measuring D-T2, as shown in figure 8, a, b, c, d in Fig. 8 are shown respectively PGSE-CPMG sequence charts, the echo-signal under multiple G, the echo signal peak dot amplitude under multiple G and D-T2 relaxation point Cloth.
Present invention innovation proposes above four steps, with the NMR relaxation analyzer actually used in the prior art Compare, with substantive significant difference and with unexpected advantageous effect.
Step 5, to the different sequence signals acquired in abovementioned steps 4, relaxation spectra inversion is carried out, obtains point in relaxation time Butut.The present invention can obtain one-dimensional spectrogram or two-dimentional spectrogram.Inversion algorithm in the step can be used effective in this field Or common a variety of methods can be achieved.Preferably, wherein any one, such as TSVD, SQLR, nerve can be used in the present invention Network etc..
As shown in figure 12, the present invention is based on the emulation nuclear magnetic resonance relaxation spectrum analysis systems of numerical simulation technology comprising： The object of analog sample information model, relaxation analysis related hardware parameter structure module, difference analysis measurement method (pulse train) Manage model construction module, NMR relaxation data acquisition Physics-mathematics model structure module and relaxation spectra inversion module.Its In, analog sample information model, for building analog sample information model；Relaxation analyzes related hardware parameter and builds module, uses Related hardware parameter is analyzed in structure relaxation；The physical model of analysis measurement method (pulse train) builds module, for building Difference measures sequence physical model；NMR relaxation data acquire Physics-mathematics model and build module, for initial data Analog acquisition；Relaxation spectra inversion module is used for relaxation spectra inversion, obtains the distribution map in relaxation time.
The two-dimentional D-T2 spectrograms that the mesh formats of PGSE-CPMG sequence measurings D-T2 as shown in Figure 10 are shown indicate ginseng Number setting, sample selection, sequence selection, hardware setting, data acquisition is shown, initial data is shown, relaxation refutation process, result Display etc..The present invention can be directed to analog sample, realize data acquisition identical with true instrument and performance and relaxation Compose refutation process.It is quite analogous to an equipment course of work and result.
In the present invention, the definition of relational language：
Relaxation spectrum：relaxation spectrum；The distribution map of the relaxation component and ratio that include in sample
Sample：The different test specimens of simulation；Such as water+oil principle standard sample, porous media sample etc.；In sample extremely Include a kind of component, generally 2 components or multicomponent less；
Component：The material composition for including in sample；Such as water+oil principle standard sample, including water and oily two kinds of components；
T1 is composed：Sample longitudinal relaxation time is distributed；
T2 is composed:The lateral relaxation time of sample is distributed；
D is composed：The diffusion coefficient of sample is distributed；
T1-T2 is composed:The Two dimensional Distribution of longitudinal direction-lateral relaxation time of sample；
D-T2 is composed：The Two dimensional Distribution of diffusion-lateral relaxation time of sample；
IR sequences：To obtain the data acquisition sequence of T1 relaxation information；
SR sequences：Another kind is obtaining the data acquisition sequence of T1 relaxation information；
CPMG sequence：To obtain the data acquisition sequence of T2 relaxation information；
PGSE sequences：To obtain the data acquisition sequence of diffusion coefficient information；
PGSE-CPMG sequences：To obtain the data acquisition sequence of D-T2 two dimension relaxation information；
IR-CPMG sequences：To obtain the data acquisition sequence of T1-T2 two dimension relaxation information；
Inverting：Inversion realizes the determination to the relaxation component in the echo string data of acquisition by mathematical method；
TSVD：SVD inversion algorithms based on cutoff value.
Related basic theories of the present invention.
NMR relaxation analytical technology is divided into two processes：
A) data acquisition, either T2, T1 and D or two-dimentional spectral technology, data acquisition are substantially designs Different coefficient matrixes, the data acquisition carried out according to certain rule (are the different kernel coefficients of construction for experiment Matrix K)；
B) relaxation inverting obtains the distributed process of relaxation spectrum by gathered data；Such as the measurement for T2, data For：M=KF+ ε, K=exp (- t/T2)；Measurement for T1, data are：M=KF, K=1-exp (- t/T2) (SR sequences) Or K=1-2exp (- t/T2) (IR sequences)
1, data acquisition sequence
1.1 reverse-revert methods survey T1
It is based onBy the way that different tau is arranged, different signals is acquired, signal amplitude follows T1 indexes increase rule；
1.2 saturation recovery methods survey T1
It is based onBy the way that different tau is arranged, signal amplitude increases also in compliance with index Add rule；
1.3 CPMG sequences measure T2
It is based onBy acquiring multiple echo-signals, the amplitude of echo-signal follows T2 decaying Rule；
1.4 PGSE sequence measurings D
Based on M/M0=exp [- (γ G δ)2D Δs], by the way that multiple and different gradient intensities is arranged, the amplitude of echo-signal declines Subtract exponentially related to the diffusion coefficient of sample；
1.5 PGSE-CPMG sequence measurings D-T2
PGSE sequences and CPMG sequence are combined together, D and T2 can be measured simultaneously；
1.6 IR-CPMG measure T1-T2
IR sequences and CPMG sequence are combined together, T1 and T2 can be measured simultaneously；
2, relaxation spectra inversion
The essence of relaxation spectra inversion is to solve for following equation：
Y=Ax+ ξ (1)
Wherein：Y=(y1,y2,…,yn)TFor the deamplification of the spin echo of measurement；
A=[exp (- ti/Tj)]n×m, it is coefficient matrix；Different data acquisition sequences, A are different；
X=(x1,x2,…xm)TFor relaxation time TjThe range value of corresponding each point, as relaxation spectrum；
Tj(j=1,2 ... m) layout for preset T, it is contemplated that the coverage area of spectrum surpasses from 0.1ms to 10000ms Cross three orders of magnitude.It is generally used logarithm uniform stationing close after first dredging；
ξ indicates random noise；
To the solution of above formula, that is, the process of F is acquired, the minimum side that can be distributed with fitting by solving measured data Difference obtains, i.e.,：
Min||M–KF||2. (2)
Specifically analytical expression is：F=(KTK) -1KTM (3)
The inverse matrix of (...) -1 representing matrix in formula；Then for exponential distribution matrix K, KTK is a singular matrix, There is no an inverse matrixs.This singularity is exactly the pathosis characterization of Laplace transform, illustrates the distribution of its spectrum to data Minor change it is very sensitive.That is, there is countless distributions can meet (3) formula.This project uses singular Value decomposition (SVD) method carries out Singular Value Decomposition method.
For COEFFICIENT K, singular value decomposition is carried out to it by SVD and is represented by：
K=U Σ VT
U in formula, V are respectively left singular component (left-hand singular vectors) and right singular component (right-hand singular vectors) is unitary matrix (unitary matrices).Σ is all singular values by K The diagonal matrix of composition；
So using SVD methods, (2) formula can be write as：
Min||M–UΣVTF||2.
In this regard, the simplest analytical expression of least square method is：
F=V Σ-1UTM
Above formula can transform to：
Wherein
nsFor the singular value number of K；Vi is corresponding right singular component value；Ui is corresponding left singular component value；SiFor i A non-zero singular value.By (4) formula it is easy to see that relaxation distribution F is the linear combination of a series of right singular component value in fact. Such as figure simulation result using more singular values it is found that can be obtained the spectrum result closer to truth.It is simultaneously it is found that minimum Singular value component Si but has F most important influence.Therefore the noise meeting being unavoidably introduced into signal acquisition process in M Amplified to significantly impact the accuracy of F by minimum singular value.
It is to limit the minimum singular value of selection on the solution that F solving results influence to limit noise, and solution is made to obtain Relaxation Spectral structure F and actual relaxation Spectral structure F0Between difference be on close level with Electronics noice, i.e., by preceding r singular value Determining F and F0Deviation control Electronics noice level.
Using more singular values, that is, smaller singular value is allowed to participate in the F that linear combination obtains, with true F0Deviation meeting Smaller, that is to say, that obtained relaxation spectrum result can more reflect truth, but the amplification effect meeting that Electronics noice influences Bad stability that is more serious, that is, composing.Using less singular value, i.e., only allows larger singular value to participate in linear combination and obtain The F arrived, with true F0Deviation can be larger, that is to say, that obtained relaxation spectrum result is deteriorated with the truth goodness of fit, but Stability can be better.Therefore, it is the pass in SVD methods reasonably to select the number (i.e. the size of minimum singular value) of singular value Key link.
The present invention protects content to be not limited to above example.Under the spirit and scope without departing substantially from present inventive concept, this Field technology personnel it is conceivable that variation and advantage be all included in the present invention, and with the appended claims be protection model It encloses.

Claims (8)

1. a kind of emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology, which is characterized in that the method packet It includes：
Build analog sample information model；
Build relaxation device hardware parameter；
Structure measures sequence physical model；
The analog sample information model, relaxation analysis hardware parameter, measurement sequence physical model based on aforementioned structure, structure NMR relaxation data acquire Physics-mathematics model, complete the analog acquisition of the initial data of the Physics-mathematics model；
The data of previous step institute analog acquisition are subjected to relaxation spectra inversion, obtain the distribution map in relaxation time.
2. emulation nuclear magnetic resonance relaxation spectrum analysis method as described in claim 1, which is characterized in that the analog sample information Model includes principle universal sample information model, typical fields sample message model.
3. emulation nuclear magnetic resonance relaxation spectrum analysis method as claimed in claim 2, which is characterized in that the principle universal sample Information model includes the combination of one or more standard samples.
4. emulation nuclear magnetic resonance relaxation spectrum analysis method as described in claim 1, which is characterized in that the relaxation device hardware Parameter includes main field B0, main field inhomogeneities deltaB0, radiofrequency field B1, gradient fields G.
5. emulation nuclear magnetic resonance relaxation spectrum analysis method as described in claim 1, which is characterized in that the sequence includes IR sequences Row, SR sequences, CPMG sequence, PGSE sequences, PGSE-CPMG sequences, and/or IR-CPMG sequences.
6. emulation nuclear magnetic resonance relaxation spectrum analysis method as described in claim 1, which is characterized in that be based on the analog sample Information model, the measurement sequence physical model and the hardware parameter, the T1 for obtaining each component of the sample relax Henan item, T2 relaxation terms, T2* relaxation decays item, diffusive attenuation item, inversion recovery item, magnetization vector overturning item, rotating coordinate system are just Intersection number according to acquisition item, the one or more and sample scale factor of aforementioned item is carried out it is even multiplied to sample component signal, Add up multiple aforementioned sample constituent signals, obtains ideal signal, then superposed simulation noise, obtains the original number of analog acquisition According to.
7. a kind of emulation nuclear magnetic resonance relaxation spectrum analysis system based on numerical simulation technology, which is characterized in that the system packet It includes：
Analog sample information model builds module, for building analog sample information model；
Relaxation device hardware parameter builds module, for building relaxation equipment related hardware parameter；
Analysis measurement method physical model builds module, for building different measurement sequence physical models；
NMR relaxation data acquire Physics-mathematics model and build module, are used for the analog acquisition of initial data；
Relaxation spectra inversion module is used for relaxation spectra inversion, obtains the distribution map in relaxation time.
8. a kind of purposes of the emulation nuclear magnetic resonance relaxation spectrum analysis system based on numerical simulation technology, which is characterized in that described Application of analysis system composes the mating virtual experimental training platform, instrument development and test school of application technology in NMR relaxation Accurate, the reference of sample test spectrogram, relaxation spectra inversion research data source.
CN201810163795.6A 2018-02-27 2018-02-27 Simulation nuclear magnetic resonance relaxation spectrum analysis method and system based on numerical simulation technology Active CN108520088B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810163795.6A CN108520088B (en) 2018-02-27 2018-02-27 Simulation nuclear magnetic resonance relaxation spectrum analysis method and system based on numerical simulation technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810163795.6A CN108520088B (en) 2018-02-27 2018-02-27 Simulation nuclear magnetic resonance relaxation spectrum analysis method and system based on numerical simulation technology

Publications (2)

Publication Number Publication Date
CN108520088A true CN108520088A (en) 2018-09-11
CN108520088B CN108520088B (en) 2022-04-05

Family

ID=63433329

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810163795.6A Active CN108520088B (en) 2018-02-27 2018-02-27 Simulation nuclear magnetic resonance relaxation spectrum analysis method and system based on numerical simulation technology

Country Status (1)

Country Link
CN (1) CN108520088B (en)

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109598245A (en) * 2018-12-07 2019-04-09 上海理工大学 Edible oil transverse relaxation attenuation curve signal characteristic extracting methods based on 1D-CNN
CN110009608A (en) * 2019-03-26 2019-07-12 沈阳农业大学 A method of the blueberry that rots is differentiated based on low-field nuclear magnetic resonance technology
CN110910965A (en) * 2019-11-15 2020-03-24 东南大学 Method for screening rubber particle surface activation technology based on molecular dynamics
CN111058828A (en) * 2019-11-13 2020-04-24 华东师范大学 Rapid measurement system and method of fluid T1 relaxation spectrum based on surface nuclear magnetic resonance technology
CN111685768A (en) * 2020-06-29 2020-09-22 中国科学院苏州生物医学工程技术研究所 Magnetic resonance T2 quantitative imaging and evaluation method and system for hip dislocation
CN111898734A (en) * 2020-07-10 2020-11-06 中国科学院精密测量科学与技术创新研究院 NMR (nuclear magnetic resonance) relaxation time inversion method based on MLP (Multi-layer linear programming)
CN112305003A (en) * 2019-07-31 2021-02-02 上海纽迈电子科技有限公司 Edible oil analysis model establishment method, edible oil analysis method and edible oil analysis device
CN112305006A (en) * 2020-10-26 2021-02-02 东南大学 Tumor tissue specimen identification method and system based on nuclear magnetic resonance signals
CN112305004A (en) * 2019-08-02 2021-02-02 中国石油化工股份有限公司 Polymer material aggregation state analysis method and system based on time domain nuclear magnetic resonance
WO2022000948A1 (en) * 2020-06-29 2022-01-06 无锡鸣石峻致医疗科技有限公司 Nuclear magnetic resonance system-based substance measurement method, and system
CN115078438A (en) * 2022-06-19 2022-09-20 西南石油大学 Method for establishing pore network model based on nuclear magnetic resonance test digital core
CN117347416A (en) * 2023-10-12 2024-01-05 无锡鸣石峻致医疗科技有限公司 Non-invasive detection method and device for substance components and storage medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010056749A1 (en) * 2008-11-12 2010-05-20 Jang-Yeon Park Short te 3d radial sampling sequence for mri
CN102023174A (en) * 2009-09-22 2011-04-20 吴雪梅 Analysis method of high resolution nuclear magnetic resonance FID signal
US20150260813A1 (en) * 2014-03-14 2015-09-17 Schlumberger Technology Corporation System and method for processing magnetic resonance signals

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2010056749A1 (en) * 2008-11-12 2010-05-20 Jang-Yeon Park Short te 3d radial sampling sequence for mri
CN102023174A (en) * 2009-09-22 2011-04-20 吴雪梅 Analysis method of high resolution nuclear magnetic resonance FID signal
US20150260813A1 (en) * 2014-03-14 2015-09-17 Schlumberger Technology Corporation System and method for processing magnetic resonance signals

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
夏天等: "《虚实结合的MRI实验实训项目开发与实践》", 《实验室研究与探索》 *
张子云，袁广宇，徐晓峰，李世刚，戴鹏编著: "《物理实验教程丛书 近代物理实验》", 30 September 2015, 合肥：中国科学技术大学出版社 *
汪红志等: "《核磁共振成像技术虚拟软件开发》", 《中国医学物理学杂志》 *
钱俊生等编: "《最新科学技术全书 第5册》", 31 May 2001, 北京：线装书局；北京：人民日报出版社 *
陈珊珊等: "《低场核磁共振弛豫谱反演算法研究》", 《生物医学工程学杂志》 *

Cited By (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109598245B (en) * 2018-12-07 2023-02-07 上海理工大学 Edible oil transverse relaxation attenuation curve signal feature extraction method based on 1D-CNN
CN109598245A (en) * 2018-12-07 2019-04-09 上海理工大学 Edible oil transverse relaxation attenuation curve signal characteristic extracting methods based on 1D-CNN
CN110009608A (en) * 2019-03-26 2019-07-12 沈阳农业大学 A method of the blueberry that rots is differentiated based on low-field nuclear magnetic resonance technology
CN110009608B (en) * 2019-03-26 2023-04-07 沈阳农业大学 Method for distinguishing rotten blueberries based on low-field nuclear magnetic resonance technology
CN112305003A (en) * 2019-07-31 2021-02-02 上海纽迈电子科技有限公司 Edible oil analysis model establishment method, edible oil analysis method and edible oil analysis device
CN112305004A (en) * 2019-08-02 2021-02-02 中国石油化工股份有限公司 Polymer material aggregation state analysis method and system based on time domain nuclear magnetic resonance
CN111058828A (en) * 2019-11-13 2020-04-24 华东师范大学 Rapid measurement system and method of fluid T1 relaxation spectrum based on surface nuclear magnetic resonance technology
CN111058828B (en) * 2019-11-13 2023-02-03 华东师范大学 Rapid measurement system and method of fluid T1 relaxation spectrum based on surface nuclear magnetic resonance technology
CN110910965B (en) * 2019-11-15 2023-07-18 东南大学 Method for screening rubber particle surface activation technology based on molecular dynamics
CN110910965A (en) * 2019-11-15 2020-03-24 东南大学 Method for screening rubber particle surface activation technology based on molecular dynamics
US11766189B2 (en) 2020-06-29 2023-09-26 Wuxi Marvel Stone Healthcare Co., Ltd. Nuclear magnetic resonance system-based substance measurement method and system
CN111685768A (en) * 2020-06-29 2020-09-22 中国科学院苏州生物医学工程技术研究所 Magnetic resonance T2 quantitative imaging and evaluation method and system for hip dislocation
WO2022000948A1 (en) * 2020-06-29 2022-01-06 无锡鸣石峻致医疗科技有限公司 Nuclear magnetic resonance system-based substance measurement method, and system
CN111898734A (en) * 2020-07-10 2020-11-06 中国科学院精密测量科学与技术创新研究院 NMR (nuclear magnetic resonance) relaxation time inversion method based on MLP (Multi-layer linear programming)
CN111898734B (en) * 2020-07-10 2023-06-23 中国科学院精密测量科学与技术创新研究院 NMR relaxation time inversion method based on MLP
CN112305006B (en) * 2020-10-26 2023-09-01 东南大学 Tumor tissue specimen identification method and system based on nuclear magnetic resonance signals
CN112305006A (en) * 2020-10-26 2021-02-02 东南大学 Tumor tissue specimen identification method and system based on nuclear magnetic resonance signals
CN115078438A (en) * 2022-06-19 2022-09-20 西南石油大学 Method for establishing pore network model based on nuclear magnetic resonance test digital core
CN117347416A (en) * 2023-10-12 2024-01-05 无锡鸣石峻致医疗科技有限公司 Non-invasive detection method and device for substance components and storage medium
CN117347416B (en) * 2023-10-12 2024-05-31 无锡鸣石峻致医疗科技有限公司 Non-invasive detection method and device for substance components and storage medium

Also Published As

Publication number Publication date
CN108520088B (en) 2022-04-05

Similar Documents

Publication Publication Date Title
CN108520088A (en) Emulation nuclear magnetic resonance relaxation spectrum analysis method based on numerical simulation technology and analysis system
CN105980876B (en) Method for isotropism disperse and/or anisotropy disperse in quantitative sample
CN109270107B (en) Multi-dimensional nuclear magnetic resonance measurement method
JP6339299B1 (en) Magnetic resonance fingerprinting with reduced sensitivity to main magnetic field inhomogeneities
CN107144591B (en) A method of measurement is independently with nuclear spin to indirect coupling mode
CN113466280B (en) Simulated nuclear magnetic resonance spectrum analysis method and system convenient for expanding molecular information base and application thereof
RU2693837C2 (en) Magnetic-resonance fingerprint method
Zheng et al. Peculiar velocity decomposition, redshift space distortion, and velocity reconstruction<? format?> in redshift surveys. II. Dark matter velocity statistics
CN108027409A (en) Time domain mri
CN105334239A (en) Multi-dimensional nuclear magnetic resonance fluid component content measuring method and device
CN105259198B (en) The measurement method in two dimensional NMR relaxation time
Vogel et al. Effects of various types of molecular dynamics on 1D and 2D 2H NMR studied by random walk simulations
AU2003271404A1 (en) Multi-measurement NMR analysis based on maximum entropy
Parasram et al. T2 analysis using artificial neural networks
CN113743682B (en) NMR (nuclear magnetic resonance) relaxation time inversion method based on supervised deep neural network
Starčuk et al. Simulation of coupled-spin systems in the steady-state free-precession acquisition mode for fast magnetic resonance (MR) spectroscopic imaging
CN107728088B (en) A kind of quick nuclear magnetic resonance T1Imaging method
CN110187297A (en) A kind of low-field nuclear magnetic resonance relaxation detection method inhibiting signal specific
CN105486709A (en) Measuring method and device for underground thick oil molecule chain length
CN116106354A (en) Novel multidimensional nuclear magnetic resonance T1-T2 imaging method
Lorigan et al. Teaching the fundamentals of pulsed NMR spectroscopy in an undergraduate physical chemistry laboratory
Antosia Ground Wave Propagation Measurement using DAQ Module and LabVIEW
Chapman et al. Some consequences of the shift theorem for multispacecraft measurements
WANG et al. MRISim: A Software Package of Magnetic Resonance Imaging Simulator
Moon Rotational Spectroscopy: A Laboratory for Undergraduate Physical Chemistry

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant