KR20190063395A - Brain metabolite network generation method using time varying function based on MRS - Google Patents

Abstract

The present invention relates to a method for generating a brain metabolite network using a time-varying function for magnetic resonance spectroscopy-based brain metabolites, which, for monitoring the brain metabolite network, detects spectrum at certain time intervals by using magnetic resonance spectroscopy (MRS) equipment for specific regions of the brain, obtains the time-varying function of the brain metabolite in which the amount of metabolite from the detected spectrum shows a quantitative change over time, and configures a brain metabolite network by using the obtained time-varying function. The present invention can contribute to revealing the interrelationship of metabolites between regions of the brain through a correlation coefficient. In addition, two regions connected in the brain functional network which connects the regions of the brain performing the same function speculates that an amount of a metabolite performing a specific function in a corresponding region coincidentally increases and decreases through the brain metabolite network of the present invention. Accordingly, it is possible to identify how the two regions can interact and through which mechanism the two regions are functionally related as well as identifying a functional association of the two regions. Therefore, the identified results can provide precise and accurate information to the development of diagnosis and treatment method of brain diseases, and also can be used as an indicator to monitor and evaluate a treatment process.

Description

Technical Field [0001] The present invention relates to a brain metabolite network generation system using a time-varying function of a brain metabolite based on magnetic resonance spectroscopy,

The present invention relates to a method for monitoring brain metabolism by detecting a spectrum at a predetermined time interval using a magnetic resonance spectroscopy (MRS) apparatus for a specific region of the brain and measuring the amount of metabolites from the detected spectrum Obtaining the time-varying function of the brain metabolite that exhibits a quantitative change over time, and using the time-varying function of the magnetic resonance spectroscopy-based brain metabolite to construct the brain metabolite network using the acquired time- Metabolic network generation system and method.

MRS uses nuclear magnetic resonance (NMR) as a clinical technique because it can observe brain metabolites in a non-invasive way without using radiation. When an external magnetic field is applied to an object, the spin existing in the object carries out the car motion with the carburizing frequency depending on the intensity of the substance and the magnetic field. At this time, when the electromagnetic wave having the same frequency as the carburizing frequency is applied to the spin, A resonance phenomenon occurs. This phenomenon is referred to as nuclear magnetic resonance (NMR). Induced electromotive force is generated by this phenomenon. The spectroscopic technique by measuring induced electromotive force is magnetic resonance spectroscopy. This technique is used to reveal the structure and morphology of proteins or macromolecules and to obtain the biochemical information of metabolites, the chemical properties of tissues present in organs of the human body. Clinically, by observing the biochemical changes of the brain metabolism, it is used to prevent and diagnose diseases such as cancer and epilepsy by detecting the changes of the cell tissues in advance.

In a clinical study of conventional magnetic resonance spectroscopy (MRS), information on the amount of metabolites in the brain contained in the spectrum obtained at one point was obtained and quantified to determine the amount of specific metabolites in the normal and control groups Are compared and observed. However, even in the same disease or condition, the rate and cycle of metabolism change is different for each person. Therefore, when comparing the amount of metabolites only at a certain point in time, it is impossible to observe the dynamic change of the metabolites. The resultant value varies depending on the type of the apparatus.

Therefore, rather than using one spectrum at a time, we measured the amount of metabolites at each time point by obtaining the spectrum at each time point by measuring several times at a certain time interval, time variing function, which allows a more precise result to be obtained by comparing the entire cycle of metabolism.

Accordingly, the present invention is directed to: Observing the spectrum at certain time intervals using magnetic resonance spectroscopy equipment for a specific region of the brain, obtaining a time-varying function of the brain metabolite that exhibits a quantitative change over time of the amount of metabolite from the detected spectrum, We propose a method of generating a brain metabolic network by using a time-varying function for magnetic resonance spectroscopy-based brain metabolites, which constitutes a brain metabolism network using a time-varying function.

The present invention can be used for various researches and analyzes by constructing a network by calculating correlation coefficients between regions with a time-varying function for metabolites in each region.

In the prior art, Korean Patent Laid-Open No. 10-2011-0045618 relates to a method of diagnosing Parkinson's disease and a method of screening a therapeutic agent for Parkinson's disease by measuring quantitative change of brain metabolism using magnetic resonance spectroscopy, Obtain information on the amount of metabolites and quantify them. By comparing the amounts of specific metabolites in the normal group and the Parkinson's disease-induced group, No dynamic changes can be observed.

SUMMARY OF THE INVENTION The object of the present invention is to provide a method for monitoring brain metabolism by detecting a spectrum at a predetermined time interval using a magnetic resonance spectroscopy (MRS) apparatus for a specific region of the brain, To obtain a time-varying function of a brain metabolite that exhibits a quantitative change over time and to construct a cerebral metabolic network using the acquired time-varying function. Method.

Another problem to be solved by the present invention is to provide a seed-based correlation analysis (SCA) method for identifying a specific voxel or region as a seed and showing a high correlation with the voxel or region, Independent component analysis (ICA) is a method of constructing a network of brain metabolites by constructing and evaluating networks through the application of graph theory. And a method for generating a brain metabolite network using the time-varying function.

In order to solve the above problems, the present invention provides a method for monitoring brain metabolism, comprising the steps of: receiving, from a magnetic resonance spectroscopy (MRS) instrument, step; A spectral time-varying function acquiring step of acquiring a time-varying function of a cerebral metabolism from the spectrum received in the spectral receiving step, the amount of metabolite showing a quantitative change with time; The operation processing unit is configured to generate a brain metabolism network using the time-varying function acquired in the time-varying function acquiring step.

The spectral image received by the arithmetic processing unit in the spectral receiving step is a multi-voxel magnetic resonance spectrum (MRS) obtained by using a 2D MR spectroscopy (MRSI) technique or a 3D MRSI technique.

 The spectrum represents the degree of chemical shift according to the amount of metabolite, and the chemical shift (?

(Where f is the Lamor frequency of the hydrogen nucleus in the metabolite and f _ref is the frequency of tetramethylsilane (TMS) or 4-dimethyl-4-silapentane-1-sulfonic acid (DSS)).

In this spectrum, the chemical shift value at the center of the peak indicates the metabolite, and the width of the peak indicates the amount of the metabolite.

The spectrum is measured using a magnetic resonance spectroscopy (MRS) apparatus in a phantom solution containing a plurality of metabolic materials including alanine (Alanine, Ala) and each metabolite alone. And generates basis set data. For the basis set data, spectra were measured by making a phantom solution containing only each metabolite for 18 brain metabolites, and the 18 brain metabolites were analyzed for alanine (Alanine, Ala), aspartate (Aspartate, Asp), Creatine (Cr), Phosphocreatine (PCr), Gamma-Aminobutyric acid (GABA), Glucose, Glc, Glutamate Glutamine, Gln, Glutathione (GSH), Glycerophosphocholine (GPC), Phosphocholine, PCho, Myo-Inositol, Ins, Lactate (Lsc) , N-Acetylaspartate (NAA), N-Acetylaspartylglutamate (NAAG), Phosphocreatine (PCr), Phosphorylethanolamine (PE), Taurine (Taurine, Tau).

In the spectrum time-varying function acquisition step, 2D magnetic resonance spectrum (MRS) signal measured time interval for obtaining a time-series signal, the time required to measure once, TR × N _x × N _y × NEX (where, TR is N _x is the number of voxels arranged in the transverse direction of the multi-voxel to obtain the spectrum in 2D MRSI and N _y is the length of the multi-voxel to obtain the spectrum in 2D MRSI, And NEX is the number of times the experiment is repeated under the same condition).

In the network configuration step, for the network configuration, the arithmetic processing unit determines a specific voxel or region as a seed and then identifies a region showing a high correlation with the voxel or region (seed-based correlation analysis; SCA) can be used.

When the MRSI data are measured at n different times (t1, t2, ..., t10), the amounts of each metabolite are obtained in each voxel at each time, and as a result, the amount of each metabolite per voxel is n And expressed as a time-variant amount of data.

,

로 하고, 시변량의 표준편차를 s_x,s_y 라고 하면, 두 복셀의 상관 계수(r)는 In the network construction step, when the SCA method is used, the time-varying amounts of the metabolite amounts in the two voxels are x and y, respectively, and the amounts of metabolites measured at the specific time t _i are x _i and y _i , Average of

,

, And the standard deviation of the time variant is s _x , s _y , the correlation coefficient (r) of the two voxels is

The correlation coefficient r has a value between -1 and 1, and the arithmetic processing unit sets the correlation coefficient r to a predetermined threshold _rthresh ). It is determined that there is a positive correlation when the correlation coefficient r is large and it is determined that there is a negative correlation if the correlation coefficient r is small and when the correlation coefficient r is large, Or is not small, it is judged that there is no correlation.

The arithmetic processing unit displays the correlation coefficient as a standardized z value used in statistical analysis, encrypts it with a color and displays it on an anatomical image, and the two areas in which the correlation coefficient is obtained are connected to an edge of the network And edge colors when they are connected in a positive correlation and negative colors when they are connected in a negative correlation are different colors. The arithmetic processing unit may represent a red edge when connected by a positive correlation and a blue edge when connected by a negative correlation.

At the network configuration stage, a graph theory approach can be used for network configuration.

A network consists of a node and an edge.

In the network configuration step, when a graph theory approach is used, whether a connection between nodes is represented by a connectivity matrix (C), a connection matrix of a brain network composed of N nodes forms an N × N matrix,

The i row j column element c _ij of the connection matrix is a value indicating the connectivity between the i-th node and the j-th node.

In the case of a binary network, the connection matrix c _ij

(Where r is the correlation coefficient and _rthresh Is a threshold value of the correlation coefficient). When the connection matrix value is 0, the two nodes are not connected to each other. If the connection matrix value is not 0, they are regarded as being connected to each other. Express.

In the case of a weighted network, the connection matrix c _ij is

When the connection matrix value is 0, the two nodes are not connected to each other, and when the connection matrix value is not 0, they are regarded as connected to each other, and the two nodes are connected to each other by an edge.

In the network configuration step, the arithmetic processing unit divides the brain functionally into a plurality of different areas, expresses each area as a node, calculates a correlation coefficient r between the nodes to construct a connection matrix, Is a gray level, and a network composed of connection matrices represented by gray levels is called a weighted network. In the weighted network, when r = 0, the two nodes do not connect to the edge. If r ≠ 0, Edge, and is configured to determine the thickness of the edge in proportion to the value of r.

The node degree, the shortest path length, the centrality, and the clustering coefficient are used to analyze the network formed between the metabolites in the brain through the network constructed in the network configuration step. Lt; / RTI > The network parameters include a number of triangles around a node, an average distance between nodes and all other nodes, a length of a characteristic path, the efficiency of a particular node, the global efficiency, the clustering coefficient of a node, the local efficiency of a node, efficiency.

The present invention also provides a recording medium storing a computer program source for a method for generating a brain metabolism network.

The method for generating a cerebral metabolic network using a time-varying function of a magnetic resonance spectroscopy-based brain metabolism of the present invention is a method for monitoring brain metabolism using a magnetic resonance spectroscopy (MRS) Acquiring a time-varying function of a brain metabolite that exhibits a quantitative change in the amount of metabolite over time from the detected spectrum, and constructing a brain metabolism network using the acquired time-varying function. This is because, rather than using one spectrum at a time, the spectra are measured at a predetermined time interval, and the quantities of the metabolites are quantified at each time point, By comparing these, it is possible to obtain more accurate results by comparing the total change period of the metabolites.

The present invention also relates to a method for identifying a specific voxel or region as a seed and then identifying a region showing a high correlation with the voxel or region thereof (seed-based correlation analysis (SCA) (ICA), and constructing and evaluating networks through the application of graph theory, constitute the brain metabolism network.

The present invention can contribute to revealing the correlation of metabolites between regions of the brain through a correlation coefficient. In addition, two regions connected by a brain functional network connecting brain regions of the same function are connected to each other through the brain metabolite network of the present invention, It is possible to identify not only the functional relationship between the two domains, but also how they are interacted and functionally related through which mechanism. It can also be used to provide more accurate and precise information on the diagnosis and treatment of brain diseases, and can be used as an indicator to monitor and evaluate the treatment process.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram schematically illustrating the configuration of a system for generating a brain metabolism network using a time-varying function for a magnetic resonance spectroscopy-based brain metabolism of the present invention. FIG.
FIG. 2 is an explanatory diagram for explaining spectra in various voxels of the brain obtained through a general 2D magnetic resonance spectroscopy (MRSI) technique.
FIG. 3 is an explanatory diagram for explaining the number of voxels according to a setting in a multi-voxel to obtain a spectrum in 2D MRSI in general.
4 is a schematic diagram for explaining a 2D MRSI pulse train in general.
5 is a schematic diagram for explaining a 3D MRSI pulse train in general.
6 is a schematic diagram for explaining the gradient magnetic fields Gx and Gy in Fig.
FIG. 7 is an explanatory diagram for explaining signal acquisition from the time when an echo signal is generated.
Figure 8 is an example of a single voxel MR spectrum measured in the hippocampal region of a mouse.
FIG. 9 is an example of a time-varying graph of NAA measured by performing multi-voxel (3 × 3) spectral experiments at 10 different viewing angles.
10 is a schematic diagram for explaining that a network is composed of nodes and edges.
11 is an explanatory view illustrating a process of dividing a brain functionally into different regions and representing the brain with nodes and edges.
12 is an explanatory diagram illustrating a network configuration of a brain having twelve nodes.
13 shows each area of the AAL template.

Hereinafter, a method for generating a cerebral metabolic network using a time-varying function of a magnetic resonance spectroscopy-based brain metabolism of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram schematically illustrating the configuration of a system for generating a cerebral metabolic network using a time-varying function for a magnetic resonance spectroscopy-based brain metabolism of the present invention, a magnetic resonance spectroscopy (MRS) apparatus 100, An operation unit 200, a data base 250, a key input unit 310, and an output unit 320. The operation unit 200 includes a display unit 110,

The magnetic resonance spectroscopy (MRS) apparatus 100 detects spectra at predetermined time intervals for a specific region of the brain.

The signal preprocessing unit 110 performs signal preprocessing for removing noise from a signal received from the MRS equipment 100 and converting it into a digital signal.

The arithmetic processing unit 200 obtains a time-varying function of a cerebral metabolite showing a quantitative change with time in the amount of metabolites from the spectrum received from the signal preprocessing unit 110, (brain metabolite network), and brain metabolism network-related data is stored in a database (DB) 250 and outputted through an output unit 320. [ The operation processing unit 200 may include a computer, a microprocessor, and the like.

Here, the key input unit 310 is necessary for setting parameters, a video slide direction, and the like.

The operation processing unit 200 generates a cerebral metabolic network through a process of acquiring a spectrum, acquiring a spectral time-varying function, and configuring a network. Therefore, these steps will be described below.

≪ Spectrum acquisition step &

First, the method of acquiring the spectrum through the MRS acquisition apparatus includes a single voxel MRS technique and a multi-voxel MRS technique. The present invention focuses on the multi-voxel MRS method because it is necessary to obtain spectra of various regions simultaneously. Multi-voxel spectra are obtained through two-dimensional (2D) and three-dimensional (3D) multivoxel MRS techniques. In general, the multivoxel MRS technique is named and used as a chemical shift imaging (CSI) technique or a magnetic resonance spectral imaging (MRSI) technique. In the present invention, MRSI technique is used for unifying for convenience of explanation.

Generally, the spectra obtained by the 2D MRSI technique are measured for each voxel in several voxels of the brain, as shown in FIG.

In the present invention, spectra are obtained using 2D MRSI and 3D MRSI techniques. Since the multivoxel 11 to acquire the spectrum by the 2D MRSI (Magnetic Resonance Spectrum Image) technique is set in a lattice shape in which N _x pieces are arranged in N _y horizontal directions in the vertical direction (y) as shown in FIG. 3, Is _Ny x N _x .

In FIG. 3, after setting the MRS to be obtained in the 16 × 16 voxel in the yellow square (ie, the multivoxel 11 to acquire the spectrum), the area in the blue square (ie, the area 12 where the high resolution spectrum is to be obtained) Only the spectrum of the 10 × 8 voxel in the blue line region, which can obtain the high-resolution spectrum of the MR spectrum of the entire 16 × 16 voxel, is shown. That is, FIG. 3 shows an example in which a region 11 in which a magnetic resonance spectrum (MRS) is to be measured by 2D MRSI (i.e., in a yellow square in FIG. 3) is set to a 16x16 voxel to acquire a magnetic resonance spectrum And a region 12 for obtaining a high-resolution spectrum (that is, in a blue square in FIG. 3) of the 16 × 16 voxels to be measured for the magnetic resonance spectrum (MRS) The region 12 in which the voxel is to be set to obtain the high-resolution spectrum is subjected to precision shimming, which is a process of correcting a non-uniform magnetic field to a uniform magnetic field, to show only a high-resolution spectrum.

The multiple voxel spectra are acquired with a chemical shift imaging (CSI) pulse train or a magnetic resonance spectral imaging (MRSI) pulse train. The MRSI pulse sequence can be referred to as the term shown in FIG. 3 as a whole.

Fig. 4 is a schematic diagram for explaining a 2D MRSI pulse train, and Fig. 5 is a schematic diagram for explaining a 3D MRSI pulse train.

In FIG. 4, three RF pulses of 90 °, 180 ° and 180 ° (that is, 90 ° - 180 ° - 180 °) are used to form a region 11 to be measured for the 2D magnetic resonance spectrum (MRS) ). The first 90-degree pulse is 2D magnetic select resonance spectrum (MRS) region (11) z-direction position and the thickness (△ _z) of the voxels (that is, a yellow square area) to measure the signal, and the first 180 ° pulse is y Select the directional area, and the second 180 degrees selects the x-directional area. On the other hand, since the x region and the y region are wide regions, it is necessary to divide the region into multiple voxels in order to increase the spatial resolution of the spectrum. For this purpose, a x-gradient magnetic field (Gx) and a Y-gradient magnetic field (y-gradient, Gy), which are phase encoding gradient magnetic fields (arrows), are applied. In the case of Figure 3 it can be a number of 265 pieces of 16 × 16 because G _x and G _y G _x and G _y of each gradient magnetic field to set the size (gradient amplitude) of step 16 to gatdo combination of voxels. Therefore, in the MRSI pulse (H _{RF in} FIG. 4), 256 combinations of G _x and G _y are sequentially applied to obtain a magnetic resonance image signal. The MR signal acquisition _{(SK (k x, k y} ; t)) is given by the equation (1) and equation (2).

Here, k _x and k _y represent the x and y coordinates in the two-dimensional k- space, t denotes a time interval, the MR signal is _{detected, SK (k x, k y} ; t) is a k- (MR) image signal in space, and S (x, y; t) denotes a magnetic resonance (MR) image signal in time space. FT _2D {S (x, y; t)} denotes a value obtained by Fourier transforming an image sample having an x coordinate and a y coordinate detected during a time interval t. Generally, a magnetic resonance imaging signal is placed on a measurement space (which is referred to as a frequency space or a remainder space or k-space). In other words, Equation 1 represents a magnetic resonance (MR) image signal in a two-dimensional k-space, and Equation 1 is transformed into a magnetic resonance (MR) image signal at a position (x, y) of each voxel in an image of a time space through a 2D- The spectral image signal S (x, y; t) is obtained as shown in Equation (2).

Where TR is the repetition time of the MRSI pulse corresponding to the respective carburized frequency (f _L ), TE is the echo time of the MRSI pulse corresponding to each car wash frequency (f _L ), and n _L is the location T _{1 ℓ} , T _{2 ℓ} , T ^* _{2 ℓ} is the relaxation time (T _{1 ℓ} ) of the vertical magnetization of the hydrogen nuclei having the voxel wavenumber frequency f _ℓ , spin-spin The relaxation time (T _2ℓ ) of the horizontal magnetization due to the interaction, and the horizontal magnetization relaxation time (T ^* _2ℓ ) due to the magnetic field unevenness. In addition, C _ℓ (x, y) are (x, y) represents a density (proton density) of hydrogen atomic nucleus A has a wash frequency f _ℓ within voxel organization centered on (x, y) is SK (k _x , k _y ; t), and does not depend on the respective car wash frequencies of the hydrogen nuclei.

On the other hand, k _x and k _y are determined from the relational expression as shown in Equation (3) and represent one point in the two-dimensional k-space.

는 자기 회전 비(磁氣 回轉 比)이고, Gx(τ)는 시간 τ에서 인가된 x축 경사자장의 크기를 말하고, Gy(τ)는 시간 τ에서 인가된 y축 경사자장의 크기를 말한다. 일반적으로 자기회전비는 핵의 종류에 따라지는 상수로, 수소원자핵의 경우,

=2π

42.577 MHz/T 이다.here

Is the magnitude of the y-axis gradient magnetic field applied at time τ, Gx (τ) is the magnitude of the x-axis gradient magnetic field applied at time τ, and Gy (τ) is the magnitude of the y-axis gradient magnetic field applied at time τ. In general, the rate of magnetic rotation depends on the type of nucleus. In the case of hydrogen nuclei,

= 2π

42.577 MHz / T.

6 is a schematic diagram for explaining the gradient magnetic fields Gx and Gy in Fig. The region 15 surrounded by the red dotted line in FIG. 4 is a rhomboid shape, but it is not a rhombus shape but a trapezoidal shape as shown in FIGS. 6 (a) and 6 (b). This is well known in the art. To explain this part, too much explanation is required. Furthermore, this part is not related to the present invention, and a detailed description thereof will be omitted.

The gradient magnetic fields Gx and Gy depicted in the area 15 surrounded by the red dotted line in FIG. 4 can be represented as shown in FIG. 6, and k _x and k _y are given by Equation (4).

In Equation (4), τ ₂ , τ _1x , and τ _1y are time variables, and G _x and G _y represent the magnitude of the gradient of the oblique magnetic field. That is, G _x represents the magnitude of the set x-axis gradient magnetic field, and G _y represents the magnitude of the y-axis gradient magnetic field. Also, τ ₂ is a time interval at which the x-axis oblique magnetic field is applied and a time interval at which the y-axis oblique magnetic field is applied, and the time interval at which the x-axis oblique magnetic field is applied is equal to the time interval at which the y- τ _1x represents a time interval during which the x-axis oblique magnetic field is equal to or larger than the predetermined x-axis oblique magnetic field size, and τ _1y represents the time interval during which the y-axis oblique magnetic field is greater than or equal to the predetermined y- Lt; / RTI >

Gx and Gy are applied between the 90-degree RF pulse and the first 180-degree RF pulse in the MRSI pulse train, and the measurement of the signal measures the echo signal generated after the second 180-degree pulse is applied as a magnetic resonance spectrum (MRS) .

FIG. 7 is an explanatory diagram for explaining signal acquisition from the time when an echo signal is generated, and 2048 times of signals are acquired at predetermined time intervals from the time when the echo signal is generated. 7, an echo signal is generated at the maximum point of the MRS signal (MRS signal), and the signal is obtained 2048 times from the time when the echo signal is generated, which is understood to limit the present invention. In other words, in the case of FIG. 7, the signal measurement (i.e., sampling) is obtained 2048 times (this value is selectable) in time intervals Δt (= 1 / BW, where BW = receive bandwidth). Here, Δt denotes a sampling interval, and a receive bandwidth denotes a frequency width at which an MRI signal is received, and is given as an input parameter when acquiring an MRI image. Therefore, the 2408 data measured at intervals of Δt for the signal SK (k _x , k _y ; t) having a k value corresponding to one point (k _x , k _y ) of a given k- .

The 2048 MRI signals (MR signal) signals stored in each 16 × 16 pixels of the two-dimensional k-space are transformed into 16 × 16 pixels of the image space (ie, the image of the time space) through the 2D- (X, y) of each of the 16 voxels as shown in Equation (6). That is, Equation (6) differs from Equation (2) by? T instead of t.

This data is obtained by measuring the magnetic resonance spectrum (MRS) signal generated at each pixel of the image space according to 2048 time intervals. Therefore, the magnetic resonance (MR) spectrum in each voxel can be obtained by one-dimensional Fourier transform of 2048 data according to the time of the voxel.

Where SF (x, y; f) is the Fourier transform of the time of S (x, y; t)

Figure 8 is a single voxel MR spectrum obtained from the hippocampal area of the mouse brain in a 9.4T MRS instrument. The horizontal axis of the spectrum represents the degree of chemical shift, the unit is ppm (percent per million, ppm), and the vertical axis represents the amount of metabolite (mM). The chemical shift (δ) is defined as Equation (8), which represents the degree to which the hydrogen atom in each metabolite varies in the Larmor frequency due to the chemical structure of the surrounding atoms.

In equation (8), f is the Lamor frequency of the hydrogen nucleus in the metabolite, f _ref is the reference material TMS (tetramethylsilane) or DSS (4-dimethyl-4-silapentane-1-sulfonic acid) . In the present invention, DSS was used.

This chemical shift is determined irrespective of the strength of the magnetic field and represents the intrinsic properties of the material. Therefore, the chemical shift value at the center of the peak in the spectrum represents the metabolite (ie, the type of dash material) and the width of the peak represents the amount of metabolite. Therefore, spectrum analysis can measure how much metabolite is present in the voxel.

Spectrum analysis was carried out using a commercially available MRS analysis program, LC Model, using a phantom solution containing only each metabolite for 18 brain metabolites, measuring the spectrum and using it as a basis set.

The 18 metabolites used in the analysis are as follows.

Aminobutyric acid (GABA), Glucose (Glucose), Glucose (Glucose), Glucose (Glucose), Glucose Gloc, Glc, Glutamate, Gln, Glutamine, Glutathione, GSH, Glycerophosphocholine, Phosphocholine, PCho, Myo-Inositol , Ins), Lactate (Lsc), N-Acetylaspartate (NAA), N-Acetylaspartylglutamate (NAAG), Phosphocreatine (PCr) Phosphorylethanolamine (PE), taurine (Taurine)

On the other hand, the time taken to measure the 2D MRS signal is TR x N _x x N _y . However, since the signal-to-noise ratio (SNR) of the acquired signal is not sufficient to precisely analyze the spectrum, the experiment is repeated under the same conditions as the number of excitations (NEX) do. The number of excursions (NEX) is to be input in advance to the MRSI experimental apparatus when it is desired to acquire data. Therefore, a signal acquisition time of the 2D MRS spectrum through MRSI experiment _{TR × N x × N y ×} NEX, 3D MRSI is to omit the detailed formula and described, since when expanded dimension to 2D MRSI in the cross direction (z-direction) I will.

<Acquisition of Spectrum Time-Varying Function>

Signal measurement time interval for obtaining a time-series signal of the 2D MRS should be a time or more _{TR × N x × N y ×} NEX required to measure a time.

A 2D MR spectrum obtained during TR x N _x x N _y x NEX is obtained via 2D MRSI. Ideally, TR x N _x x N _y x NEX should be very short, but in practice it takes more than 10 minutes. Therefore, a single 2D MR spectrum is not obtained at a specific time of measurement but is obtained from a signal acquired over 10 minutes, which is an average value of MR spectra occurring for 10 minutes. Therefore, to obtain the time series signal, the time interval (△ TT) for measuring time series data and the number of time series data (NT) must be determined. For this purpose, the minimum time required to acquire a 2D MRS spectrum suitable for spectrum analysis was determined through experiments, by optimizing 2D MRSI. The analysis of the time series data of a single voxel spectrum was used to measure the change with time of the spectrum TT is determined to be suitable for achieving the desired performance, and NT is determined so as to satisfy the minimum number of time series data required to obtain the correlation of the voxels obtained through the time series data.

<Network configuration step>

There are two network configuration methods to be used in the present invention.

The first is a seed-based correlation analysis (SCA) that identifies a particular voxel or region as a seed and then shows a high correlation with the voxel or region. In this method, a network is constructed by selecting a seed region (ROI) and then correlating with the remaining ROIs. In this case, the ROI (ROI) is set to be the smallest one voxel and the voxel to be largely set. The correlation between the two ROIs is determined by the change in the amount of the specific metabolite amount Can be known by comparing them.

When MRS data is measured in multiple voxels using MRSI, the amount of each metabolite in each voxel within the ROI set to obtain MRSI is quantified. Therefore, when MRSI data is measured at n different times (t1, t2, ..., t10), the amounts of each metabolite are obtained in each voxel at each time. As a result, the amount of each metabolite per voxel is expressed as a time-variant of n data.

Figure 9 shows the results of a multivoxel (3x3) MRS experiment at 10 times t ₁ , t ₂ , ... , t ₁₀ , and then the amount of NAA (N-Acetylaspartate) in the various metabolites was measured to determine the time-varying amount.

,

로 하고, 시변량의 표준편차를 s_x,s_y 라고 하면, 상관 계수(r)는 수학식 9와 같이 계산된다.In order to construct a network for each metabolite, a correlation matrix is required between two different voxels. In the present invention, a commonly used correlation coefficient (r) is used. Let x and y be the time-varying amounts of the metabolites in the two voxels, respectively, and the amount of metabolites measured at the specific time t _i is called x _i , y _i ,

,

, And the standard deviation of the time-varying amount is s _x , s _y , the correlation coefficient r is calculated as shown in Equation (9).

Where r is the correlation coefficient between -1 and 1. In order to determine whether the two ROIs are interconnected through the correlation coefficient r, the correlation coefficient threshold value _rthresh (> 0) to set r> r _thresh , It is regarded that there is a positive correlation when r <-r _thresh , and it is regarded as having a negative correlation when r <-r _thresh , and it is regarded as having no correlation when | r | <r _thresh . Two areas that are correlated by these criteria are considered to be linked to the metabolites. In the statistical analysis, the z value that normalized the correlation coefficient is used, and it is expressed in color and shown in the anatomical image. For example, if the z value is a positive number, it is represented in red. The larger the positive value, the more the color can be represented by a mixture of red and white. Here, the standardized z value refers to a value obtained by converting the correlation coefficient into the Z score according to the Fisher z transformation.

As will be described later, the two regions connected to the network are connected to the edge. In this case, they are represented as a red edge when they are connected by a positive correlation, and a blue edge when they are connected by a negative correlation do.

The second method is a graph theory approach in which graph theory is applied to the brain's network. The network consists of a node and an edge, as in Fig.

The brain is functionally divided into different regions and represented by nodes and edges. This is because data in which a region for the brain is set in advance is stored in the DB 250, and the operation processing unit 200, According to the data, the brain region is divided into the brain image input to the operation processing unit 200 and represented by nodes and edges. At this time, as the data for setting the area for the brain, an internationally-known AAL template image that distinguishes the brain functionally is used. When each brain of a person is normalized using the AAL template, the individual brain image is automatically and functionally separated because each position of the brain image and the AAL template corresponds 1: 1, and one functionally separated region One node. That is, if the brain is functionally divided into 150, the number of nodes becomes 150.

11 (a) and 11 (b) are explanatory diagrams for explaining the process of dividing the brain functionally into different regions and representing the process with nodes and edges, FIG. 11 (a) FIG. 11 (c) is an image in which the functional area is represented by a node in FIG. 11 (b), FIG. 11 (d) It is a network image that expresses whether or not it is edge.

As shown in FIG. 11, for the network configuration, the brain is functionally divided into different regions, each region is expressed as a node, and two nodes connected to each other are represented by an edge.

Connectivity between nodes is represented by a connectivity matrix (C). The connection matrix of the brain network consisting of N nodes is represented by an N × N matrix as shown in Equation (10). Here, each element of the connection matrix C represents the degree of connectivity between nodes. Each node is given a number.

The i row j column element c _ij of the connection matrix is given as a value indicating the connectivity between the i-th node and the j-th node (hereinafter referred to as 'inter-node connectivity value'). This value depends on the type of network. In this case, each element (c _ij ), that is, the inter-node connectivity value of the connection matrix, is obtained by obtaining a correlation coefficient between the node and the node, and the absolute value of each correlation coefficient thus obtained is set to a predetermined correlation coefficient threshold value r _thresh Is expressed as 1 (correlated), and when it is smaller than the correlation coefficient threshold _rthresh , it is represented as 0 (no correlation).

In the case of a binary network, c _ij is given by Equation (11).

If there is a correlation, in order to have a weight according to the degree of correlation, a predetermined correlation coefficient threshold _rthresh The correlation coefficient value at that time is represented by the inter-node connectivity value.

In the case of a weighted network, c _ij is given by Equation (12).

In this case, when the connection matrix value is 0, the two nodes are not connected to each other. If the connection matrix value is not 0, they are considered to be connected to each other.

FIG. 12 is an explanatory view for explaining a network configuration of a 12-node brain. FIG. 12 (a) is a functional block diagram of a brain, which is divided into 12 different areas, 12 (b) shows a connection matrix calculated by correlation coefficients obtained by calculating correlation coefficients between nodes in FIG. 12 (a), and FIG. 12 (c) 12 (d) shows a reconfigured connection matrix by setting a threshold value to a weight obtained in FIG. 12 (c), and FIG. 12 (e) 12 (f) shows a binary connection matrix constructed in the connection matrix of FIG. 12 (d), and FIG. 12 (g) shows a weighting network constructed by using the connection matrix of FIG. Is a binary network constructed by using a binary matrix of

That is, 12 nodes are distributed as shown in FIG. 12 (a). The correlation coefficient r between these nodes is calculated to construct a connection matrix, and this matrix value is expressed in gray level as shown in FIG. 12 (b). The network composed of the connection matrix expressed as above is called a weighted network, and is expressed as (c) in FIG. In this weighted network, if r = 0, the two nodes are not connected to the edge. If r ≠ 0, they are connected to the edge, and the thickness of the edge is determined in proportion to the r value. In order to increase the probability of connectivity between nodes in the weighted network, if r < r _thresh , r = 0, the connection matrix is expressed as (d) in FIG. 12, . 12 (d) is binarized to obtain a binary connection matrix as shown in FIG. 12 (f), and a binary network as shown in FIG. 12 (g) can be constructed using the binary connection matrix.

The following network parameters are used to analyze the network formed between metabolites in the brain via the network thus constructed. Network parameters are divided into global parameters and local parameters. We can evaluate the network structure and characteristics by calculating characteristics such as node degree, shortest path length, centrality, and clustering coefficient as global variables. These network parameters were analyzed using network parameters defined in 2010 (Rubinov M and Sporns O., NeuroImage, 2010, "Complex network measures of brain connectivity: uses and interpretations" (Rubinov M, Sporns O., 2010, Complex network measures of brain connectivity: uses and interpretations. NeuroImage 52: 1059-1069)

The degree of a node (k _i ) of the node is obtained by Equation (13). That is, the degree of the node can be said to be the sum of the values of connectivity between nodes.

However, SN is a set of all nodes.

The shortest path length (d _ij ) is obtained by the following equation (14).

는 노드 i와 j사이의 최단거리를 말한다. 여기서 Cuv는 노드 i와 노드 j사이에 있는 노드들의 노드간 연결성값이며, 최단 경로 길이 (d_ij)는 노드 i와 노드 j사이에 있는 노드들의 노드간 연결성값 들의 합이라 할 수 있다.here

Is the shortest distance between nodes i and j. Where Cuv is the inter-node connectivity of the nodes between node i and node j, and the shortest path length (d _ij ) is the sum of the inter-node connectivity values of the nodes between node i and node j.

The number of triangles around a node (t _i ) around the node _i is found by Equation (15).

When there are N nodes having the connectivity value between node i and node i, the connectivity value (cij) between node i and node j and the node i (Cih) of the node h and the value of the connectivity value (cih) between the node j and the node h are the values of one triangle network. By substituting the values of j and h, As a result, it is applied to all nodes having connectivity value between node i and node i, and the value of the triangle is calculated, and half of the sum of the values is the number of triangles around node i.

The average distance between node i and all other nodes (L _i ) is obtained from equation (16).

If there are N nodes with node connectivity value between node i and node i, the average distance between node i and all other nodes is obtained by finding the shortest path length between node i and all other nodes, summing them, Is divided by the value obtained by subtracting 1 from the number (N) of nodes having the inter-node connectivity value with node i.

The characteristic path length (L) is obtained by equation (17). The characteristic path length L is a value obtained by obtaining the average distance Li between one node and all the other nodes for all the nodes and dividing the sum by the number of all nodes N, Also called the length.

Efficiency of node _i (E _i ) is obtained by equation (18).

That is, the efficiency E _i of the node i is obtained by calculating the reciprocal of the shortest path length (d _ij ) between the node i and all the other nodes, summing the sum and subtracting 1 from the total number (N) .

The global efficiency (E) is obtained by equation (19). The total efficiency (E) is a value obtained by dividing the sum of efficiency of each node by the total number of nodes (N), which is an average of efficiency of each node.

The clustering coefficient of node i (Ci) of node i is obtained by equation (20). Clustering coefficients are also referred to as cluster coefficients or aggregation (or cluster) coefficients.

Of node i clustering coefficient (Ci), the node i the triangle number (t _i) of the surrounding, the product of the value obtained by subtracting 1 from the order (k _i) of the node, the order (k _i) of the node value, It refers to divided value. The clustering factor evaluates the functional segregation based on whether or not the neighboring nodes are directly connected to each other.

The clustering coefficient (C) is obtained by the following equation (21). The clustering coefficient C is a value obtained by adding the clustering coefficients Ci of the respective nodes and dividing the value by the total number of nodes (N), which is an average of the clustering coefficients of the respective nodes.

The local efficiency of node i (E _{loc, t} ) of node i is found by Equation (22).

The reciprocal of the shortest path length (d _ij ) of node j and node h and the reciprocal of the node i and node j, while changing the values of j and h in node j and node h, The value obtained by multiplying the connectivity value cij by the node i and the connectivity value cih between nodes of the node h is calculated and the values thus obtained are added together and the summed value is multiplied by the degree of node k _i , (k _i ), which is obtained by multiplying the value obtained by subtracting 1 from the local efficiency (E _{loc, i} ) of the node i. In the local efficiency, since the inverse number of the shortest path length between neighboring nodes is calculated and used in all the local networks configured only in the vicinity of a certain node, the functional separation including the indirect connection can be evaluated.

The local efficiency (E _loc ) is found by equation (23).

Here, n is the number of nodes in the local network, and the local efficiency (E _loc ) is a value obtained by dividing the sum of the local efficiencies (E _{loc, i} ) of each node by the number of nodes in the local network. It is an average.

There are many ways to set up a node for a brain network configuration. A first method is to configure a network by setting each of the 116 regions functionally separated from an AAL (Automated Anatomical Labeling) template as one node as shown in FIG.

The AAL template is an image, each area is divided into images, and the positions of the pixels corresponding to each area are stored in a database. Therefore, when each pixel is selected with the mouse in the AAL template, the pixel position can be known, so that it can be known which of the 116 areas the pixel corresponds to.

Another method is to divide the regions obtained 2D or 3D MRS data anatomically and set them as nodes.

In the present invention, both methods are used.

FIG. 13 shows each area of the known AAL template, and the names of the separated areas in FIG. 13 are shown in Table 1.

1. Precentral _L
  ( Center last time _left) 2. Precentral_R
  ( Center last time _Right side) 3. Frontal_Sup_L
  ( Forehead _left) 4. Frontal_Sup_R
  ( Forehead _Right side) 5. Frontal_Sup_Orb_L
  ( Orbit above the forehead _left) 6. Frontal _Sup_Orb_R
  ( Orbit above the forehead _Right side) 7. Frontal_Mid_L
  (Left forehead _ left) 8. Frontal_Mid_R
  (Right half of forehead) 9. Frontal_Mid_Orb_L
  ( Orbit of the forehead _left) 10. Frontal_Mid_Orb_R
  ( Orbit of the forehead _Right side) 11. Frontal_Inf_Oper_L
  ( Lobes below the forehead _left) 12. Frontal_Inf_Oper_R
  ( Lobes below the forehead _Right side) 13. Frontal_Inf_Tri_L
  ( Triangle below the forehead _left) 14. Frontal_Inf_Tri_R
  ( Triangle below the forehead _Right side) 15. Frontal_Inf_Orb_L
  ( Orbit below the forehead _left) 16. Frontal_Inf_Orb_R
  ( Orbit below the forehead _Right side) 17. Rolandic_Oper_L
  ( Roland Dick _left) 18. Rolandic_Oper_R
  ( Roland Dick _Right side) 19. Supp_Motor_Area_L
  (Auxiliary exercise area - left) 20. Supp_Motor_Area_R
  (Auxiliary exercise area_Right) 21. Olfactory_L
  ( Rear part _left) 22. Olfactory_R
  ( Rear part _Right side) 23. Frontal_Sup_Medial_L
  ( Forehead yang _left) 24. Frontal_Sup_Medial_R
  ( Forehead yang _Right side) 25. Frontal_Med_Orb_L
  ( Imaan orbit _left) 26. Frontal_Med_Orb_R
  ( Imaan orbit _Right side) 27. Rectus_L
  ( Straight _left) 28. Rectus_R
  ( Straight _Right side) 29. Insula _L
  ( Brain leaf _left) 30. Insula_R
  ( Brain leaf _Right side) 31. Cingulum_Ant_L
  ( The front part of the subject _left) 32. Cingulum_Ant_R
  ( The front part of the subject _Right side) 33. Cingulum_Mid_L
  ( Mature middle middle stage _left) 34. Cingulum_Mid_R
  ( Middle of subject _Right side) 35. Cingulum_Post_L
  ( Back of object _left) 36. Cingulum_Post_R
  ( Back of object _Right side) 37. Hippocampus_L
  (Hippocampus left) 38. Hippocampus _R
  (Hippocampal right) 39. ParaHippocampal _L
  ( Hippocampus _left) 40. ParaHippocampal _R
  ( Hippocampus _Right side) 41. Amygdala _L
  (Amygdala left) 42. Amygdala _R
  (Right side of the amygdala) 43. Calcarine _L
  ( Joe _left) 44. Calcarine _R
  ( Joe _Right side) 45. Cuneus _L
  ( Snow _left) 46. Cuneus _R
  ( Snow _Right side) 47. Lingual _L
  (Language center _ left) 48. Lingual _R
  (Language center _ right) 49. Occipital_Sup_L
  ( Upper back of head _left) 50. Occipital _Sup_R
  ( Upper back _Right side) 51. Occipital _Mid_L
  ( Mid-back _left) 52. Occipital _Mid_R
  ( Mid back _Right side) 53. Occipital_Inf_L
  ( Inner lobe of ridge _left) 54. Occipital _Inf_R
  ( Inner lobe of ridge _Right side) 55. Fusiform _L
  ( The _left) 56. Fusiform _R
  ( The _Right side) 57. Postcentral _L
  ( Center back gore _left) 58. Postcentral _R
  ( Center back gore _Right side) 59. Parietal _Sup_L
  ( Poodle _left) 60. Parietal _Sup_R
  ( Poodle _Right side) 61. Parietal Inf _L
  ( Doodle Lobe _left) 62. Parietal _ Inf _R
  ( Doodle Lobe _Right side) 63. SupraMarginal _L
  ( Above corners _left) 64. SupraMarginal _R
  ( Above corners _Right side) 65. Angular_L
  ( Each time _left) 66. Angular _R
  ( Each time _Right side) 67. Precuneus _L
  ( Front leaflet _left) 68. Precuneus _R
  ( Front leaflet _Right side) 69. Paracentral _Lobule_L
  ( Center lobed leaf _left) 70. Paracentral_Lobule_R
  ( Center lobed leaf _Right side) 71. Caudate _L
  ( Liverless _left) 72. Caudate _R
  ( Liverless _Right side) 73. Putamen _L
  ( Pieces _left) 74. Putamen _R
  ( Pieces _Right side) 75. Pallidum _L
  (Left) 76. Pallidum _R
  (Right) 77. Thalamus_L
  (Prize_ Left) 78. Thalamus _R
  (Prize_Right) 79. Heschl _L
  ( Hechel _left) 80. Heschl _R
  ( Hechel _Right side) 81. Temporal_Sup_L
  ( Temporomandibular _left) 82. Temporal _Sup_R
  ( Temporomandibular _Right side) 83. Temporal _Pole_Sup_L
  ( A temporal upper electrode _left) 84. Temporal _Pole_Sup_R
  ( A temporal upper electrode _Right side) 85. Temporal_Mid_L
  ( Middle temporal lobe _left) 86. Temporal _Mid_R
  ( Middle temporal lobe _Right side) 87. Temporal _Pole_Mid_L
  ( Mid-shaft pole _left) 88. Temporal _Pole_Mid_R
  ( Middle temporal pole _Right side) 89. Temporal_ Inf _L
  ( Temporal lobe _left) 90. Temporal _ Inf _R
  ( Temporal lobe _Right side) 91. Cerebelum _Crus1_L
  ( Cerebellar lowering 1 _left) 92. Cerebelum _ Crus1 _R
  ( Cerebellar lowering 1 _Right side) 93. Cerebelum_Crus2_L
  ( Cerebellar lowering 2 _left) 94. Cerebelum _Crus2_R
  ( Cerebellar lowering 2 _Right side) 95. Cerebelum _3_L
  ( Cerebellum 3 _left) 96. Cerebelum _3_R
  ( Cerebellum 3 _Right side) 97. Cerebelum_4_5_L
  ( Cerebellum 4 _5_ left) 98. Cerebelum _4_5_R
  ( Cerebellum 4 _5_ right) 99. Cerebelum _6_L
  (Cerebellum _6_ left) 100. Cerebelum _6_R
  (Cerebellum _6_ right) 101. Cerebelum _7b_L
  (Cerebellum _7b_ left) 102. Cerebelum _7b_R
  (Cerebellum _7b_right) 103. Cerebelum _8_L
  (Cerebellum _8_ left) 104. Cerebelum _8_R
  (Cerebellar _8_ right) 105. Cerebelum _9_L
  (Cerebellum _9_ left) 106. Cerebelum _9_R
  (Cerebellum _9_right) 107. Cerebelum _10_L
  (Cerebellum _10_ left) 108. Cerebelum _10_R
  (Cerebellum _10_right) 109. Vermis _1_2
  ( Insistence _1_2) 110. Vermis _3
  ( Insistence _3) 111. Vermis _4_5
  ( Insistence _4_5) 112. Vermis _6
  ( Insistence _6) 113. Vermis _7
  ( Insistence _7) 114. Vermis _8
  ( Insistence _8) 115. Vermis _9
  ( Insistence _9) 116. Vermis _10
  ( Insistence _10)

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Modification is possible. Accordingly, it is intended that the scope of the invention be defined by the claims appended hereto, and that all equivalent or equivalent variations thereof fall within the scope of the present invention.

100: Magnetic Resonance Spectroscopy (MRS) equipment 110: Signal preprocessing unit
200: operation processing unit 250:
310: key input unit 320: output unit

Claims (21)

A spectral receiving step, at a predetermined time interval, for monitoring the brain metabolism, from the magnetic resonance spectroscopy (MRS) equipment, a spectrum of the obtained brain region is received;
A spectral time-varying function acquiring step of acquiring a time-varying function of a cerebral metabolism from the spectrum received in the spectral receiving step, the amount of metabolite showing a quantitative change with time;
Wherein the operation processing unit generates a cerebral metabolic network using the time-varying function obtained in the time-varying function acquiring step;
&Lt; / RTI &gt; The method according to claim 1,
Characterized in that the spectral image received by the arithmetic processing unit in the spectral receiving step is a multi-voxel magnetic resonance spectrum (MRS) obtained using a 2D MR spectroscopy (MRSI) technique or a 3D MRSI technique. Way. 3. The method of claim 2,
The spectrum represents the degree of chemical shift according to the amount of metabolite, and the chemical shift (?

(Where f is the Lamor frequency of the hydrogen nucleus in the metabolite and f _ref is the frequency of tetramethylsilane (TMS) or 4-dimethyl-4-silapentane-1-sulfonic acid (DSS)
&Lt; / RTI &gt; The method of claim 3,
Wherein the chemical shift value of the central point of the peak in the spectrum indicates a metabolite and the width of the peak indicates an amount of a metabolite. 5. The method of claim 4,
The spectrum is measured using a magnetic resonance spectroscopy (MRS) apparatus in a phantom solution containing a plurality of metabolic materials including alanine (Alanine, Ala) and each metabolite alone. And generating a set of basis data. 6. The method of claim 5,
For the basis set data, spectra were measured by making a phantom solution containing only each metabolite for 18 brain metabolites, and the 18 brain metabolites were analyzed for alanine (Alanine, Ala), aspartate (Aspartate, Asp), Creatine (Cr), Phosphocreatine (PCr), Gamma-Aminobutyric acid (GABA), Glucose, Glc, Glutamate Glutamine, Gln, Glutathione (GSH), Glycerophosphocholine (GPC), Phosphocholine, PCho, Myo-Inositol, Ins, Lactate (Lsc) , N-Acetylaspartate (NAA), N-Acetylaspartylglutamate (NAAG), Phosphocreatine (PCr), Phosphorylethanolamine (PE), Taurine (Taurine, Tau). A method for generating a cerebral metabolite network. The method according to claim 1,
In the spectral time-varying function acquisition step
The signal measurement time interval for acquiring the time series signal of the 2D magnetic resonance spectrum (MRS) is TR x N _x N _y x NEX
(Where TR is the repetition time of the MRSI pulse corresponding to each car- riage frequency, N _x is the number of voxels arranged in the transverse direction of the multi-voxel to obtain the spectrum in 2D MRSI, and N _y is the spectrum obtained in 2D MRSI The number of voxels arranged in the longitudinal direction of the multivoxel, and NEX is the number of times the experiment is repeated under the same conditions)
Of the total amount of the metabolites. The method according to claim 1,
In the network configuration step, for the network configuration, the arithmetic processing unit determines a specific voxel or region as a seed and then identifies a region showing a high correlation with the voxel or region (seed-based correlation analysis; RTI ID = 0.0 &gt; (SCA). &Lt; / RTI &gt; 9. The method of claim 8,
When the MRSI data are measured at n different times (t1, t2, ..., t10), the amounts of each metabolite are obtained in each voxel at each time, and as a result, the amount of each metabolite per voxel is n Data is expressed as a time-varying amount of data. 10. The method of claim 9,
In the network configuration step, when using the SCA method,
Let x and y be the time-varying amounts of the metabolites in the two voxels, respectively, and the amount of metabolites measured at the specific time t _i is called x _i , y _i ,

,

, And the standard deviation of the time variant is s _x , s _y , the correlation coefficient (r) of the two voxels is

, The correlation coefficient (r) has a value between -1 and 1,
The arithmetic processing unit sets the correlation coefficient r to a predetermined threshold _rthresh ).
It is determined that there is a positive correlation when the correlation coefficient r is large and it is determined that there is a negative correlation if the correlation coefficient r is small and when the correlation coefficient r is large, Or if not, it is determined that there is no correlation. 11. The method of claim 10,
The operation processing unit displays the correlation coefficient as a standardized z value used in statistical analysis, encrypts the correlation coefficient and displays it on an anatomical image,
The edge areas of the two areas where the correlation coefficient is determined are connected to each other at an edge of the network. When the edge areas are connected by positive correlation, &Lt; / RTI &gt; is a different color. 12. The method of claim 11,
Wherein the arithmetic processing unit is represented by a red edge when connected by a positive correlation and by a blue edge when connected by a negative correlation. The method according to claim 1,
Characterized in that, for the network configuration, in the network configuration step, a graph theory approach is used. 14. The method of claim 13,
Wherein the network comprises a node and an edge. &Lt; Desc / Clms Page number 13 &gt; 15. The method of claim 14,
In the network configuration phase, when using the graph theory approach,
Connectivity between nodes is represented by a connectivity matrix (C), and the connection matrix of a brain network consisting of N nodes forms an N × N matrix,

Wherein the i-th row column element c _ij of the connection matrix is a value indicating the connectivity between the i-th node and the j-th node. 16. The method of claim 15,
In the case of a binary network, the connection matrix c _ij

(Where r is the correlation coefficient and _rthresh Is the threshold of the correlation coefficient)
Lt;
If the connection matrix value is 0, the two nodes are not connected to each other,
Wherein when the connection matrix value is not 0, it is regarded as being connected to each other, and the two nodes are connected to each other by an edge. 17. The method of claim 16,
In the case of a weighted network, the connection matrix c _ij is

Lt;
If the connection matrix value is 0, the two nodes are not connected to each other,
Wherein when the connection matrix value is not 0, it is regarded as being connected to each other, and the two nodes are connected to each other by an edge. 18. The method of claim 17,
In the network configuration step, the arithmetic processing unit divides the brain functionally into a plurality of different areas, expresses each area as a node, calculates a correlation coefficient r between the nodes to construct a connection matrix, And a network constituted by a connection matrix expressed in gray level is referred to as a weighted network,
Wherein, in the weighted network, when r = 0, the two nodes are not connected to the edge, and if r ≠ 0, the edges are connected to the edge, and the thickness of the edge is determined in proportion to the r value. How to create a network. 19. The method according to any one of claims 11 to 18,
The node degree, the shortest path length, the centrality, and the clustering coefficient are used to analyze the network formed between the metabolites in the brain through the network constructed in the network configuration step. The method comprising the steps of: detecting a network parameter that includes the network parameter. 20. The method of claim 19,
The network parameters include the number of triangles around a node, the average distance between nodes and all other nodes, The efficiency of a particular node, the global efficiency, the clustering coefficient of the node i, the local efficiency of the node i, the local efficiency ). &Lt; / RTI &gt; 19. A recording medium storing a computer program source for a method of generating a cerebral metabolite network according to any one of claims 1 to 18.

Images