JP5037866B2 - Magnetic resonance imaging system - Google Patents

Description

  The present invention measures nuclear magnetic resonance (hereinafter referred to as `` NMR '') signals from hydrogen, phosphorus, etc. in a subject and images nuclear density distribution, relaxation time distribution, etc. In particular, the present invention relates to a technique that achieves both high speed parallel imaging and high image quality.

  The MRI device measures NMR signals (echo signals) generated by the spins of the subject, especially the tissues of the human body, and forms the shape and function of the head, abdomen, limbs, etc. in two or three dimensions. It is a device that images. In imaging, the echo signal is given different phase encoding depending on the gradient magnetic field and is frequency-encoded and measured as time-series data. The measured NMR signal is reconstructed into an image by two-dimensional or three-dimensional Fourier transform.

In the MRI apparatus as described above, as one solution for shortening the imaging time, the number of echo signals to be acquired is reduced by thinning out the phase encoding to shorten the time until image acquisition. However, when an image is reconstructed using data that has been thinned out by phase encoding, aliasing artifacts are generated in the image. For this reason, several methods have been realized to remove the aliasing artifacts with high accuracy and acquire time-series images with high time resolution, and one of them is a method that uses the sensitivity difference of each coil constituting a multiple coil (SENSE Law; Non-Patent Document 1).
“SENSE: Sensitivity Encoding for Fast MRI”, Klaas P. Pruessmann et al., Magnetic Resonance in Medicine 42: 952-962 (1999)

In addition, before and after the main measurement, which is obtained by thinning out the phase encoding, a method of adding training measurement to grasp the movement of the subject (heartbeat and periodic pulsation caused by it) has also been proposed (Non-Patent Document). 2). This method acquires the dynamic elements of the target in training measurement and the static elements in the main measurement, and combines them in the image reconstruction algorithm to output a good time-series image in which aliasing artifacts are suppressed. . In this measurement, in order to improve the temporal resolution by increasing the spatial resolution, echo signals are thinned out and acquired over the entire k-space, but in the training measurement, only the low-frequency region of the k space is acquired. By such k-space scanning, an image obtained from the main measurement has high spatial resolution, but aliasing artifacts appear. On the other hand, the image obtained from the training measurement has a low spatial resolution, but has an image quality sufficient for estimating the temporal change of the pixel value. A motion of the imaging target is estimated from the pixel value likelihood distribution of the training image acquired in this way, and is combined with a still image estimated from the main measurement, thereby reconstructing a real-time image having high temporal and spatial resolution. Such a high-speed imaging method is said to have a higher artifact suppression effect than the SENSE method described above.
“Kt BLAST and kt SENSE: Dynamic MRI With High Frame Rate Exploiting Spatiotemporal Correlations”, Jeffery Tsao et al., Magnetic Resonance in Medicine 50: 1031-1042 (2003)

The measurement order of the training measurement and the main measurement is roughly classified into a pre-scan type (Non-Patent Document 3) and a self-calibration type (Patent Document 1).
The pre-scan type (Non-patent Document 3) acquires a single image by combining dynamic elements and static elements obtained from training measurements and main measurements at different measurement times in an image reconstruction algorithm. . This method assumes that the subject to be photographed has the same movement (hereinafter referred to as “identity of body movement”) between the training measurement and the main measurement. In this measurement, since the echo signal is acquired by thinning out the phase encoding, the acquired time-series image can realize a high frame rate (that is, high time resolution).

On the other hand, the self-calibration type (Patent Document 1) performs the training measurement and the main measurement almost simultaneously by acquiring data by densely decimating the low frequency of the k space and thinning out the high frequency during the main measurement. Then, after data acquisition, the acquired k-space data is separated into dynamic elements and static elements, and then image reconstruction is performed. In other words, this method acquires a dynamic element and a static element at approximately the same time by one measurement in order to maintain the same body motion. For this reason, the same body motion is maintained between the training measurement and the main measurement, and it is possible to effectively suppress the aliasing artifact and perform stable image reconstruction.
“Accelerating Cardiac Cine 3D Imaging Using kt BLAST”, Sebastian Kozerke et al., Magnetic Resonance in Medicine 52: 19-26 (2004) Japanese Patent Laid-Open No. 11-262479

However, in the pre-scan type imaging, the aliasing artifacts may not be effectively suppressed and remain on the image due to the fact that the body motion identity between the two measurements is not maintained due to various factors. That is, the image reconstruction may become unstable. In addition, in self-calibration type imaging, the number of echoes necessary for composing one image increases, so the number of time-series images acquired per unit time (that is, the frame rate) decreases.
For this reason, the operator can take into account the characteristics of the pre-scan type and self-calibration type described above and the necessary conditions at the time of shooting, and enable the pre-scan type that enables high time resolution and stable image reconstruction. In this situation, one of the self-calibration types must be selected. Therefore, a measurement method that combines the features of both the pre-scan type and the self-calibration type, that is, high temporal resolution while maintaining the stability of image reconstruction (that is, the ability to suppress aliasing artifacts stably). A measurement method is desired but not realized.
Therefore, the present invention is to realize a measurement method that realizes high time resolution while maintaining stability of image reconstruction.

In order to achieve the above object, the MRI apparatus of the present invention is configured as follows. That is,
Measurement control means for measuring nuclear magnetic resonance signals generated from a subject placed in a static magnetic field to obtain k-space data necessary for image reconstruction, and re-image the subject using k-space data. A signal processing unit that configures the body movement information of the subject, and the measurement control unit obtains the first data corresponding to the low range of the k space and performs phase encoding over the entire region of the k space. The second data corresponding to the area thinned out in the direction is acquired, and at least one of the acquisition amount and the acquisition frequency of the first data acquisition is controlled based on the body motion information. The image is reconstructed using the data and the second data.

  Specifically, an adaptive self-calibration type configuration is proposed as a new configuration that combines the features of both the pre-scan type and the self-calibration type. In this adaptive self-calibration type high-speed imaging method, self-calibration type data acquisition is performed, but the number of training data phase encodings (hereinafter referred to as N) to be acquired is set variably. N is updated as needed according to the body movement of the subject. In other words, the body movement of the subject is monitored, and if the identity of the subject's body movement tends to be lost over time, N is increased to increase the amount of update of the dynamic information. When there is a tendency to increase the identity, reduce N and reduce the amount of dynamic information updates.

  According to the MRI apparatus of the present invention, it is possible to realize a high time resolution while maintaining the stability of image reconstruction, and to perform measurement having both features of a pre-scan type and a self-calibration type. Is possible.

  Hereinafter, preferred embodiments of the MRI apparatus of the present invention will be described in detail with reference to the accompanying drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiments of the invention, and the repetitive description thereof is omitted.

  First, an overall outline of an example of the MRI apparatus according to the present invention will be described with reference to FIG. FIG. 12 is a block diagram showing the overall configuration of an embodiment of the MRI apparatus according to the present invention. This MRI apparatus uses a NMR phenomenon to obtain a tomographic image of a subject.As shown in FIG. 2, the MRI apparatus includes a static magnetic field generation system 2, a gradient magnetic field generation system 3, a transmission system 5, A reception system 6, a signal processing system 7, a sequencer 4, and a central processing unit (CPU) 8 are provided.

   The static magnetic field generation system 2 generates a uniform static magnetic field in the direction perpendicular to the body axis in the space around the subject 1 if the vertical magnetic field method is used, and in the direction of the body axis if the horizontal magnetic field method is used. Thus, a permanent magnet type, normal conducting type or superconducting type static magnetic field generating source is arranged around the subject 1.

  The gradient magnetic field generation system 3 includes a gradient magnetic field coil 9 wound in the three axes of X, Y, and Z, which is a coordinate system (static coordinate system) of the MRI apparatus, and a gradient magnetic field power source 10 that drives each gradient magnetic field coil. The gradient magnetic fields Gx, Gy, and Gz are applied in the three axis directions of X, Y, and Z by driving the gradient magnetic field power supply 10 of each coil according to a command from the sequencer 4 described later. At the time of imaging, a slice gradient magnetic field pulse (Gs) is applied in a direction orthogonal to the slice plane (imaging cross section) to set a slice plane for the subject 1, and the remaining two orthogonal to the slice plane and orthogonal to each other A phase encoding gradient magnetic field pulse (Gp) and a frequency encoding gradient magnetic field pulse (Gf) are applied in the direction, and position information in each direction is encoded in the echo signal.

  The sequencer 4 is a control means that repeatedly applies a high-frequency magnetic field pulse (hereinafter referred to as “RF pulse”) and a gradient magnetic field pulse in a predetermined pulse sequence, and operates under the control of the CPU 8 to collect tomographic image data of the subject 1. Various commands necessary for the transmission are sent to the transmission system 5, the gradient magnetic field generation system 3, and the reception system 6.

  The transmission system 5 irradiates the subject 1 with RF pulses in order to cause nuclear magnetic resonance to occur in the nuclear spins of the atoms constituting the living tissue of the subject 1, and includes a high frequency oscillator 11, a modulator 12, and a high frequency amplifier. 13 and a high frequency coil (transmission coil) 14a on the transmission side. The high-frequency pulse output from the high-frequency oscillator 11 is amplitude-modulated by the modulator 12 at a timing according to a command from the sequencer 4, and the amplitude-modulated high-frequency pulse is amplified by the high-frequency amplifier 13 and then placed close to the subject 1. By supplying to the high frequency coil 14a, the subject 1 is irradiated with the RF pulse.

  The receiving system 6 detects an echo signal (NMR signal) emitted by nuclear magnetic resonance of nuclear spins constituting the biological tissue of the subject 1, and receives a high-frequency coil (receiving coil) 14b and a signal amplifier 15 on the receiving side. And a quadrature detector 16 and an A / D converter 17. After the NMR signal of the response of the subject 1 induced by the electromagnetic wave irradiated from the high frequency coil 14a on the transmission side is detected by the high frequency coil 14b arranged close to the subject 1 and amplified by the signal amplifier 15, The signal is divided into two orthogonal signals by the quadrature phase detector 16 at the timing according to the command from the sequencer 4, and each signal is converted into a digital quantity by the A / D converter 17 and sent to the signal processing system 7.

  The signal processing system 7 performs various data processing and display and storage of processing results, and has an external storage device such as an optical disk 19 and a magnetic disk 18 and a display 20 composed of a CRT, etc. Is input to the CPU 8, the CPU 8 executes processing such as signal processing and image reconstruction, and displays the tomographic image of the subject 1 as a result on the display 20, and the magnetic disk 18 of the external storage device. Record in etc.

  The operation unit 25 inputs various control information of the MRI apparatus and control information of processing performed in the signal processing system 7, and includes a trackball or mouse 23 and a keyboard 24. The operation unit 25 is disposed close to the display 20, and the operator controls various processes of the MRI apparatus interactively through the operation unit 25 while looking at the display 20.

  In FIG. 12, the high-frequency coil 14a and the gradient magnetic field coil 9 on the transmission side face the subject 1 in the static magnetic field space of the static magnetic field generation system 2 into which the subject 1 is inserted, in the case of the vertical magnetic field method. If the horizontal magnetic field method is used, the subject 1 is installed so as to surround it. The high-frequency coil 14b on the receiving side is installed so as to face or surround the subject 1.

  At present, the radionuclide to be imaged by the MRI apparatus is a hydrogen nucleus (proton) which is a main constituent material of the subject as being widely used clinically. By imaging information on the spatial distribution of proton density and the spatial distribution of relaxation time in the excited state, the form or function of the human head, abdomen, limbs, etc. is imaged two-dimensionally or three-dimensionally.

  Next, an example of a pulse sequence for measuring an echo signal from a subject in the MRI apparatus will be described. FIG. 13 shows a gradient echo pulse sequence which is an example of a pulse sequence. However, the present invention is not limited to the pulse sequence shown in FIG. 13, but can be applied to other pulse sequences.

  In FIG. 13, RF, Gs, Gp, Gr, A / D, and echo represent RF pulse, slice gradient magnetic field, phase encode gradient magnetic field, frequency encode gradient magnetic field, A / D conversion, and echo signal, respectively, and 1301 represents a high frequency pulse. 1302 is a slice gradient magnetic field pulse, 1303 is a phase encode gradient magnetic field pulse, 1304 is a frequency encode gradient magnetic field pulse, 1305 is a sampling window, 1306 is an echo signal, and 1307 is a repetition time (interval of 1301). In the gradient echo pulse sequence, the echo signal obtained by each phase encoding is provided by changing the amount of the phase encoding gradient magnetic field pulse 1303 (= area surrounded by the gradient magnetic field pulse waveform and the time axis) for each repetition 1307 and giving different phase encoding. 1306 is detected. This operation is repeated by the number of phase encodings, and an echo signal necessary for reconstruction of one image is acquired at an image acquisition time 1308. As the number of phase encodings, values such as 64, 128, 256, 512, etc. are usually selected per image. Each echo signal is usually obtained as a time-series signal composed of 128, 256, 512, and 1024 sampling data. These data are two-dimensionally Fourier transformed to create one MR image.

Next, an outline of a pre-scan type imaging method and a self-calibration type imaging method realized in the MRI apparatus will be described.
First, a prescan type photographing method will be described with reference to FIG. In FIG. 1, the training measurement is indicated by block T, and the main measurement is indicated by block A. In training measurement, low-frequency data of k-space is mainly acquired, and in this measurement, echo signals are thinned out and acquired over the entire area of k-space. Hereinafter, this block notation is used to indicate the data acquisition range in the k space. The vertical axis in FIG. 1 represents the ky axis of k space corresponding to phase encoding, and the horizontal axis represents time. The data arrangement for the other axes (kx, kz) in the k space is omitted because it is the same as the normal measurement.

  FIG. 1 (a) shows the measurement order in the pre-scan type. That is, the training data 101 is acquired first, and then the main measurement data 102 is acquired. FIG. 1 (b) represents the data acquisition point 104 of the training measurement block T101. In the training measurement block T, only the low frequency region of the k space is densely measured. In each training measurement block T, the same ky point (that is, the same phase encoding) is measured. Further, FIG. 1 (c) represents the data acquisition point 106 of the main measurement block A102, and the main measurement block A represents that measurement is performed with phase encoding thinned out from the low range to the high range of the k space. . In each main measurement block A, the ky point incremented for each block is measured. (For example, in the case of 3x shooting, if ky = 1, 4, 7, 10,... Is acquired in frame 1, ky = 2, 5, 8, 11,. (Sampling method of acquiring ky = 3, 6, 9, 12,... and acquiring the same ky as that of frame 1 in frame 4.) In FIG. 1, the training measurement block T101 and the main measurement block A 102 are acquired. The time width is similarly described, but actually varies depending on the number of acquired data. Generally, the time width of the main measurement block A102 is longer than the time width of the training measurement block T101. That is, the data acquisition number in the main measurement block A102 is larger than the data acquisition number in the training measurement block T101.

  Next, a self-calibration type photographing method will be described with reference to FIG. FIG. 2 (a) shows a measurement sequence in the self-calibration type, and shows that the training measurement T201 and the main measurement A202 are alternately performed. The data acquisition points on the k space in the training measurement block T201 and the main measurement data block A202 are the same as those in FIGS. 1 (b) and 1 (c), respectively. As a result, the self-calibration type k-space data has a low frequency in the phase encoding direction, with the data 203 in the dotted frame combining the training measurement block T and the main measurement data block A one by one as the minimum acquisition unit. The data is acquired densely, and the high frequency is thinned out, resulting in sparsely acquired data. This is shown in Fig. 2 (b).

Next, based on the above description, an embodiment of the MRI apparatus of the present invention will be described. In the present embodiment, body motion information is detected, and at least one of the acquisition amount and the acquisition frequency of training data is controlled based on the body motion information. Hereinafter, the measurement method of this embodiment is referred to as an adaptive self-calibration method.
First, the outline of the adaptive self-calibration method will be described with reference to FIG. In the adaptive self-calibration method, detection of body motion information is performed in parallel with measurement in order to detect body motion of a subject. Since this body motion information varies depending on the application, details of the body motion information detection method will be described separately below. FIG. 3 shows an example of how the ratio between the training measurement and the main measurement changes depending on the presence or absence of body motion identity due to the body motion of the subject.

  FIG. 3 (a) shows a schematic diagram of the imaging sequence when the disturbance 301 of the same body motion due to body motion occurs. The training measurement block T alternates between this measurement block A immediately after the occurrence of the disturbance 301. The example inserted in is shown. Before the occurrence of the disturbance 301, the training data and the main measurement data had the same body movement, so the training measurement block T is omitted and only the main measurement block T is performed. . At this time, the imaging sequence is equivalent to the pre-scan type, and the number of phase encodings for training data is set to 0 (zero). Then, immediately after the occurrence of the disturbance 301, the process proceeds to an imaging sequence in which the training measurement block T is inserted. At this time, the imaging sequence is equivalent to the self-calibration type, and the number of phase encodes for training data is increased to N.

Fig. 3 (b) shows a schematic diagram of the imaging sequence when the body movement of the subject converges and becomes stable without disturbance, and shows an example in which the training measurement block T is gradually reduced. ing. Initially, an imaging sequence in which the training measurement block T and the main measurement block A are alternately repeated is performed. At this time, the imaging sequence is equivalent to the self-calibration type, and the number of phase encoding for training data is set to N. In this state, when there is no disturbance for a predetermined period, the same body motion is maintained between the training data and the actual measurement data. Since the training data need not be acquired every time if the same body motion is maintained, the training measurement block T is gradually decreased, and finally the training measurement block T is omitted and this measurement block Migrated so that only T is done. At this time, the imaging sequence is equivalent to a pre-scan imaging sequence, and the number of phase encoding for training data is set to 0 (zero).
As described above, the amount of training data acquired is controlled based on the detected body motion information.

  Next, a specific processing flow of the adaptive self-calibration method will be described with reference to FIG. FIG. 4 is a flowchart showing the processing flow of the adaptive self-calibration method centered on the control of the number of phase encodings as the amount of training data in training measurement. Hereinafter, an outline of each processing step will be described for each processing step.

In step 401, the number of phase encoding for training data is initialized at the start of imaging. In the case of starting shooting from pre-scan type measurement, for example, 32 can be set. On the other hand, when shooting is started from the self-calibration type measurement, 0 (zero) is set.
In step 402, a variation threshold is set corresponding to the individuality of the subject.
In step 403, training data acquisition for the set number of phase encodings for training data is started. For example, it is acquired using a gradient echo pulse sequence as shown in FIG.

In step 404, it is determined whether self-monitoring or external monitoring is performed as a monitoring method for detecting body movement information. When self-monitoring is performed, the process proceeds to step 405, and when external monitoring is performed, the process proceeds to step 406. Details of self-monitoring and external monitoring will be described later.
If it is determined in step 405 that self-monitoring is performed, self-monitoring is performed.
If it is determined in step 406 that external monitoring is performed, external monitoring is performed.

In step 407, a difference between the body movement position detected by monitoring and a predetermined reference position as a reference is obtained to calculate a fluctuation amount of the body movement.
In step 408, the body motion fluctuation amount is compared with a predetermined threshold value. If the variation amount is larger than the threshold value, the process proceeds to step 409, and if smaller, the process proceeds to step 410.

  If the fluctuation amount exceeds the threshold value in step 409, the number of training data phase encodes is increased as a disturbance to the identity of body motion. In particular, when a disturbance occurs in a pre-scan type shooting sequence in which the training measurement block T is omitted and only the main measurement block T is performed, the shooting sequence shifts to a state equivalent to the self-calibration type. To do. In this way, by increasing the number of training data phase encodings and increasing the amount of dynamic information, it is possible to respond flexibly to the body movement of the subject, which is an advantage of the self-calibration type.

  If the fluctuation amount is equal to or smaller than the threshold value in step 410, it is determined that the same body motion is maintained and there is no disturbance, and the number of phase encoding for training data is decreased. Finally, the training measurement block T is omitted, and only the main measurement block T is performed. At this time, the imaging sequence is equivalent to a pre-scan imaging sequence, and the number of phase encoding for training data is set to 0 (zero). Thus, by reducing the number of training data phase encodings and reducing the amount of dynamic information, it is possible to achieve high temporal resolution, which is an advantage of the pre-scan type.

In step 411, the main measurement data is acquired. For example, the main measurement data is acquired using a gradient echo pulse sequence as shown in FIG.
In step 412, image reconstruction is performed using the acquired training data and main measurement data. Details of the image reconstruction method will be described later.
In step 413, it is determined whether or not the time-series shooting is finished. If not, the process proceeds to step 403.

  By repeating the above steps 403 to 413 until the end of imaging, a good time-series image with suppressed artifacts is acquired regardless of the amount of body movement of the subject.

  In addition, adaptive self-calibration measurement converges the number of phase encoding for training data to the optimal number according to the body movement of each subject, so either self-calibration type measurement or pre-scan type measurement is used at the start of imaging. It doesn't matter if it starts. When shooting is started with pre-scan measurement, only training measurement is performed first, and then only main measurement is performed. However, as described above, the pre-scan measurement has a configuration almost equivalent to the adaptive self-calibration measurement. Therefore, even if the training data phase encoding number is updated by the training data phase encoding number control algorithm shown in FIG. 4, no contradiction occurs, and the training data phase encoding number is converged to the optimum training data phase encoding number later. On the other hand, when shooting is started with self-calibration type measurement, it is considered adaptive self-calibration data with training data phase encoding number N, so the number of phase encodings for training data is determined by the above phase number control algorithm (Fig. 4). Is updated and converges to the optimal value.

Next, details of the image reconstruction method in step 412 will be described with reference to FIG.
First, the number of training data frames and the number of training data phase encodings used for image reconstruction will be described. The number of phase encodings for training data required for image reconstruction is C, and the phase encoding for training data acquired i (= 0, 1, 2, ...) frames from the actual measurement data 501 and 503 to be reconstructed Let N _i be the number,

を満たす最小のIを算出する。この結果から、本計測データ501、503からIフレーム前のトレーニングデータまでを有効トレーニングデータ502、504として、画像再構成に用いる。たとえば，C＝128とすることができる。ただし，再構成する本計測データから起算してIフレーム前よりも後に，トレーニングデータ位相数が0(ゼロ)からNに増加している場合はトレーニングデータ中に擾乱が含まれていると判断し，Iの値に関わらず擾乱より後に取得したトレーニングデータのみを採用する。
また再構成される時系列画像の時間分解能や再構成画像の安定性を保障するために，(2)のように各フレームで取得するトレーニングデータ位相数に上下限値N_max、N_minを設定することができる。

The smallest I that satisfies is calculated. From this result, the main measurement data 501, 503 to the training data before I frame are used as effective training data 502, 504 for image reconstruction. For example, C = 128. However, if the number of training data phases increases from 0 (zero) to N after the I frame before counting from the actual measurement data to be reconstructed, it is determined that the training data contains disturbances. , Only the training data acquired after the disturbance is adopted regardless of the value of I.
In order to guarantee the time resolution of the reconstructed time-series image and the stability of the reconstructed image, the upper and lower limits N _max and N _min are set for the number of training data phases acquired in each frame as shown in (2) can do.

これらの上下限値N_max、N_minの範囲内で，各トレーニングデータ位相エンコード数が制御される。その制御処理の一例を図8に示す。
時系列画像の時間分解能を保障するためには、トレーニングデータ用位相エンコード数が上限値N_max以下に制限することができる。この場合、図4のステップ409のみがステップ801、802、803に替わる。ステップ801でトレーニングデータ用位相エンコード数が増加された場合、ステップ802で上限値N_maxを超えているかが判定され、超えていた場合はステップ803でトレーニングデータ用位相エンコード数が上限値N_maxに再設定される。

The number of training data phase encodings is controlled within these upper and lower limits N _max and N _min . An example of the control process is shown in FIG.
In order to guarantee the time resolution of the time-series image, the number of phase encoding for training data can be limited to the upper limit value N _max or less. In this case, only step 409 in FIG. 4 is replaced with steps 801, 802, and 803. If number of phase encodes training data is increased in step 801, whether it exceeds the upper limit N _max is determined in step 802, the phase encode number for training data in step 803 if it has exceeded the upper limit value N _max Will be reset.

On the other hand, in order to guarantee the reconstruction stability of the time-series images, the number of training data phase encodings can be limited to the lower limit value N _min or more. In this case, only step 410 replaces steps 804, 805, and 806. If the number of phase encoding for training data is decreased in step 804, it is determined in step 805 whether the value is less than the minimum value N _min. If it is less, the number of phase encoding for training data is reset to the minimum value N _min in step 806. Is done. The limit values N _max and N _min used in these settings are selected according to the characteristics of the movement of the shooting target, the purpose of shooting, and the like. For example, N _max = 64 and N _min = 16 can be set.

Next, which time training data is used for image reconstruction will be described with reference to FIG.
FIG. 5A is a schematic diagram when an image is reconstructed using data acquired in the main measurement 501 immediately after a disturbance occurs. When an image is reconstructed using the data acquired in the main measurement 501, training data having high body motion identity with the data of the main measurement 501 is selected and image reconstruction is performed. The training data 502 acquired after the disturbance and close to the acquisition time of the main measurement 501 has high body movement identity with the data of the main measurement 501, and conversely, the body movement with the training data held before the disturbance The identity is low. Therefore, only data 502 obtained after the disturbance is treated as training data and used for image reconstruction. In such a case, since the number of effective training data phase encodes is small and does not satisfy the required number C, image reconstruction is performed by image reconstruction A described in detail below.

  On the other hand, FIG. 5 (b) is a schematic diagram when an image is reconstructed using data acquired by the main measurement 503 acquired by a stable subject. When reconstructing an image using the data acquired in the main measurement 503, the fact that the identity of the body motion is retained up to some past data 504 is used, and the data 504 is treated as training data. In such a case, since the number of phase encodings for effective training data is large and becomes the required number C or more, image reconstruction is performed by image reconstruction B described in detail below.

  Finally, details of the image reconstruction method will be described with reference to FIG. FIG. 6 is a diagram showing details of image reconstruction (step 412). As described above, the image reconstruction method is switched depending on the number of effective training data phase encodings. In step 601, the number of phase encodings for effective training data is compared with the required number C. If the required number is exceeded, image reconstruction B is performed in step 602. If less than the required number, image reconstruction is performed in step 603. A is done.

  Image reconstruction A in step 602 is a reconstruction method when the number of phase encodings for effective training data is less than the necessary number, and image reconstruction is performed using the method detailed in (Patent Document 1). . In other words, the sensitivity distribution of the receiving coil is obtained using the training data, and the image with the aliasing obtained by reconstructing the measurement data is developed using the coil sensitivity distribution, so that the image without the aliasing is obtained. Is acquired.

  Reconstruction B in step 603 is a reconstruction method in the case where the number of phase encodings for effective training data is the required number C or more, and image reconstruction is performed by a process as shown in FIG. That is, in step 701, the acquired training data and main measurement data are each reconstructed into a time-series image by two-dimensional Fourier transform. In step 702, each pixel of the obtained time-series image is subjected to Fourier transform in the time direction to be converted into a frequency spectrum. In step 703, the spectrum obtained from the training data is used as a mask for extracting dynamic information. Since this mask is updated as needed using the effective training data, the accuracy of extracting dynamic information is improved. On the other hand, in step 704, the spectrum obtained from the measurement data is separated into a static component with frequency = 0 and other dynamic components. In step 705, the dynamic component is masked using the mask data updated in step 703, and the static component is added again. In step 706, the spectrum thus obtained is subjected to inverse Fourier transform in the time direction, thereby obtaining a reconstruction result image using the training data and the main measurement data.

  As described above, according to the present embodiment, it is possible to perform measurement capable of realizing high time resolution while maintaining the stability of image reconstruction, which has both the features of the pre-scan type and the self-calibration type. .

Hereinafter, some specific examples of the embodiment will be described.
First, a first specific example of the present embodiment will be described. This specific example is an example in which body motion of a subject is externally monitored and the number of phase encoding for training data is controlled based on the amount of variation from the average period of body motion. Hereinafter, this specific example will be described by taking the case of cardiac imaging as an example.

  In the imaging of the heart part, since there are many portions that move rapidly in the field of view such as the heart itself, blood vessels, and blood flow, stability of image reconstruction with respect to fluctuations is required. When this embodiment is applied to cardiac imaging, an RR interval (beating cycle) of an electrocardiogram (ECG) is used as a body motion information source to be monitored. Therefore, in the flowchart shown in FIG. 4, the process branches to the external monitoring path of step 406, and the RR interval is monitored using an electrocardiogram apparatus. At this time, in the calculation of the fluctuation amount in step 407, calculation of the average RR interval between the past 10 heartbeats (calculated by moving average; step 901) shown in FIG. 9 and calculation regarding deviation from the average value (step 902) are performed. Do. However, step 901 is not limited to 10 heartbeats, and may be 9 heartbeats or less or 11 heartbeats or more. Then, the difference between the acquired RR interval and the average RR interval so far is used as a fluctuation amount and compared with a threshold value to determine the identity of body motion (step 408). However, the training data at the time of image reconstruction is not the training data close to the time of acquisition of the main measurement data (FIG. 5), but the training data whose elapsed time from the previous R wave is close to the main measurement data. In this way, fluctuations in heart rate that occur during measurement are captured by fluctuations in the R-R interval, so in addition to heart rate disturbances that occur during measurement, the subject can flexibly respond to trends in heart rate fluctuations of each subject. be able to.

Here, the control of the number of phase encoding for training data is performed as shown in FIG. 11, for example. That is, the R-R interval is acquired from the electrocardiogram device as a monitor signal, and the average R-R interval 1101 of the past 10 heartbeats, which is a criterion for determining the identity of body motion, is 1000 milliseconds (60 bpm when converted to a heart rate) And Also, the number N of training data phase encodings acquired last is N = 16, and the variation threshold (step 402) set at the start of imaging is ± 10%. In this case, if the next input RR interval 1102 is in the range of 900 to 1100 milliseconds, it is determined that the identity of the body movement is maintained (step 408), and the training to be acquired next The number of phase encodes for data is changed from 16 to 8 1103. Conversely, if the RR interval 1102 does not fall within the fluctuation threshold ± 10% (that is, if it does not fall within 900-1100 milliseconds), it is determined that the identity of body movement is not maintained. Then, the number N of training data phase encodings to be acquired next is changed 1104 from 16 to 24.
As described above, it is possible to acquire heart images in time series with high temporal resolution while maintaining the stability of image reconstruction regardless of the disturbance in the heart rate.

  Next, as a second specific example of this embodiment, a case where this embodiment is applied to perfusion imaging will be described below. In contrast-enhanced perfusion imaging, the echo signal intensity changes abruptly over time due to the inflow / outflow of contrast medium to the imaging region. For this reason, stability of image reconstruction with respect to fluctuations in echo signal intensity is required. Therefore, when this embodiment is applied to contrast perfusion imaging, echo signal intensity is used as a body motion information source to be monitored. Since this is extracting body motion information from the acquired data, it corresponds to branching to self-monitoring (step 405) in the flowchart of FIG. Further, in the calculation of the fluctuation amount in step 407, as shown in FIG. 10, the echo signal intensity is acquired (step 1001), and the difference from the echo signal intensity at the time of the previous image acquisition is obtained as the fluctuation amount (step 1002). . Then, the amount of variation is compared with a threshold value to determine the identity of body movement (step 408). As a result, it is possible to acquire data by accurately detecting a change in contrast.

  As described above, heart images can be acquired in time series with high temporal resolution while maintaining the stability of image reconstruction, regardless of changes in echo signal intensity due to inflow or outflow of contrast medium. Become.

  Next, the case where this embodiment is implemented in Interventional MRI will be described as a third specific example of this embodiment. In Interventional MRI, a guide wire or catheter is inserted into a blood vessel and moved, and the imaging field of view is moved following the movement. In many cases, a puncture needle or other device is inserted from outside the body.

  Therefore, as an example of applying this embodiment to Interventional MRI, information to be monitored is echo signal intensity in order to cope with a change in imaging conditions such as a change in imaging field of view. Since this is to extract variation information from the acquired data, it corresponds to branching to self-monitoring (step 405) in the flowchart of FIG. Therefore, the identity of body movement is determined by the processing shown in FIG. 10 as in the second specific example described above.

  As another example of applying this embodiment to interventional MRI, it is also possible to monitor a change in position of a device such as a catheter using a device that monitors the operation of the operator. Since the position of the device changes as the device moves, the amount of change in position (that is, the amount of movement) is set as the amount of change. This case corresponds to branching to external monitoring (step 406) in the flowchart of FIG.

  As described above, heart images can be acquired in time series with high temporal resolution while maintaining the stability of image reconstruction regardless of changes in imaging conditions in interventional MRI.

The figure which shows the general prescan type imaging | photography method. Diagram showing a general self-calibration imaging method The figure which shows an example of how the ratio of training measurement and this measurement changes with the presence or absence of the disturbance of the identity of the body motion by the body motion of the subject. The flowchart showing control of the number of phase encoding for training data of this invention. The figure explaining which training data of which time is used for image reconstruction. The flowchart showing selection control of an image reconstruction method. 10 is a flowchart showing details of image reconstruction B. The flowchart showing the control which makes the number of phase encoding for training data into an upper-lower limit value. The flowchart showing the detail of the external monitoring at the time of cardiac imaging | photography. Flow chart showing details of external monitoring during Perfusion / Interventional MRI imaging. The figure which shows the example which controls the number of phase encoding for training data at the time of cardiac imaging | photography. The block diagram which shows the whole structure of the MRI apparatus which concerns on this invention. The figure which shows an example of a gradient echo pulse sequence.

Explanation of symbols

  1 subject, 2 static magnetic field generation system, 3 gradient magnetic field generation system, 4 sequencer, 5 transmission system, 6 reception system, 7 signal processing system, 8 central processing unit (CPU), 9 gradient magnetic field coil, 10 gradient magnetic field power supply, DESCRIPTION OF SYMBOLS 11 High frequency transmitter, 12 Modulator, 13 High frequency amplifier, 14a High frequency coil (transmission coil), 14b ... High frequency coil (reception coil), 15 Signal amplifier, 16 Quadrature phase detector, 17 A / D converter, 18 Magnetic disk , 19 Optical disc, 20 Display, 21 ROM, 22 RAM, 23 Trackball or mouse, 24 Keyboard

Claims (9)

Measurement control means for measuring nuclear magnetic resonance signals generated from a subject placed in a static magnetic field and acquiring k-space data necessary for image reconstruction;
Signal processing means for reconstructing an image of the subject using the k-space data,
The measurement control unit performs the acquisition of the first data corresponding to the low range of the k space and the acquisition of the second data corresponding to the region thinned out in the phase encoding direction in the entire region of the k space,
In the magnetic resonance imaging apparatus in which the signal processing means reconstructs the image using the first data and the second data,
Means for detecting body movement information of the subject,
The magnetic resonance imaging apparatus, wherein the measurement control unit controls the amount of acquisition of the first data per time based on the body motion information. Measurement control means for measuring nuclear magnetic resonance signals generated from a subject placed in a static magnetic field and acquiring k-space data necessary for image reconstruction;
Signal processing means for reconstructing an image of the subject using the k-space data,
The measurement control means acquires first data corresponding to a low region of k space, and acquires second data corresponding to a region thinned out in the phase encoding direction in the entire region of k space in time series,
In the magnetic resonance imaging apparatus, the signal processing means reconstructs the image in time series using the first data and at least one second data.
Means for detecting body movement information of the subject,
The measurement control means, based on the body movement information, inserts the first data acquisition between the second data acquisition, and controls the acquisition amount of the first data per time,
The magnetic resonance imaging apparatus, wherein the signal processing means reconstructs the image in time series using at least one of the first data and at least one of the second data. Measurement control means for measuring nuclear magnetic resonance signals generated from a subject placed in a static magnetic field and acquiring k-space data necessary for image reconstruction;
Signal processing means for reconstructing an image of the subject using the k-space data,
The measurement control means obtains at least one time of the first data corresponding to the low range of the k space and at least one time of the second data corresponding to the entire k space thinned out in the phase encoding direction. And alternately,
In the magnetic resonance imaging apparatus, the signal processing means reconstructs the image in time series using at least one of the first data and at least one of the second data.
Means for detecting body movement information of the subject,
The magnetic resonance imaging apparatus, wherein the measurement control unit controls the amount of acquisition of the first data per time based on the body motion information. The magnetic resonance imaging apparatus according to any one of claims 1 to 3,
The magnetic resonance imaging apparatus, wherein the measurement control means controls the acquisition frequency of the first data based on the body movement information. In the magnetic resonance imaging apparatus according to claim 4,
The magnetic resonance imaging apparatus, wherein the measurement control unit dynamically switches between pre-scan type imaging and self-calibration type imaging by controlling an acquisition amount of the first data. The magnetic resonance imaging apparatus according to any one of claims 1 to 5,
The measurement control unit increases the acquisition amount of the first data when the fluctuation amount of the body motion information is larger than a predetermined threshold, and when the fluctuation amount of the body motion information is smaller than the predetermined threshold, A magnetic resonance imaging apparatus characterized in that the amount of acquisition of one data is reduced. The magnetic resonance imaging apparatus according to any one of claims 1 to 6,
The signal processing means selects the first data close to the acquisition time of the second data after the disturbance of the body motion information as effective first data, and performs image reconstruction using the second data A magnetic resonance imaging apparatus characterized by being used in the above. The magnetic resonance imaging apparatus according to claim 7.
When the number of the effective first data is less than the number necessary for the image reconstruction, the signal processing means
Using the effective first data, obtain a sensitivity distribution of the receiving coil used to acquire the effective first data,
A magnetic resonance imaging apparatus, wherein aliasing of an image obtained by reconstructing the second data using the sensitivity distribution is removed. The magnetic resonance imaging apparatus according to claim 7.
When the number of the effective first data is larger than the number necessary for the image reconstruction, the signal processing means
Obtaining a time-series image from the effective first data, Fourier-transforming in the time direction for each pixel, creating mask data for extracting a dynamic component;
Obtain a time-series image from the second data, perform Fourier transform in the time direction for each pixel, mask the dynamic component using the mask data, and inverse Fourier transform the static component in the time direction to obtain an image. A magnetic resonance imaging apparatus.

Images