JP5595872B2 - Magnetic resonance imaging apparatus and program - Google Patents

Description

  The present invention relates to a magnetic resonance imaging apparatus and a program for sampling k-space data according to a sampling pattern.

As a method for shortening the scan time when scanning a subject, a method is known in which image reconstruction is performed by undersampling without acquiring data of the entire region of k-space. As an image reconstruction method, for example, Compressed
A method called Sensing is known (see Patent Document 1).

JP 2009-268901 A

  The method of Patent Document 1 has a problem that image quality deteriorates when the sampling rate is lowered. Therefore, it is desired that a high-quality image can be acquired while maintaining a short scan time.

A first aspect of the present invention is a magnetic resonance imaging apparatus that performs a scan for sampling k-space data according to a sampling pattern,
Prediction means for predicting a k-space region where data with a large amplitude is highly likely to be concentrated based on sampled k-space data during the scan;
Sampling pattern updating means for updating the sampling pattern based on the region predicted by the prediction means;
A magnetic resonance imaging apparatus.

A second aspect of the present invention is a magnetic resonance imaging apparatus program for performing a scan for sampling k-space data according to a sampling pattern,
Prediction processing for predicting a k-space region where data with a large amplitude is highly likely to be concentrated based on sampled k-space data in the middle of the scan;
A sampling pattern update process for updating the sampling pattern based on the region predicted by the prediction process;
Is a program for causing a computer to execute.

  Since it is possible to predict a k-space region where data with a large amplitude is highly likely to be concentrated, it is possible to acquire a high-quality image.

1 is a schematic diagram showing a magnetic resonance imaging apparatus according to a first embodiment of the present invention. It is a block diagram of the prediction means 91. FIG. It is a figure explaining the flow when scanning the test object. It is a figure which shows schematically the sampling pattern SP initialized. Is a diagram illustrating a data DK ₁ k-space at time _{t 1.} It is a figure which shows the reconstructed image schematically. Is a diagram showing how to convert the image data DI ₁ to the data of the k-space. It is a figure which shows the sampling pattern SP 'of the updated k space. It is a diagram illustrating an example of a data DK _e final k-space obtained by the scanning. It is a diagram showing an image data DI _e. It is a block diagram of the prediction means 91 of the MRI apparatus of the 2nd form. In a 2nd form, it is explanatory drawing of the flow when the test object 12 is scanned. It is a figure which shows a test | inspection range. It is a figure which shows the mode after moving the test | inspection range C. FIG. It is a figure which shows a mode when the test | inspection range C moves to the position of the some left upper side rather than the center of k space. It is a figure which shows the amplitude sum map M roughly. It is a figure which shows the updated sampling pattern SP '. It is a block diagram of the prediction means 91 of the MRI apparatus of the 3rd form. In a 3rd form, it is explanatory drawing of the flow when the subject 12 is scanned. It is a diagram illustrating a data DK ₁₀ of k-space after placing zero.

  Hereinafter, although the form for inventing is demonstrated, the form for inventing is not limited to the following forms.

(1) First Embodiment FIG. 1 is a schematic diagram showing a magnetic resonance imaging apparatus according to the first embodiment of the present invention, and FIG.

  A magnetic resonance imaging (MRI) apparatus 100 includes a magnetic field generator 2, a table 3, a receiving coil 4, and the like.

  The magnetic field generator 2 includes a bore 21 in which the subject 12 is accommodated, a superconducting coil 22, a gradient coil 23, and a transmission coil 24. The superconducting coil 22 applies a static magnetic field B0, the gradient coil 23 applies a gradient magnetic field, and the transmission coil 24 transmits an RF pulse. In place of the superconducting coil 22, a permanent magnet may be used.

  The table 3 has a cradle 31. The cradle 31 is configured to be movable to the bore 21. The subject 12 is transported to the bore 21 by the cradle 31.

  The receiving coil 4 is attached from the abdomen of the subject 12 to the chest. The receiving coil 4 receives a magnetic resonance signal from the subject 12.

  The MRI apparatus 100 further includes a sequencer 5, a transmitter 6, a gradient magnetic field power supply 7, a receiver 8, a central processing unit 9, an operation unit 10, and a display unit 11.

  Under the control of the central processing unit 9, the sequencer 5 sends information for executing a scan to the transmitter 6 and the gradient magnetic field power supply 7.

  The transmitter 6 outputs a drive signal for driving the RF coil 24 based on the information sent from the sequencer 5.

  The gradient magnetic field power supply 7 outputs a drive signal for driving the gradient coil 23 based on the information sent from the sequencer 5.

  The receiver 8 performs predetermined signal processing on the magnetic resonance signal received by the receiving coil 4 and outputs the data obtained by the signal processing to the central processing unit 9.

  The central processing unit 9 implements various operations of the MRI apparatus 100 such as transmitting necessary information to the sequencer 5 and the display unit 11 and reconstructing an image based on data received from the receiver 8. The operation of each unit of the MRI apparatus 100 is controlled. The central processing unit 9 is constituted by a computer, for example. The central processing unit 9 includes a prediction unit 91, a sampling pattern update unit 92, and the like.

  The predicting unit 91 predicts a k-space area where data with a large amplitude is highly likely to be concentrated based on the sampled k-space data during the scan. As shown in FIG. 2, the prediction unit 91 includes an image reconstruction unit 91a, a conversion unit 91b, and an area specifying unit 91c.

  The image reconstruction unit 91a reconstructs an image using the sampled k-space data during the scan.

  The conversion unit 91b converts the image data obtained by the image reconstruction unit 91a into k-space data.

  Based on the k-space data obtained by the conversion unit 91b, the area specifying unit 91c specifies a k-space area where data with large amplitude is highly likely to concentrate.

  The sampling pattern update unit 92 updates the sampling pattern based on the region predicted by the prediction unit 91.

  The central processing unit 9 is an example of the prediction unit 91 and the sampling pattern update unit 92, and functions as these units by executing a predetermined program. The central processing unit 9 is an example of a computer described in a means for solving the problem.

  The operation unit 10 is operated by the operator 13 and inputs various information to the central processing unit 9. The display unit 11 displays various information.

  The MRI apparatus 100 is configured as described above. Hereinafter, a flow when the subject 12 is scanned with the MRI apparatus 100 will be described.

FIG. 3 is a diagram illustrating a flow when scanning the subject 12.
At time t ₀ , scanning is started according to the initially set sampling pattern in the k space (see FIG. 4).

  FIG. 4 is a diagram schematically showing a sampling pattern SP in the initially set k space.

  FIG. 4 shows a kz-ky plane in a three-dimensional k-space. Here, the kz-ky plane is indicated by 16 × 16 squares arranged in two dimensions. The sampling points are indicated by white circles “◯” written in a square. The part where the white circle “◯” is not written means that the sampling point is not initially set.

  In FIG. 4, a 16 × 16 sampling pattern SP is shown, but in general, a 256 × 256 sampling pattern or the like is used. However, in the first embodiment, for convenience of explanation, it is assumed that sampling is performed according to the sampling pattern SP shown in FIG.

Returning to FIG. 3, the description will be continued.
When scanning is started according to the sampling pattern SP (see FIG. 4), it is determined whether the number of sampled data has reached N _i (i = 1 to z). Here, it is determined whether i = 1, that is, whether the number of sampled data has reached N ₁ . N ₁ may be, for example, several (for example, five) or several tens (for example, 30). In the first mode, it is assumed that the number of sampled data has reached N ₁ at time t ₁ (see FIG. 5).

FIG. 5 shows k-space data DK ₁ at time t ₁ .
In FIG. 5, the sampling points are represented by black circles “●” and white circles “◯”. Sampling points with black circles “●” represent points where sampling was performed between time points t _{0 to} t ₁ . Sampling points with white circles “◯” represent points where data has not yet been sampled at time t ₁ .

When the number of sampled data reaches N ₁ , the scan is continued and the first processing flow F ₁ (see FIG. 3) is executed.

In the processing flow F ₁ , first, in step ST ₁ , there is a possibility that the prediction means 91 (see FIGS. 1 and 2) concentrates large amplitude data based on the k-space data DK ₁ (see FIG. 5). Predict high areas. Step ST1 has steps ST11 and ST12 in order to perform this prediction. Below, each step ST11 and ST12 is demonstrated in order.

In step ST11, the image reconstruction unit 91a (see FIG. 2), using the data DK ₁ of k space shown in FIG. 5, to reconstruct an image at time _{t 1} (see FIG. 6).

FIG. 6 is a diagram schematically showing the reconstructed image.
FIG. 6A shows k-space data DK ₁ (see FIG. 5), and FIG. 6B shows image data DI ₁ at time t ₁ obtained by image reconstruction.

As a method of image reconstruction, a compressed sensing method or the like can be used. After obtaining the image data DI ₁ , the process proceeds to step ST12.

In step ST12, converting means 91b (see FIG. 2), image data DI ₁ obtained by Fourier transform to convert the data of the k-space (see FIG. 7).

FIG. 7 is a diagram showing how the image data DI ₁ is converted into k-space data.
7 (a) is the image data DI ₁ shows a (FIG. 6 (b) refer), FIG. 7 (b), data DK of k-space obtained by performing Fourier transformation on the image data DI _{1 1} ′ Is shown.

In FIG. 7B, the k-space data DK ₁ ′ is divided into two regions A and B based on a threshold value TH that is a criterion for whether the amplitude of the k-space data DK ₁ ′ is large or small. The region B is a region where the amplitude of the data DK ₁ ′ in the k space is smaller than the threshold value TH and is indicated by cross hatching. On the other hand, the region A is a region in which the amplitude of the k-space data DK ₁ ′ is larger than the threshold value TH and is surrounded by a thick solid line. In the region A, a point having a particularly large amplitude is indicated in white, and a point having an amplitude smaller than that of the white point is indicated by a symbol “Δ”.

The area specifying unit 91c (see FIG. 2) specifies the k-space area A in which data with large amplitude is highly likely to be concentrated based on the k-space data DK ₁ ′. Since the area A is an area having an amplitude larger than the threshold value TH, it is considered that the area having a high possibility that data having a large amplitude is concentrated. On the other hand, since the region B is a region having an amplitude smaller than the threshold value TH, it is considered that the region where data having a large amplitude is less likely to concentrate. Therefore, in the first embodiment, the region specifying unit 91c specifies the region A as a region where data with large amplitude is highly likely to concentrate.

  Therefore, by executing step ST1, it is possible to predict a region A where data with large amplitude is highly likely to concentrate.

  In the first form, the sampling pattern is updated based on this prediction. In order to update the sampling pattern, the process proceeds to step ST2.

  In step ST2, the sampling pattern update unit 92 (see FIG. 1) updates the sampling pattern based on the predicted region A. A method for updating the sampling pattern will be described below.

  The sampling pattern updating unit 92 updates the initially set k-space sampling pattern SP (see FIG. 4) so that points with high amplitude in the k-space data can be preferentially sampled.

FIG. 8 is a diagram showing the updated k-space sampling pattern SP ′.
Note that the regions A and B shown in the updated sampling pattern SP ′ in the k space represent the regions A and B shown in FIG. 7B, respectively.

  In FIG. 8, black circles “●” and white circles “◯” represent the same sampling points as the sampling pattern SP (see FIG. 4) in the initially set k space. Sampling points with black circles “●” represent points that have already been sampled, and sampling points with white circles “O” represent points that have not yet been sampled. A double circle “◎” represents a sampling point newly generated by updating the sampling pattern. In the first embodiment, since new sampling points “◎” are generated in the region A where data with large amplitude is highly likely to be concentrated, all points included in the region A are set as sampling points. And a high quality image can be obtained.

Note that if there are too many newly generated sampling points, the scan time may be extended. Therefore, in the first embodiment, even if a new sampling point is generated, in order not to extend the scan time, sampling in which data has not yet been collected among the sampling points that are initially set shown in FIG. Delete some of the points. In FIG. 8, the deleted sampling points are indicated by broken lines. Here, the sampling points located in the region B are deleted. Since the region B is considered to be a region where data with a large amplitude is unlikely to be concentrated, the sampling point located in the region B is considered to have little influence on the image quality. Therefore, even if the sampling points located in the region B are deleted, a sufficiently high quality image can be obtained. When the sampling pattern is updated, the scan is continued according to the updated sampling pattern SP ′ in the k space after the time point t ₁ ′ (see FIG. 3). When the number of sampled data becomes N ₂ , the second processing flow F ₂ is executed. Note that N ₂ can be, for example, N ₂ = 2N ₁ .

In the second process flow _{F 2,} in step ST11, the image reconstruction unit 91a is, the data of the scan start time _{t 0} ~ process flow _{F 2} at the start sampled _{N 2} pieces of k-space during the _{t 2} using, to reconstruct an image at time t _2. When the image is reconstructed, the process proceeds to step ST12.

In step ST12, image data is converted into k-space data in the same procedure as in the _first process flow F1, and a region A where data with large amplitude is highly likely to be concentrated is obtained. In step ST2, the sampling pattern is updated. Therefore, after time t ₂ ′, scanning is continued according to the sampling pattern of the k space updated by the _second processing flow F ₂ .

Similarly, when the number of sampled data reaches N _i , the i-th processing flow F _i is executed and the sampling pattern is updated. For example, at time t _k , the number of sampled data reaches N _k , so the k-th processing flow F _k is executed and the sampling pattern is updated.

In the k-th process flow F _k , in step ST11, the image reconstruction unit 91a stores N _k pieces of k-space data sampled between the scan start time t ₀ and the process flow F _k start time t _k. using, to reconstruct an image at time t _k. When the image is reconstructed, the process proceeds to step ST12.

In step ST12, image data is converted into k-space data in the same procedure as in the _first process flow F1, and a region A where data with large amplitude is highly likely to be concentrated is obtained. In step ST2, the sampling pattern is updated. Therefore, after time t _k ′, scanning is continued according to the sampling pattern of the k space updated by the k-th processing flow F _k .

Even after the k-th processing flow F _k is executed, when the number of sampled data reaches N _i , the i-th processing flow F _i is executed. For example, since the number of sampled data reaches N _z at the time point t _z , the z-th processing flow F _z is executed, and the sampling pattern is updated. Therefore, after time t _z ′, scanning is continued according to the sampling pattern of the k space updated by the z-th processing flow F _z . Then, at time t _e, which can be sampled all the data required sampling points, and ends the scan.

FIG. 9 is a diagram illustrating an example of the final k-space data DK _e obtained by scanning. In FIG. 9, the points at which data is sampled are indicated by black circles “●”.

When the scan is finished, the final k-space data DK _e is converted into image data DI _e (see FIG. 10), and the flow is finished.

  In the first mode, it is possible to predict a region where data with a large amplitude is highly likely to be concentrated during scanning, so that a high-quality image can be acquired.

  In the above description, a three-dimensional k-space is described, but the present invention is also applicable to a two-dimensional k-space.

(2) Second embodiment The MRI apparatus of the second embodiment is different from the MRI apparatus 100 of the first embodiment in the prediction means, but the other configurations are the same. As for the MRI apparatus, the prediction means will be mainly described.

FIG. 11 is a block diagram of the prediction means 91 of the MRI apparatus according to the second embodiment.
FIG. 11A is a block diagram of the prediction unit 91, and FIG. 11B is a block diagram of the filter unit 911 shown in FIG. 11A.

  The prediction unit 91 includes an image reconstruction unit 91a, a conversion unit 91b, a filter unit 911, and an area specifying unit 91c.

  The image reconstruction unit 91a reconstructs an image using the sampled k-space data during the scan.

  The conversion unit 91b converts the image data obtained by the image reconstruction unit 91a into k-space data.

  The filter unit 911 filters the k-space data obtained by the conversion unit 91b. As shown in FIG. 11B, the filter unit 911 includes an inspection range moving unit 911a and a feature amount calculating unit 911b.

  The inspection range moving unit 911a moves the inspection range when obtaining the feature amount of the k-space data obtained by the conversion unit 91b.

  The feature amount calculation unit 911b calculates the feature amount of the k-space data in the inspection range at each movement position of the inspection range.

  Based on the k-space data obtained by the filter unit 911, the area specifying unit 91c specifies a k-space area where data with large amplitude is highly likely to concentrate.

  The prediction means 91 is configured as described above. The central processing unit 9 is an example of the prediction unit 91 shown in FIG. 11, and functions as this unit by executing a predetermined program. Next, the flow when scanning the subject 12 in the second embodiment will be described with reference to FIG.

  FIG. 12 is an explanatory diagram of a flow when scanning the subject 12 in the second embodiment.

In the second embodiment, the step ST1 in the processing flow _F 1 to F _z, except that the step ST13 is added, is the same as the first embodiment. Therefore, in the description of FIG. 12, step ST13 will be mainly described.

Start scanning at time t _0, the sampled data count when it reaches the N _1, 1-time processing flow F ₁ is started.

Note that steps ST11 and ST12 of the processing flow _{F 1} is the same as the first embodiment description is omitted. In step ST12, after converting the image data DI ₁ to data DK ₁ 'k-space (see FIG. 7), the process proceeds to step ST13.

In step ST13, the k-space data DK ₁ 'is filtered. Hereinafter, an example of the filtering process will be described.

In step ST13, first, the inspection range moving unit 911a (see FIG. 11B) sets an inspection range for calculating the feature amount of the k-space data DK ₁ ′.

FIG. 13 is a diagram showing the inspection range.
In FIG. 13, the inspection range C is set so that the center of the inspection range C is located in the upper left corner data of the k space. When the inspection range C is set, the feature amount calculation unit 911b (see FIG. 11B) obtains the sum of the amplitudes of the respective data included in the inspection range C. In Figure 13, the inspection area C, because it contains data _D 11 _{to D 31,} obtains an amplitude sum of these data _D 11 _{to D 31.} Here, the amplitude sum and _{A 11.} Feature calculating unit 911b at a position corresponding to the center of the examination region C, assigns the amplitude sum _{A 11.}

After determining the amplitude sum _{A 11,} the inspection range moving means 911a moves the check range C data one minute only (see Figure 14).

FIG. 14 is a diagram illustrating a state after the inspection range C is moved.
In FIG. 14, the inspection range C before movement is indicated by a broken line, and the inspection range C after movement is indicated by a solid line. The feature amount calculation unit 911b calculates the sum of the amplitudes of the respective data included in the inspection range C after movement. In Figure 14, the inspection area C, since the data _D 11 _{to D 32} are included to determine the amplitude sum of these data _D 11 _{to D 32.} Here, the amplitude sum and _{A 12.} Feature calculating unit 911b at a position corresponding to the center of the examination region C, assigns the amplitude sum _{A 12.}

After determining the amplitude sum A _12, the same procedure, while moving the test range, calculates an amplitude sum (see Figure 15).

  FIG. 15 is a diagram illustrating a state where the inspection range C has moved to a position slightly above the left of the center of the k space.

The feature amount calculation unit 911b calculates the sum of the amplitudes of the data included in the inspection range C. In Figure 15, the inspection area C, because it contains data _D 46 _{to D 86,} obtains an amplitude sum of these data _D 46 _{to D 86.} Here, the amplitude sum is _A66 . The feature amount calculating unit 911b assigns the amplitude sum A ₆₆ to the position corresponding to the center of the inspection range C.

  Similarly, the amplitude sum is calculated by moving the inspection range C over the entire area of the k-space. Therefore, a map of the sum of amplitudes of k-space data can be obtained. FIG. 16 schematically shows the obtained amplitude sum map M.

  In FIG. 16, a region B in which the amplitude sum is smaller than the threshold value TH is indicated by cross hatching, and a region A in which the amplitude sum is greater than the threshold value TH is indicated by being surrounded by a thick solid line. In the region A, a point having a particularly large amplitude sum is shown in white, a point having a smaller amplitude sum than a white point is indicated by a symbol “Δ”, and a point having a smaller amplitude sum is This is indicated by the symbol “x”.

  Since the area A is an area where the sum of amplitudes is larger than the threshold value TH, it is considered that the area with high possibility that data with large amplitudes is concentrated. On the other hand, since the region B is a region where the sum of amplitudes is smaller than the threshold value TH, it is considered that the region where data with large amplitudes is less likely to concentrate. Therefore, the area specifying unit 91c (see FIG. 11A) specifies the area A as an area where data with large amplitude is highly likely to concentrate. Therefore, by creating the amplitude sum map M, it is possible to predict a region A where data with a large amplitude is highly likely to concentrate.

  In the second mode, the sampling pattern is updated based on this prediction. In order to update the sampling pattern, the process proceeds to step ST2.

  In step ST2, the sampling pattern update unit 92 (see FIG. 1) updates the sampling pattern based on the predicted region A. A method for updating the sampling pattern will be described below.

  The sampling pattern update unit 92 updates the initially set sampling pattern SP (see FIG. 4) so as to preferentially sample points where the amplitude of the k-space data is high.

FIG. 17 is a diagram illustrating the updated k-space sampling pattern SP ′.
Note that regions A and B shown in the updated k-space sampling pattern SP ′ represent the regions A and B shown in FIG. 16, respectively.

  In FIG. 17, black circles “●” and white circles “◯” represent the same sampling points as the sampling pattern SP (see FIG. 4) in the initially set k space. Sampling points with black circles “●” represent points that have already been sampled, and sampling points with white circles “O” represent points that have not yet been sampled. A double circle “◎” represents a sampling point newly generated by updating the sampling pattern. In the second embodiment, a new sampling point “◎” is generated in the region A where data with large amplitude is highly likely to concentrate. Therefore, since the points included in the region A where data with a large amplitude is highly likely to be concentrated can be preferentially sampled, a high-quality image can be obtained. In the second embodiment, some points in the area A are excluded from the sampling points so that the sampling points do not concentrate too much on a specific area.

Further, as in the first embodiment, even if a new sampling point is generated, data is not yet collected among the initially set sampling points shown in FIG. 4 so that the scan time is not extended. Remove some missing sampling points. In FIG. 17, the deleted sampling points are indicated by broken lines. Here, the sampling points located in the region B are deleted. Since the region B is considered to be a region where data with a large amplitude is unlikely to be concentrated, the sampling point located in the region B is considered to have little influence on the image quality. Therefore, even if the sampling points located in the region B are deleted, a sufficiently high quality image can be obtained. When the sampling pattern is updated, the scan is continued in accordance with the updated sampling pattern SP ′ in the k space after the time point t ₁ ′ (see FIG. 12). When the number of sampled data becomes N ₂ , the second processing flow F ₂ is executed.

Hereinafter, similarly, when the number of sampled data reaches N _i , the i-th processing flow F _i is executed, and the scan is performed at the time t _te when the data of all necessary sampling points can be sampled. finish. When the scan is finished, the final image data is obtained by reconstructing the image using the final k-space data obtained by the scan.

  Also in the second mode, similarly to the first mode, it is possible to predict a region where data with a large amplitude is highly likely to be concentrated in the middle of a scan, so that a high-quality image can be acquired.

  In the second mode, a map M of feature quantities of k-space data is created, and the sampling pattern is updated based on the map M. As described above, when updating the sampling pattern, the feature amount map M of the k-space data may be referred to.

  In the second embodiment, the amplitude sum is obtained as the k-space data and the feature amount. However, another feature amount such as an average value of amplitude may be obtained.

(3) Third embodiment The MRI apparatus of the third embodiment is different from the MRI apparatus 100 of the first embodiment in the prediction means, but the other configurations are the same. As for the MRI apparatus, the prediction means will be mainly described.

FIG. 18 is a block diagram of the prediction means 91 of the MRI apparatus of the third embodiment.
18A is a block diagram of the prediction unit 91, and FIG. 18B is a block diagram of the filter unit 911 shown in FIG. 18A.

  The prediction unit 91 includes a zero filling unit 910, a filter unit 911, and a region specifying unit 91c.

  The zero filling unit 910 performs zero filling on a region in which k-space data is not sampled during the scan.

  The filter unit 911 filters the zero-filled k-space data. As shown in FIG. 18B, the filter unit 911 includes an inspection range moving unit 911a and a feature amount calculating unit 911b.

  The inspection range moving unit 911a moves the inspection range when obtaining the feature amount of the zero-filled k-space data.

  The feature amount calculation unit 911b calculates the feature amount of the k-space data in the inspection range at each movement position of the inspection range.

  Based on the k-space data obtained by the filter unit 911, the area specifying unit 91c specifies a k-space area where data with large amplitude is highly likely to concentrate.

  The prediction means 91 is configured as described above. The central processing unit 9 is an example of the prediction unit 91 shown in FIG. 18, and functions as this unit by executing a predetermined program. Next, the flow when scanning the subject 12 in the third embodiment will be described with reference to FIG.

  FIG. 19 is an explanatory diagram of a flow when scanning the subject 12 in the third embodiment.

In the third embodiment, in step ST1 of the processing flow _F 1 to F _z, instead of steps ST11 and ST12, except that the step ST10 is provided is the same as the second embodiment. Therefore, in the description of FIG. 19, step ST10 will be mainly described.

Start scanning at time t _0, at time t _1, the sampled data count reaches the N _1, and the data DK ₁ k-space as shown in FIG. 5 were obtained.

Sampled data count when it reaches the N _1, 1-time processing flow F ₁ is started.

In process flow _{F 1,} first, in step ST10, the zero-filling section 910 (see FIG. 18 (a)) is to place a zero in the region that is not sampled data DK ₁ of k space shown in Fig. 5 (Fig. 20 reference).

Figure 20 is a diagram showing a data DK ₁₀ of k-space after placing zero.
In FIG. 20, the points where sampling is performed are represented by black circles “●”, and regions where zeros are arranged are indicated by cross hatching. After arranging zero, the process proceeds to step ST13.

In step ST13, the data _{DK 10} k-space filtering. The filtering process is performed by obtaining the amplitude sum while moving the inspection range C and creating a map of the amplitude sum as in the second embodiment. If the map of amplitude sum is created, it will progress to step ST2 and will update a sampling pattern.

Hereinafter, similarly, when sampled data count reaches the N _i number, executes the processing flow F _i, at time t _e, which can be sampled all the data required sampling points, and ends the scan. When the scan is finished, the final image data is obtained by reconstructing the image using the final k-space data obtained by the scan.

  Even in the third mode, it is possible to predict a region where data with a large amplitude is highly likely to be concentrated in the middle of scanning, so that a high-quality image can be acquired.

2 Magnetic field generator 3 Table 4 Receiving coil 5 Sequencer 6 Transmitter 7 Gradient magnetic field power supply 8 Receiver 9 Central processing unit 10 Operation unit 11 Display unit 12 Subject 13 Operator 21 Bore 22 Superconducting coil 23 Gradient coil 24 Transmitting coil 31 Cradle 91 Prediction means 91a Image reconstruction means 91b Conversion means 91c Region specification means 92 Sampling pattern update means 100 MRI apparatus 910 Zero filling means 911 Filter means 911a Inspection range moving means 911b Feature quantity calculation means

Claims (16)

A magnetic resonance imaging apparatus for performing a scan for sampling k-space data according to a sampling pattern,
Prediction means for predicting a k-space region where data with a large amplitude is highly likely to be concentrated based on sampled k-space data during the scan;
Sampling pattern updating means for updating the sampling pattern based on the region predicted by the prediction means;
A magnetic resonance imaging apparatus.   The magnetic resonance imaging apparatus according to claim 1, wherein the sampling pattern is updated once or a plurality of times during the scan. The prediction means includes
Image reconstruction means for reconstructing an image using sampled k-space data during the scan;
Conversion means for converting the image data obtained by the image reconstruction means into k-space data;
Area specifying means for specifying a k-space area where data having a large amplitude is highly likely to be concentrated based on the k-space data obtained by the converting means;
The magnetic resonance imaging apparatus according to claim 2, comprising: The image reconstruction means includes
The magnetic resonance imaging apparatus according to claim 3, wherein when the number of sampled data reaches N _i (i is an integer of 1 to z), an image is reconstructed using the sampled k-space data. . The region specifying means includes
4. The k-space region where data having a large amplitude is highly likely to concentrate is identified based on a threshold value that is a criterion for whether the amplitude of the k-space data obtained by the conversion unit is large or small. 5. The magnetic resonance imaging apparatus according to 4. The prediction means includes
Image reconstruction means for reconstructing an image using sampled k-space data during the scan;
Conversion means for converting the image data obtained by the image reconstruction means into k-space data;
Filter means for filtering k-space data obtained by the conversion means;
An area specifying means for specifying an area of k-space where high-amplitude data is highly likely to be concentrated based on the k-space data obtained by the filter means;
The magnetic resonance imaging apparatus according to claim 2, comprising: The image reconstruction means includes
The magnetic resonance imaging apparatus according to claim 6, wherein when the number of sampled data reaches N _i (i is an integer of 1 to z), an image is reconstructed using the sampled k-space data. . The region specifying means includes
The region of the k space where data having a large amplitude is highly likely to be concentrated is specified based on a threshold value which is a criterion for whether the amplitude of the k space data obtained by the filter means is large or small. 8. A magnetic resonance imaging apparatus according to 7. The filter means includes
Inspection range moving means for moving the inspection range when obtaining the feature quantity of the k-space data obtained by the conversion means;
A feature amount calculating means for calculating a feature amount of data in the k space in the inspection range at each movement position of the inspection range;
The magnetic resonance imaging apparatus according to claim 6, comprising:   The magnetic resonance imaging apparatus according to claim 9, wherein the feature amount is a sum of amplitudes of k-space data. The prediction means includes
Zero filling means for performing zero filling on an area in which k-space data is not sampled during the scan;
Filter means for filtering zero-filled k-space data;
An area specifying means for specifying an area of k-space where high-amplitude data is highly likely to be concentrated based on the k-space data obtained by the filter means;
The magnetic resonance imaging apparatus according to claim 2, comprising: The zero filling means includes
The magnetic resonance imaging apparatus according to claim 11, wherein zero filling is performed when the number of sampled data reaches N _i (i is an integer of 1 to z). The region specifying means includes
12. A region of k space where data with a large amplitude is highly likely to be concentrated is identified based on a threshold value that is a criterion for whether the amplitude of the k space data obtained by the filter means is large or small. 12. The magnetic resonance imaging apparatus according to 12. The filter means includes
Inspection range moving means for moving the inspection range when obtaining the feature amount of the zero-filled k-space data;
A feature amount calculating means for calculating a feature amount of data in the k space in the inspection range at each movement position of the inspection range;
The magnetic resonance imaging apparatus according to claim 11, comprising:   The magnetic resonance imaging apparatus according to claim 14, wherein the feature amount is a sum of amplitudes of k-space data. A program of a magnetic resonance imaging apparatus that performs a scan for sampling k-space data according to a sampling pattern,
Prediction processing for predicting a k-space region where data with a large amplitude is highly likely to be concentrated based on sampled k-space data in the middle of the scan;
A sampling pattern update process for updating the sampling pattern based on the region predicted by the prediction process;
A program to make a computer execute.

Images