JP5306413B2 - Magnetic resonance imaging system - Google Patents

Description

  The present invention relates to a magnetic resonance imaging apparatus that reconstructs an image of a subject based on a magnetic resonance signal emitted from the subject.

  In magnetic resonance imaging, a nuclear spin of a subject placed in a static magnetic field is magnetically excited with a high-frequency signal of the Larmor frequency, and an image of the subject is obtained from a magnetic resonance signal (MR signal) generated by this excitation. Is an imaging method for reconstructing.

  FIG. 10 is a diagram showing an example of the flow of the main part of the process for reconstructing a three-dimensional (3D) image.

  As shown in FIG. 10, the reconstruction is realized by a combination of many processes. These processes are prepared as individual program modules, and are called and executed in a predetermined order. Each program module describes a processing procedure from input of data necessary for processing to output of processed data.

  FIG. 10 simply shows the flow of the main body part without the loop structure. In actual reconstruction, for example, a loop structure as shown in FIG. 11 is applied.

  Such a loop structure includes a large number of loops, but is managed as a series of processes as shown in FIG.

  As described above, all loops have been managed as a series of processes in the past, so even if you want to change some of the processes, you need to review the entire management procedure, and easily handle them. It was difficult.

  The present invention has been made in consideration of such circumstances, and an object of the present invention is to make it possible to easily change some of the processes.

The present invention combines means for receiving a magnetic resonance signal radiated from a subject and a plurality of program modules that execute different processes based on the received magnetic resonance signal while applying a plurality of loops. A magnetic resonance imaging apparatus comprising: a reconstruction unit configured to reconstruct an image of the subject by reconstruction processing; and a plurality of main body procedures for executing each of the plurality of program modules in the reconstruction unit. And a first input / output procedure for data input / output related to conversion involving a change in data type and size, and a second input / output for data input / output related to arithmetic processing without data type / size change. In executing one of the main body procedures, if the main body procedure to be executed relates to the conversion process, The first input-output procedure, and execution means for performing data input and output by the second output procedure if body procedure relating to the processing of the execution, the plurality of program modules of the plurality of loops Managing means for managing a loop applied for each repetition of the first program module and a second program module using data obtained by the first program module among the plurality of program modules And means for managing a loop different from that managed by the management means among the plurality of loops.

The figure which shows the structure of the magnetic resonance imaging apparatus (MRI apparatus) which concerns on one Embodiment of this invention. The figure which shows an example of the internal structure of the reconstruction part 13 in FIG. The figure which shows the classification | category of the program module for the process part 13d in FIG. 2 to perform various main body processes. The flowchart which shows the process sequence of the scheduler 13c. The flowchart which shows the process sequence of the process part 13d in FIG. 2 at the time of performing DC. The flowchart which shows the process sequence of the process part 13d in FIG. 2 based on the input / output module M11 shown in FIG. The flowchart which shows the process sequence of the process part 13d in FIG. 2 based on the input / output module M21 shown in FIG. The block diagram which shows the modification structural example of the reconstruction part 13 in FIG. The block diagram which shows the modification structural example of the reconstruction part 13 in FIG. The figure which shows an example of the flow of the main-body part of the process for reconstructing a three-dimensional (3D) image. The figure which shows the loop structure of the process of the main-body part shown in FIG.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram showing a configuration of a magnetic resonance imaging apparatus (MRI apparatus) according to the present embodiment. The MRI apparatus shown in FIG. 1 includes a static magnetic field magnet 1, a gradient magnetic field coil 2, a gradient magnetic field power source 3, a bed 4, a bed control unit 5, a transmission RF coil 6, a transmission unit 7, a reception RF coil 8, a reception unit 9, A gantry control unit 10, a collection control unit 11, a data input unit 12, a reconstruction unit 13 and a database 14 are provided.

  The static magnetic field magnet 1 has a hollow cylindrical shape and generates a uniform static magnetic field in an internal space. As the static magnetic field magnet 1, for example, a permanent magnet, a superconducting magnet or the like is used.

  The gradient magnetic field coil 2 has a hollow cylindrical shape and is disposed inside the static magnetic field magnet 1. The gradient coil 2 is a combination of three coils corresponding to the X, Y, and Z axes orthogonal to each other. The gradient coil 2 generates a gradient magnetic field in which the above three coils are individually supplied with electric current from the gradient magnetic field power source 3 and the magnetic field strength is inclined along the X, Y, and Z axes. The Z-axis direction is, for example, the same direction as the static magnetic field. The gradient magnetic fields of the X, Y, and Z axes correspond to, for example, the slice selection gradient magnetic field Gs, the phase encoding gradient magnetic field Ge, and the readout gradient magnetic field Gr, respectively. The slice selection gradient magnetic field Gs is used to arbitrarily determine an imaging section. The phase encoding gradient magnetic field Ge is used to encode the phase of the magnetic resonance signal in accordance with the spatial position. The readout gradient magnetic field Gr is used to encode the frequency of the magnetic resonance signal in accordance with the spatial position.

  The subject P is inserted into the cavity (imaging port) of the gradient coil 2 while being placed on the top 41 of the bed 4. The top 41 of the bed 4 is driven by the bed control unit 5 and moves in the longitudinal direction and the vertical direction. Usually, the bed 4 is installed such that the longitudinal direction thereof is parallel to the central axis of the static magnetic field magnet 1.

  The transmission RF coil 6 is disposed inside the gradient magnetic field coil 2. The transmission RF coil 6 receives a high frequency pulse from the transmission unit 7 and generates a high frequency magnetic field.

  The transmission unit 7 includes an oscillation unit, a phase selection unit, a frequency conversion unit, an amplitude modulation unit, a high frequency power amplification unit, and the like. The oscillation unit generates a high-frequency signal having a resonance frequency unique to the target nucleus in the static magnetic field. The phase selection unit selects the phase of the high-frequency signal. The frequency conversion unit converts the frequency of the high-frequency signal output from the phase selection unit. The amplitude modulation unit modulates the amplitude of the high-frequency signal output from the frequency modulation unit, for example, according to a sync function. The high frequency power amplification unit amplifies the high frequency signal output from the amplitude modulation unit. As a result of the operation of each of these units, the transmission unit 7 transmits a high-frequency pulse corresponding to the Larmor frequency to the transmission RF coil 6.

  The reception RF coil 8 is disposed inside the gradient magnetic field coil 2. The reception RF coil 8 receives a magnetic resonance signal radiated from the subject due to the influence of the high frequency magnetic field. An output signal from the reception RF coil 8 is input to the reception unit 9.

  The receiving unit 9 generates magnetic resonance signal data based on the output signal from the reception RF coil 8.

  The static magnetic field magnet 1, the gradient magnetic field coil 2, the gradient magnetic field power source 3, the bed control unit 5, the transmission RF coil 6, the transmission unit 7, the reception RF coil 8, and the reception unit 9 constitute a gantry.

  The gantry control unit 10 comprehensively controls the operation of each of the above units in order to obtain magnetic resonance data necessary for reconstructing a required image. The gantry control unit 10 gives the magnetic resonance data output from the receiving unit 9 to the data input unit 12.

  The collection control unit 11 requests the gantry control unit 10 to acquire magnetic resonance data required by the reconstruction unit 13.

  The data input unit 12 performs processing for inputting the magnetic resonance data that has been collected, that is, so-called raw data, to the reconstruction unit 13 via the gantry control unit 10.

  The reconstruction unit 13 performs image reconstruction based on the magnetic resonance data input by the data input unit 12.

  The database 14 stores image data reconstructed by the reconstruction unit 13.

  FIG. 2 is a diagram illustrating an example of the internal structure of the reconstruction unit 13.

  The reconfiguration unit 13 includes a data stacker 13a, an event generator 13b, a scheduler 13c, and a processing unit 13d. The reconfiguration unit 13 can use, for example, a general-purpose computer device as basic hardware. The data stacker 13a, the event generator 13b, the scheduler 13c, and the processing unit 13d can be realized by causing a processor mounted on the computer device to execute a reconfiguration program. At this time, the reconstruction unit 13 may be realized by installing the reconstruction program in the computer device in advance, or a removable recording medium such as a magnetic disk, a magneto-optical disk, an optical disk, or a semiconductor memory. It may be realized by recording the above-mentioned reconfiguration program over the network or by distributing the reconfiguration program to a computer device as appropriate. Note that part or all of the above-described units can be realized by hardware such as a logic circuit. Each of the above-described units can also be realized by combining hardware and software control. The data stacker 13a includes a memory device such as a memory or a hard disk device, a part of which is built in the computer device, a memory device such as a memory or a hard disk device externally attached to the server device or the computer device, It is realized by appropriately using a removable recording medium such as a magnetic disk, a magneto-optical disk, and an optical disk.

  The data stacker 13a stacks the raw data input by the data input unit 12. The data stacker 13a stacks image groups that are generated intermediately during processing by the processing unit 13d. The data stacker 13a rearranges the stacked raw data and image groups.

  The event generator 13b monitors the operation of the data input unit 12 and detects the timing when the raw data necessary for the processing in the processing unit 13d is complete. The event generator 13b detects that necessary raw data is available regardless of the operation state of the data input unit 12 at the time of batch reconstruction.

  The scheduler 13c schedules processing in the processing unit 13d. The scheduler 13c manages the loops in which the repeated processes among the plurality of loops included in the reconfiguration process have no influence on each other. The scheduler 13c creates tag information. The tag information includes the combination contents of the arithmetic processing and conversion processing processed by the processing unit 13d, information on the order thereof, and the like.

  The processing unit 13d manages a loop in which the repeated processes among the plurality of loops included in the reconstruction process influence each other. For example, as illustrated in FIG. 10, the processing unit 13d executes various processes for realizing reconfiguration (hereinafter referred to as main processing) under the loop management by the self and the loop management by the scheduler 13c. To do. The processing unit 13d has a function of post-processing the image group and registering the post-processed image group in the database 14.

  By the way, in this MRI apparatus, the processing unit 13d individually prepares program modules for executing various main body processes. The program module is stored in a storage device such as a memory or a hard disk device built in or externally attached to the computer device, and is called up and used by the processing unit 13d as appropriate. These program modules are classified into a module group M1 for arithmetic processing and a module group M2 for conversion processing as shown in FIG.

  An input / output module M11 and a plurality of main body modules M12 belong to the module group M1. The input / output module M11 describes input / output processing of information related to main body processing classified as arithmetic processing. The main body module M12 describes processing other than input / output of information among main body processing classified as arithmetic processing.

  The module group M2 includes an input / output module M21 and a plurality of main body modules M22. The input / output module M21 describes input / output processing of information related to main body processing classified as conversion processing. The main body module M22 describes processing other than input / output of information among main body processing classified as conversion processing.

  Note that the arithmetic processing refers to processing in which the data value changes and the data shape and size do not change. The conversion process refers to a process involving a change in data shape or size. In the case of the main body processing shown in FIG. 10, phase correction (TEMPLATE, NAVI, 0th order, primary), signal value correction (TEMPLATE (also T2), NAVI, noise power, brightness correction), DC correction (DC value), Roll-off correction, filter (T2Decay, leveler, Kspace, Refine, GA), unnecessary signal removal (FID suppression processing, residual correction, whiskers), synthesis (PAC, SPEEDER), scaling (maximum, minimum, scale factor), Registers information in the database (image, operation value, icon), determines reconstruction order (priority), reads raw data (address, timing, ReferenceView, ViewShare), and calculates between images (TOF3D, FISP, Diffusion) The conversion process includes raw data rearrangement (sort table), FFT (there are various types of input, but no distinction), image shift, image conversion (reduction, cutout, rotation, icon creation) and the like.

  Next, the operation of the MRI apparatus configured as described above will be described. Here, the three-dimensional image shown in FIG. 10 will be specifically described focusing on the case where the reconstruction process is executed in a loop structure as shown in FIG. Since the main body process included in FIG. 10 is a well-known process that has been conventionally used, detailed description thereof is omitted here.

  The loop structure shown in FIG. 11 includes loops of stage, dynamic, diffusion, echo, slab, subprotcol, slicegroup, coilNum, psnAck, MS_PS_MAX_FLOW_CMP, and numMax.

  The stage loop repeatedly performs processing for each stage defined by a protocol such as AMB. The dynamic loop repeatedly performs processing for each of a plurality of collected data having different time phases for the same position. The diffusion loop repeatedly performs processing for each different B value. The echo loop repeatedly performs processing for each different TE (echo time). The slab loop iterates over each slice of different slices. The loop of subprotcol repeatedly performs processing for each different number of scans. The slicegroup loop repeatedly performs processing for each of different chunks of slices within one scan. The coilNum loop repeatedly performs processing for each of the different channels. The psnAck loop repeatedly performs processing for each of TEs having different PS (phase shift). The MS_PS_MAX_FLOW_CMP loop repeatedly performs processing for each output image of a different PS. The numMax loop repeatedly performs processing for each of the different quad multiplexing (output images).

  Among these loops, the loops of stage, dynamic, diffusion, echo, slab, subprotcol, and slicegroup have no effect on the processes repeated there. In each loop of coilNum, psnAck, MS_PS_MAX_FLOW_CMP, and numMax, processes repeated there affect each other.

  Therefore, the scheduler 13c manages the loops of stage, dynamic, diffusion, echo, slab, subprotcol, and slicegroup by the process shown in FIG.

  Although some illustrations are omitted, in steps Sa1 to Sa7, the scheduler 13c sequentially selects a stage to be processed, acquired data, B value, TE, slice block, scan, and slice group.

  In step Sa8, the scheduler 13c requests the processing unit 13d to execute DC shown in FIG. In step Sa9, the scheduler 13c waits for a completion notification from the processing unit 13d.

  If the execution of DC is requested, the processing unit 13d executes a process as shown in FIG.

  In step Sb1 and step Sb2, the processing unit 13d selects a coil to be processed and a TE (PS echo) of PS, respectively.

  In step Sb3, the processing unit 13d sets the stage, collection data, B value, TE, slice lump, scan, slice group selected by the scheduler 13c, and the coil and PS echo selected above as a processing target to DC. The included main body processing is executed according to the procedure shown in FIG.

  If the DC has been executed once, in step Sb4, the processing unit 13d confirms whether or not all PS echoes to be selected as processing targets have been selected. If there is an unselected TE, the processing unit 13d returns from step Sb4 to step Sb2. Then, the processing unit 13d repeatedly executes DC while changing the TE to be processed.

  When all the PS echoes to be selected as the processing target have been selected, the processing unit 13d proceeds from step Sb4 to step Sb5. In step Sb5, the processing unit 13d confirms whether or not all of the coils to be selected as processing targets have been selected. If there is an unselected coil, the processing unit 13d returns from step Sb5 to step Sb1. And the process part 13d repeatedly performs the process of step Sb2 thru | or step Sb5, changing the coil made into a process target.

  When all the coils to be selected as processing targets have been selected, the processing unit 13d proceeds from step Sb5 to step Sb6. In step Sb6, the processing unit 13d notifies the scheduler 13c of the completion of processing. With this, the processing unit 13d ends the processing shown in FIG.

  Thus, the processing unit 13d manages the coilNum loop and the psnAck loop.

  When the completion is notified from the processing unit 13d to the scheduler 13c as described above, the scheduler 13c proceeds from step Sa9 to step Sa10. In step Sa10, the scheduler 13c requests the processing unit 13d to execute SE shown in FIG. In step Sa11, the scheduler 13c waits for a completion notification from the processing unit 13d.

  When the execution of SE is requested, the processing unit 13d executes SE in the same manner as in the case of DC. At this time, the processing unit 13d manages the PS echo loop and the coil loop in the same manner as in the case of DC, and notifies the scheduler 13c of completion when the coil loop is completed.

  After the completion of the notification from the processing unit 13d to the scheduler 13c in this manner, the scheduler 13c is not shown, but in steps Sa12 to Sa19, ROPE is executed, PS echo is read, and PS output image is output. The processing unit 13d is sequentially requested for writing and writing of the quad output image while waiting for notification of completion of each processing. When executing ROPE, reading PS echo, writing PS output image, and writing Quad output image, coilNum loop, psnAck loop, MS_PS_MAX_FLOW_CMP loop, and numMax loop are executed respectively. This loop is managed by the processing unit 13d.

  When the completion of writing of the quad output image is notified from the processing unit 13d to the scheduler 13c, the scheduler 13c proceeds to Step Sa20. In step Sa20, the scheduler 13c confirms whether selection of all slice groups to be selected has been completed. If there is an unselected slice group, the scheduler 13c returns from step Sa20 to step Sa7. Then, the scheduler 13c repeatedly executes the processing from step Sa7 to step Sa20 while changing the slice group to be processed.

  When all the slice groups to be selected as processing targets have been selected, the scheduler 13c proceeds from step Sa20 to step Sa21. Although some of the illustrations are omitted later, in steps Sa21 to Sa26, the scheduler 13c selects all the scans, chunks of slices, TE, B values, collected data, and stages to be selected as processing targets. Check if the selection is complete. In addition, it progresses to each step of step Sa22 thru | or step Sa26, when it determines with YES in the last step. If NO is determined in each of the steps Sa21 to Sa26, the process returns to the steps Sa6, Sa5, Sa4, Sa3, Sa2, Sa1, respectively, and each of the scan, the lump of slices, the TE, the B value, the collected data, and the stage Select again and repeat the subsequent processing.

  If the scheduler 13c determines in step Sa26 that all stages to be selected as processing targets have been selected, the process shown in FIG.

  By the way, when executing the main body process, the processing unit 13d calls the main body modules M12 and M22 corresponding to the main body process to be executed from either the arithmetic processing module group M1 or the conversion processing module group M2, and Execute. However, processing related to data input / output is not described in the main body modules M12 and M22. Therefore, the processing unit 13d calls the input / output module M11 when the main module M12 is to be executed and calls the input / output module M21 when the main module M22 is to be executed. And execute.

  FIG. 6 is a flowchart showing a processing procedure of the processing unit 13d based on the input / output module M11.

  In step Sc1, the processing unit 13d acquires parameters necessary for the main body processing from the data stacker 13a. In step Sc2, the processing unit 13d reads data to be subjected to main body processing from the data stacker 13a. In step Sc3, the processing unit 13d passes the above parameters and data to the processing based on the main body module M12. The processing unit 13d performs main body processing (arithmetic processing) by separate processing based on the main body module M12.

  In step Sc4, the processing unit 13d waits for the main body processing to be completed. When the main body process is completed, the processing unit 13d proceeds from step Sc4 to step Sc5. In step Sc5, the processing unit 13d stores the parameter acquired in step Sc1 and the data after the main body processing is performed in the data stacker 13a. Here, the reason why the parameters acquired in step Sc1 are stored as they are is that the shape and size of the data do not change in the arithmetic processing.

  FIG. 7 is a flowchart showing a processing procedure of the processing unit 13d based on the input / output module M21.

  In step Sd1, the processing unit 13d acquires parameters necessary for the main body processing from the data stacker 13a. In step Sd2, the processing unit 13d reads data to be subjected to main body processing from the data stacker 13a. In step Sd3, the processing unit 13d passes the above parameters and data to the processing based on the main body module M22. The processing unit 13d performs main body processing (conversion processing) by separate processing based on the main body module M22.

  In step Sd4, the processing unit 13d waits for the main body processing to be completed. When the main body process is completed, the processing unit 13d proceeds from step Sd4 to step Sd5. In step Sd5, the processing unit 13d determines data properties (such as data shape and size) after the above-described main body processing is performed. In step Sd6, the processing unit 13d determines the storage destination of the data after the main body processing in consideration of the determined property.

  In step Sd7, the processing unit 13d creates a parameter reflecting the property determined in step Sd5. In step Sd8, the processing unit 13d stores the created parameter and the data after the main body processing in the data stacker 13a.

  As described above, according to the present embodiment, loops in which the repeated processes are unaffected are managed by the scheduler 13c, and the processing unit 13d manages only loops in which the repeated processes are affected. For this reason, when changing the main body process, it is only necessary to change the processing contents in the processing unit 13d in many cases, and it is possible to easily cope with a new MRI technique.

  Further, according to the present embodiment, after classifying the main body processing into the arithmetic processing and the conversion processing, the input / output processing related to the main body processing belonging to the arithmetic processing is executed by the common input / output module M11, and the conversion processing is performed. Since the input / output processing relating to the main body processing to be performed is executed by the common input / output module M21, each of the main body modules M12 and M22 does not need to include the input / output processing, and the data in the entire program module relating to the main body processing The amount can be reduced.

  FIG. 8 is a block diagram showing a modified configuration example of the reconstruction unit 13. In the reconstruction unit 13 shown in FIG. 8, the processing unit 13d in the embodiment is divided into a processing unit 13e and a post-processing unit 13f. The processing unit 13e does not perform post-processing, and the post-processing unit 13f performs this post-processing.

  FIG. 9 is a block diagram showing another modified configuration example of the reconstruction unit 13. In the reconstruction unit 13 shown in FIG. 9, the data stacker 13a in the embodiment is further divided into a data stacker 13g and an image stacker 13h. The data stacker 13g stacks raw data and rearranges the stacked raw data. The image stacker 13h stacks image groups and rearranges the stacked image groups.

This embodiment can be variously modified as follows.
Therefore, the processing unit 13d may manage a part of a loop in which repeated processing has no influence on each other.

  Note that the present invention is not limited to the above-described embodiment as it is, and can be embodied by modifying the constituent elements without departing from the scope of the invention in the implementation stage. In addition, various inventions can be formed by appropriately combining a plurality of components disclosed in the embodiment. For example, some components may be deleted from all the components shown in the embodiment.

  DESCRIPTION OF SYMBOLS 1 ... Static magnetic field magnet, 2 ... Gradient magnetic field coil, 3 ... Gradient magnetic field power supply, 4 ... Bed, 5 ... Bed control part, 6 ... Transmission RF coil, 7 ... Transmission part, 8 ... Reception RF coil, 9 ... Reception part, DESCRIPTION OF SYMBOLS 10 ... Gantry control part, 11 ... Collection control part, 12 ... Data input part, 13 ... Reconstruction part, 13b ... Event generator, 13h ... Image stacker, 13c ... Scheduler, 13a, 13g ... Data stacker, 13f ... Post-processing part , 13d, 13e ... processing unit, 14 ... database.

Claims (1)

Means for receiving a magnetic resonance signal radiated from the subject;
Reconstructing means for reconstructing the image of the subject by reconstructing processing by combining a plurality of program modules that execute different processes based on the received magnetic resonance signals while applying a plurality of loops. A magnetic resonance imaging apparatus,
The reconstruction means includes
A plurality of main body procedures for executing each of the plurality of program modules, a first input / output procedure for data input / output related to conversion accompanied by a change in data type and size, and a change in data type and size A second input / output procedure for data input / output relating to the arithmetic processing not involving the above, and in executing one of the main body procedures, if the main body procedure to be executed relates to the conversion processing, the first input / output procedure is executed. An execution means for performing data input / output by the second input / output procedure if the main body procedure to be executed is related to the arithmetic processing by the input / output procedure ;
Each of a first program module of the plurality of program modules in the plurality of loops and a second program module using data obtained by the first program module among the plurality of program modules. A management means for managing a loop applied for the repetition of
A magnetic resonance imaging apparatus comprising: means for managing a loop different from that managed by the management means among the plurality of loops.

Images