WO2021012972A1 - 核磁共振成像控制方法、装置及计算机可读存储介质 - Google Patents
核磁共振成像控制方法、装置及计算机可读存储介质 Download PDFInfo
- Publication number
- WO2021012972A1 WO2021012972A1 PCT/CN2020/101520 CN2020101520W WO2021012972A1 WO 2021012972 A1 WO2021012972 A1 WO 2021012972A1 CN 2020101520 W CN2020101520 W CN 2020101520W WO 2021012972 A1 WO2021012972 A1 WO 2021012972A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- magnetic resonance
- nuclear magnetic
- scan
- scanning
- resonance imaging
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/4802—Travelling-wave MR
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/54—Signal processing systems, e.g. using pulse sequences ; Generation or control of pulse sequences; Operator console
- G01R33/56—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution
- G01R33/561—Image enhancement or correction, e.g. subtraction or averaging techniques, e.g. improvement of signal-to-noise ratio and resolution by reduction of the scanning time, i.e. fast acquiring systems, e.g. using echo-planar pulse sequences
- G01R33/5615—Echo train techniques involving acquiring plural, differently encoded, echo signals after one RF excitation, e.g. using gradient refocusing in echo planar imaging [EPI], RF refocusing in rapid acquisition with relaxation enhancement [RARE] or using both RF and gradient refocusing in gradient and spin echo imaging [GRASE]
- G01R33/5616—Echo train techniques involving acquiring plural, differently encoded, echo signals after one RF excitation, e.g. using gradient refocusing in echo planar imaging [EPI], RF refocusing in rapid acquisition with relaxation enhancement [RARE] or using both RF and gradient refocusing in gradient and spin echo imaging [GRASE] using gradient refocusing, e.g. EPI
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R33/00—Arrangements or instruments for measuring magnetic variables
- G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
- G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
- G01R33/48—NMR imaging systems
- G01R33/58—Calibration of imaging systems, e.g. using test probes, Phantoms; Calibration objects or fiducial markers such as active or passive RF coils surrounding an MR active material
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Definitions
- the present invention relates to the technical field of nuclear magnetic resonance imaging, in particular to a nuclear magnetic resonance imaging control method, device and computer readable storage medium.
- Magnetic Resonance Imaging is a new type of medical imaging technology that uses the nuclear magnetic resonance phenomenon of a certain atomic nucleus in human tissues to process the resulting radio frequency signals through an electronic computer to reconstruct a certain scan level image of the human body Because of its non-ionizing radiation, multi-sequence, multi-parameter, multi-plane imaging, and high soft tissue resolution, it is widely used in disease diagnosis.
- a typical magnetic resonance imaging method is: the current signal induced by the receiving coil module is converted from analog to digital to obtain a digital signal, and the digital signal is filled into K space according to a certain encoding direction, where the K space is a kind of The data of the original magnetic resonance signal fills the space.
- the user can set the desired orientation of the reconstructed image, such as cross section or coronal plane or other angles.
- the desired orientation of the reconstructed image fill the collected data into K-space according to the encoding direction, and then After processing the data in the K-space with an image reconstruction algorithm, an MRI image of a certain scanning plane (section) of the scanning part is obtained.
- Artifacts refer to various forms of images that do not exist in the scanned object but appear on the image. Take the upper abdomen and lower abdomen as an example. Due to the breathing movement and gastrointestinal peristalsis, the MRI images usually have artifacts. In order to suppress motion artifacts, the scanning methods usually used are: (1) Scanning during the patient's breath-holding; this method has limitations, often requiring the subject to hold the breath several times, and some patients cannot fully cooperate or hold the breath for a long time. The patient brings discomfort. (2) Use a breathing monitoring device or collect navigation signals to trigger the collection of echo signals, so that the collected K-space data always corresponds to the same or similar motion state.
- the main purpose of the present invention is to provide a MRI control method, device, and computer readable storage medium, aiming to solve the problem of low calibration efficiency and low accuracy in the MRI clinical scanning process, which affects the quality of MRI images .
- the present invention provides a nuclear magnetic resonance imaging control device, which includes a processor suitable for realizing various computer program instructions and a memory suitable for storing multiple computer program instructions.
- the device is connected with magnetic resonance imaging equipment, so
- the computer program instructions are loaded by the processor and execute the following steps: set the pre-scanning field of view FOV, and control the MRI equipment under the preset FOV to activate the scan sequence to pre-scan the target scanning part to obtain the scan data; input the scan data into a user To identify the neural network model of the scan location information, and output the scan location information corresponding to the scan data from the neural network model; determine the gradient encoding direction of MRI according to the scan location information and the phase encoding direction in the gradient encoding direction relative to the preset reference Rotate the direction to a preset deflection angle; control the MRI equipment to excite the scanning sequence to perform MRI scans on the target scanning part to obtain multiple echo signals, where a gradient along the gradient encoding direction is applied during the acquisition of multiple echo signals Field
- the step of controlling the nuclear magnetic resonance imaging device to excite a scanning sequence to perform nuclear magnetic resonance scanning on the target scanning site to obtain multiple echo signals includes: controlling the radio frequency transmitting coil of the nuclear magnetic resonance imaging device to transmit radio frequency pulses to the scanning site to stimulate scanning Position of nuclear spin; control the gradient coil of the MRI equipment to generate a gradient field, the gradient field encodes the nuclear spin excited by the scanning part to generate an echo signal, the phase encoding direction of the gradient field is rotated relative to the preset reference direction Set the deflection angle; receive the echo signal through the radio frequency receiving coil of the MRI equipment.
- the preset deflection angle is determined by the included angle on the scan level between the preset reference direction of the sporty reference target object and the preset target direction of the region of interest of the scanning part.
- the computer program instructions are loaded by the processor and perform the following steps: training the neural network model in advance, and the specific steps of training the neural network model include: inputting the sample image into the information recognition model to be trained to obtain and The current output recognition result corresponding to the sample image; according to the error between the current output recognition result and the recognition result of the scanned part and tissue of the sample image, determine whether the loss function of the neural network model converges; when the loss function converges, the The neural network model training is over.
- the computer program instructions are loaded by the processor to perform the following steps: acquiring the generated nuclear magnetic resonance image from the nuclear magnetic resonance imaging device, and displaying it on the display of the device or storing it in the memory.
- the present invention also provides a nuclear magnetic resonance imaging control method, which is applied to a computer device, the computer device is connected to the nuclear magnetic resonance imaging equipment, the method includes the following steps: setting the pre-scanning field of view FOV, under the preset FOV Control the MRI equipment to activate the scan sequence to pre-scan the target scan location to obtain scan data; input the scan data into a neural network model for identifying the scan location information, and output the scan location information corresponding to the scan data from the neural network model;
- the scanning position information determines the gradient encoding direction of the MRI and the phase encoding direction in the gradient encoding direction rotates the preset deflection angle relative to the preset reference direction; controls the MRI equipment to activate the scan sequence to perform the MRI scan of the target scan position to obtain Multiple echo signals, where a gradient field along the gradient encoding direction is applied during the acquisition process of multiple echo signals; multiple echo signals are filled into K-space to obtain K-space data of the target scanning part; according to K-space The data determines the following steps
- the step of controlling the nuclear magnetic resonance imaging device to excite a scanning sequence to perform nuclear magnetic resonance scanning on the target scanning site to obtain multiple echo signals includes: controlling the radio frequency transmitting coil of the nuclear magnetic resonance imaging device to transmit radio frequency pulses to the scanning site to stimulate scanning Position of nuclear spin; control the gradient coil of the MRI equipment to generate a gradient field, the gradient field encodes the nuclear spin excited by the scanning part to generate an echo signal, the phase encoding direction of the gradient field is rotated relative to the preset reference direction Set the deflection angle; receive the echo signal through the radio frequency receiving coil of the MRI equipment.
- the preset deflection angle is determined by the included angle on the scan level between the preset reference direction of the sporty reference target object and the preset target direction of the region of interest of the scanning part.
- the MRI control method further includes the step of pre-training the neural network model, and the specific steps of training the neural network model include: inputting the sample image into the information recognition model to be trained to obtain a corresponding sample image The current output recognition result; according to the current output recognition result and the error of the recognition result of the scanning part and tissue of the sample image, determine whether the loss function of the neural network model converges; when the loss function converges, the neural network model The training is over.
- the present invention also provides a computer-readable storage medium that stores a plurality of computer program instructions and is applied to a computer device that is connected to a nuclear magnetic resonance imaging device.
- the computer program instructions The steps of the control method based on nuclear magnetic resonance imaging are executed and realized by the processor of the computer device.
- the MRI control method, device and computer readable medium of the present invention perform pre-scanning of the object to be scanned, and input the obtained scan data into the neural network model for information identification, and then according to the neural network
- the output result of the model determines the calibration parameters to complete the calibration of the MRI equipment, which realizes the improvement of the calibration efficiency and accuracy during the patient calibration process, and at the same time, the phase encoding direction and the region of interest are preset at the scan level.
- Directional deflection presets the deflection angle to reduce the influence of artifacts in the phase-encoding direction from the movement of the moving reference organ on the region of interest, thereby improving the image quality of the MRI image, which is conducive to improving the accuracy of clinical diagnosis.
- FIG. 1 is a schematic block diagram of the structure of a preferred embodiment of the nuclear magnetic resonance imaging control device of the present invention.
- Fig. 2 is a method flowchart of a preferred embodiment of the MRI control method of the present invention.
- FIG. 1 is a schematic structural diagram of a preferred embodiment of the nuclear magnetic resonance imaging control device of the present invention.
- the nuclear magnetic resonance imaging control device 1 includes, but is not limited to, a memory 11 suitable for storing various computer program instructions, a processor 12 that executes various computer program instructions, and a display 13. Both the memory 11 and the display 13 are electrically connected to the processor 12 through an electrical connection line, and are connected to the processor 12 through a data bus for data transmission.
- the processor 12 can call the nuclear magnetic resonance imaging control program 10 stored in the memory 11, and execute the nuclear magnetic resonance imaging control program 10 to control the nuclear magnetic resonance imaging device 2 to scan the target scanning part of the object to be scanned to obtain scan data, And generate nuclear magnetic resonance images based on the scan data.
- the nuclear magnetic resonance imaging control device 1 may be a personal computer, a notebook computer, a server, and other computer devices installed with the nuclear magnetic resonance imaging control program 10 of the present invention.
- the nuclear magnetic resonance imaging control device 1 is connected with a nuclear magnetic resonance imaging device 2, which can scan different parts of the human body of a target object for nuclear magnetic resonance scanning to obtain multiple different echo signals
- the processor 12 executes the nuclear magnetic resonance imaging control program 10 to process the echo signal.
- the nuclear magnetic resonance image can be generated. Reduce the influence of the artifacts in the phase encoding direction from the movement of the moving reference organ on the region of interest, thereby improving the image quality of the MRI image, which is beneficial to improving the accuracy of clinical diagnosis.
- the memory 11 includes at least one type of readable storage medium.
- the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), and magnetic memory. , Disks, CDs, etc.
- the memory 11 may be an internal storage unit of the nuclear magnetic resonance imaging control device 1, such as the hard disk, read-only memory ROM, random access memory RAM, and electrically erasable memory EEPROM of the nuclear magnetic resonance imaging control device 1. , Flash memory FLASH or CD, etc.
- the memory 11 may also be an external storage device of the nuclear magnetic resonance imaging control device 1, for example, a plug-in hard disk equipped on the nuclear magnetic resonance imaging control device 1, a smart memory card (Smart Media Card, SMC), Secure Digital (SD) card, Flash Card, etc. Further, the memory 11 may also include both an internal storage unit of the nuclear magnetic resonance imaging control device 1 and an external storage device.
- the memory 11 can be used not only to store application software and various data installed in the nuclear magnetic resonance imaging control device 1, for example, to store the program code of the nuclear magnetic resonance imaging control program 10, etc., but also to temporarily store what has been output or will be output. Data, such as a nuclear magnetic resonance image generated by the nuclear magnetic resonance imaging device 2.
- the processor 12 may be a central processing unit (Central Processing Unit) in some embodiments.
- Central Processing Unit CPU
- controller a controller
- microcontroller a microprocessor or other data processing chips are used to call and run the program code or processing data stored in the memory 11, for example to execute the nuclear magnetic resonance imaging control program 10, etc.
- the display 13 may be a touch display screen or a general LED display screen, which can display the nuclear magnetic resonance image generated by the nuclear magnetic resonance imaging device 2.
- the nuclear magnetic resonance imaging control program 10 may also be divided into one or more modules, and the one or more modules are stored in the memory 11 and run by one or more processors ( This embodiment is executed by the processor 12) to complete the present invention.
- the module referred to in the present invention refers to a series of computer program instruction segments that can complete specific functions, and is used to describe the nuclear magnetic resonance imaging control program 10 in the nuclear magnetic resonance imaging The execution process in the control device 1.
- the MRI control program 10 is composed of program modules composed of multiple computer program instructions, including, but not limited to, a neural network creation module 101, a scanning position determination module 102, a scanning gradient determination module 103, The scan data acquisition module 104 and the nuclear magnetic resonance image generation module 105.
- the module referred to in the present invention refers to a series of computer program instruction segments that can be executed by the processor 12 of the nuclear magnetic resonance imaging control device 1 and can complete fixed functions, which are stored in the nuclear magnetic resonance imaging control device 1 In the memory 11.
- the neural network creation module 101 is used for pre-training a neural network model for identifying the scanning part information of the object to be scanned.
- the neural network model is a model used for identifying scanning part information obtained by training a large number of image samples in advance.
- the image sample is the detection data of various parts of the human body or the detection data after Fourier transform, and the recognition result of the tissue information corresponding to each detection data.
- the neural network creation module 101 inputs the sample data recognized by the scanned parts and tissues into the neural network model for training. Specifically, the neural network creation module 101 inputs the sample image into the information recognition model to be trained to obtain the current image corresponding to the sample image.
- the neural network creation module 101 determines whether the loss function of the neural network model converges according to the current output recognition result and the error of the recognition result of the scanned part of the sample image and the tissue; when the loss function converges, the The neural network model training is over.
- the scanning position determining module 102 is used to set a pre-scanning field of view (FOV), and controlling the nuclear magnetic resonance imaging device 2 to excite a scanning sequence under a preset FOV to perform pre-scanning of the target scanning position of the object to be scanned to obtain scan data.
- the object to be scanned is a target scan location of a patient undergoing an MRI scan. According to medical clinical requirements, the target scan location may be one or more of the limbs, abdomen, pelvis, chest, or head.
- the scan data may be pre-scan data collected by using a fast pre-scan sequence to excite the target scanning part of the object to be scanned, and the scan data may be a positioning image obtained by collecting the scanned part of the object to be scanned by a positioning scanning method.
- the scan data may also be a diagnostic image obtained by collecting an object to be scanned using an imaging sequence.
- the type of the scan data may be one or more of amplitude image, phase image or K-space data.
- the scanning sequence may be a low-resolution nuclear magnetic resonance sequence, such as a 3D gradient echo sequence (GRE) or a single shot fast spin echo sequence (Single Shot Fast Spin Echo, SSFSE).
- GRE 3D gradient echo sequence
- SSFSE single shot fast spin Echo
- the three-dimensional amplitude image, phase image or K-space data corresponding to the target scanning part can be obtained through pre-scanning.
- the positioning scanning method may adopt camera shooting, infrared imaging, or positioning scanning sequence scanning.
- the scanning position determination module 102 is also used for inputting scanning data into the trained neural network model, and outputting scanning position information corresponding to the scanning data from the neural network model.
- the scan data of the target scan part is obtained, the scan data can be directly input to the trained neural network model, and the neural network model recognizes and outputs scan part information corresponding to the scan data, wherein the scanned part information includes the The organization to which the scanned part belongs and the geometric information of the scanned part.
- the neural network model can select an image recognition model, and according to the input scanned image, the model will correspondingly output a recognition result, that is, the information of the target scanning part.
- the scan gradient determination module 103 is used to determine the gradient encoding direction of the nuclear magnetic resonance imaging and the phase encoding direction in the gradient encoding direction according to the scan position information; the phase encoding direction in the gradient encoding direction rotates preset relative to the preset reference direction The deflection angle, wherein the preset deflection angle is determined by the angle between the preset reference direction of the sporty reference target and the preset target direction of the region of interest of the scanning part on the scan level. Since the time to complete data acquisition in the phase encoding direction is much longer than the time to complete data acquisition in the frequency encoding direction, data differences caused by organ movement are likely to appear in the phase encoding direction, which makes the reconstructed image blurry or overlapping artifacts In the area of interest.
- the phase encoding direction in the encoding direction of this embodiment is deflected by a preset deflection angle relative to the preset reference direction, so as to reduce the effect of artifacts in the phase encoding direction on the region of interest. influences.
- the preset reference direction can be manually set, such as the left and right direction of the object to be scanned.
- the preset deflection angle is determined by the angle between the preset reference direction of the sporty reference target and the preset target direction of the region of interest of the scanning part on the scan level.
- the scan data acquisition module 104 is used to control the nuclear magnetic resonance imaging device 2 to excite a scan sequence to perform a nuclear magnetic resonance scan on the target scanning part to obtain multiple echo signals, wherein the along-gradient coding is applied during the acquisition of the multiple echo signals Directional gradient field.
- the scan data acquisition module 104 controls the radio frequency transmitting coil of the nuclear magnetic resonance imaging device 2 to emit radio frequency pulses to the scanning part to excite nuclear spins at the scanning part; controls the gradient coil of the nuclear magnetic resonance imaging device 2 to generate a gradient field ,
- the gradient field encodes the nuclear spins excited by the scanning part to generate an echo signal, the phase encoding direction of the gradient field rotates a preset deflection angle relative to the preset reference direction; the echo is received by the radio frequency receiving coil of the nuclear magnetic resonance imaging device 2 signal.
- the scanning sequence includes a radio frequency pulse sequence and a gradient pulse sequence, etc.
- the scanning parameters corresponding to the two sequences include echo time (echo time, TE), inversion time (inversion time, TI), the size of the flip angle of the RF pulse, and the measurement time (acquisition time, TA), delay (time delay, TD) one or a combination of more.
- the nuclear magnetic resonance image generating module 105 is used for filling multiple echo signals into K-space to obtain K-space data of the target scanning part. Obtain the echo signal received by the radio frequency receiving coil, and fill multiple echo signals into the K-space to obtain the K-space data of the scanned part. It can be understood that after the gradient encoding direction is determined, the filling trajectory of the K space is also determined, and the K space data of the scanning part can be obtained by filling the K space data according to the predetermined filling trajectory. Optionally, the K-space data of the scanned part may be fully collected or under-collected.
- the filling trajectory of the K-space data may be one or a combination of one or more of sequential symmetric filling, center priority acquisition filling, circuitous filling, spiral filling, and radial filling.
- the nuclear magnetic resonance image generating module 105 is also used to determine the initial nuclear magnetic resonance image according to the K-space data, and rotate the initial nuclear magnetic resonance image by a preset deflection angle to generate the nuclear magnetic resonance image; since the phase encoding direction is of interest relative to the scanning part
- the preset target direction of the area on the scan level is deflected by the preset deflection angle, then the K-space data is deflected by the preset deflection angle relative to the left and right or front and back directions of the human body, so the reconstructed initial MRI image is relative to the left and right sides of the human body.
- the direction or front-to-back direction is deflected by the preset deflection angle.
- the initial MRI image needs to be rotated reversely to the preset deflection angle to obtain clinically commonly used MRI images.
- the left and right directions of the MRI image are the left and right directions of the human body.
- the direction is the front and back direction of the human body.
- this embodiment does not limit the image reconstruction method, and it is sufficient to perform image reconstruction on the collected K-space data by using an existing image reconstruction method.
- the nuclear magnetic resonance image generating module 105 is also used to obtain the generated nuclear magnetic resonance image from the nuclear magnetic resonance imaging device, and display it on the display 13 or store it in the memory 11 for the doctor to provide reference for diagnosis and treatment.
- FIG. 2 it is a flowchart of a preferred embodiment of the MRI control method of the present invention.
- the various method steps of the nuclear magnetic resonance imaging control method are implemented by a computer software program, which is stored in a computer-readable storage medium (such as the memory 11 of this embodiment) in the form of computer program instructions.
- the computer-readable storage medium may include: read-only memory, random access memory, magnetic disk or optical disk, etc.
- the computer program instructions can be loaded by a processor (for example, the processor 12 in this embodiment) and execute the following steps.
- Step S21 pre-training a neural network model for identifying the scanning part information of the object to be scanned;
- the neural network model is a pre-trained one for identifying the scanning part information through a large number of image samples. model.
- the image sample is the detection data of various parts of the human body or the detection data after Fourier transform, and the recognition result of the tissue information corresponding to each detection data.
- the sample data recognized by the scanned parts and tissues are input into the neural network model for training.
- the input data of the neural network model includes each sample image obtained by pre-scanning and each The scan location and tissue recognition result of a pre-scan sample image.
- a large number of samples must be collected first, which can be a pre-scan sequence for a certain number of people of various types (such as the elderly, adults, children, men, women) to perform rapid MRI scans of various parts of the body.
- the corresponding scanned image database is collected, and the scanned position of each scanned image in the scanned image database and the tissue shape information of the scanned position are marked.
- the Fourier transform of each scanned image can also be performed to obtain a corresponding amplitude image, and the amplitude image is used as a training sample of the model.
- the specific process of the neural network model training includes: inputting the sample image into the information recognition model to be trained to obtain the current output recognition result corresponding to the sample image; according to the current output recognition result and the sample image The error of the recognition result of the scanning part and the tissue determines whether the loss function of the neural network model converges; when the loss function converges, the training of the neural network model ends.
- Step S22 Set a pre-scan FOV, and control the nuclear magnetic resonance imaging device 2 to excite the scan sequence under the preset FOV to perform pre-scan on the target scanning part of the object to be scanned to obtain scan data.
- first set the pre-scan field of view Field of View, FOV
- control the scanning sequence of the nuclear magnetic resonance imaging device 2 under the preset FOV to excite the target scanning part of the object to be scanned for pre-scanning to obtain scan data.
- the object to be scanned is a target scan location of a patient undergoing an MRI scan. According to medical clinical requirements, the target scan location may be one or more of the limbs, abdomen, pelvis, chest, or head.
- the scan data may be pre-scan data collected by using a fast pre-scan sequence to excite the target scanning part of the object to be scanned, and the scan data may be a positioning image obtained by collecting the scanned part of the object to be scanned by a positioning scanning method.
- the scan data may also be a diagnostic image obtained by collecting an object to be scanned using an imaging sequence.
- the type of the scan data may be one or more of amplitude image, phase image or K-space data.
- the scanning sequence may be a low-resolution nuclear magnetic resonance sequence, such as a 3D gradient echo sequence (GRE) or a single shot fast spin echo sequence (Single Shot Fast Spin Echo, SSFSE).
- the three-dimensional amplitude image, phase image or K-space data corresponding to the target scanning part can be obtained through pre-scanning.
- the positioning scanning method may adopt camera shooting, infrared imaging, or positioning scanning sequence scanning. By using the pre-scan sequence to excite the object to be scanned to obtain scan data, the entire process takes a short time and can be completed in about 3 seconds.
- Step S23 Input the scan data into the trained neural network model, and output the scan location information corresponding to the scan data from the neural network model.
- the scan data can be directly input to the trained neural network model, and the neural network model recognizes and outputs scan part information corresponding to the scan data, wherein the scanned part information includes the The organization to which the scanned part belongs and the geometric information of the scanned part.
- the neural network model can select an image recognition model, and according to the input scanned image, the model will correspondingly output a recognition result, that is, the information of the target scanning part.
- Step S24 Determine the gradient encoding direction of the MRI and the phase encoding direction in the gradient encoding direction according to the scan position information; the phase encoding direction in the gradient encoding direction is rotated by a preset deflection angle relative to a preset reference direction, wherein It is assumed that the deflection angle is determined by the angle between the preset reference direction of the sporty reference target object and the preset target direction of the region of interest of the scanning part on the scan level.
- three orthogonal gradient magnetic fields are usually used for spatial positioning.
- the gradient magnetic field in one direction is used for radio frequency pulses to selectively excite the nuclear spins of protons in a scanning plane, and the gradient magnetic field in one direction is used for To phase-encode the echo signal in one direction in the scanning plane, the gradient magnetic field in the other direction is used to read out and encode the echo signal in the other direction in the scanning plane.
- the direction of the gradient magnetic field used for phase encoding is called the phase encoding direction
- the direction of the gradient magnetic field used for frequency encoding is called the frequency encoding direction.
- the phase encoding direction in the encoding direction of this embodiment is deflected by a preset deflection angle relative to the preset reference direction, so as to reduce the effect of artifacts in the phase encoding direction on the region of interest. influences.
- the preset reference direction can be manually set, such as the left and right direction of the object to be scanned.
- the preset deflection angle is determined by the angle between the preset reference direction of the sporty reference target and the preset target direction of the region of interest of the scanning part on the scan level.
- Step S25 controlling the nuclear magnetic resonance imaging device 2 to excite the scanning sequence to perform nuclear magnetic resonance scanning on the target scanning part to obtain multiple echo signals, wherein a gradient field along the gradient encoding direction is applied during the acquisition of the multiple echo signals.
- the step of controlling the scanning sequence of the nuclear magnetic resonance imaging device 2 to excite the target scanning part to obtain multiple echo signals includes: controlling the radio frequency transmitting coil of the nuclear magnetic resonance imaging device 2 to emit radio frequency pulses to the scanning part to excite The nuclear spin of the scanning part; the gradient coil of the MRI equipment 2 is controlled to generate a gradient field, the gradient field encodes the nuclear spin excited by the scanning part to generate an echo signal, and the phase encoding direction of the gradient field is relative to the preset reference direction Rotate the preset deflection angle; receive the echo signal through the radio frequency receiving coil of the nuclear magnetic resonance imaging device 2.
- the scanning sequence includes a radio frequency pulse sequence and a gradient pulse sequence, etc.
- the scanning parameters corresponding to the two sequences include echo time (echo time, TE), inversion time (inversion time, TI), the size of the flip angle of the RF pulse, and the measurement time (acquisition time, TA), delay (time delay, TD) one or a combination of more.
- Step S26 Fill the K-space with multiple echo signals to obtain K-space data of the target scanning part. Obtain the echo signal received by the radio frequency receiving coil, and fill multiple echo signals into the K-space to obtain the K-space data of the scanned part. It can be understood that after the gradient encoding direction is determined, the filling trajectory of the K space is also determined, and the K space data of the scanning part can be obtained by filling the K space data according to the predetermined filling trajectory. Optionally, the K-space data of the scanned part may be fully collected or under-collected.
- the filling trajectory of the K-space data may be one or a combination of one or more of sequential symmetric filling, center priority acquisition filling, circuitous filling, spiral filling, and radial filling.
- Step S27 Determine the initial nuclear magnetic resonance image according to the K-space data, and rotate the initial nuclear magnetic resonance image by a preset deflection angle to generate the nuclear magnetic resonance image; since the phase encoding direction is preset relative to the region of interest of the scanning part on the scanning level The target direction is deflected by the preset deflection angle, then the K-space data is deflected by the preset deflection angle relative to the left-right or front-rear direction of the human body, so the reconstructed initial MRI image is deflected by the preset relative to the left-right or front-rear direction of the human body Deflection angle, therefore, it is necessary to reversely rotate the initial MRI image to a preset deflection angle to obtain commonly used clinical MRI images.
- the left and right directions of the MRI image are the left and right directions of the human body
- the up and down directions are the front and back directions of the human body.
- this embodiment does not limit the image reconstruction method, and it is sufficient to perform image reconstruction on the collected K-space data by using an existing image reconstruction method.
- step S28 the generated nuclear magnetic resonance image is obtained from the nuclear magnetic resonance imaging device, and displayed on the display 13 or stored in the memory 11.
- the nuclear magnetic resonance imaging control method further includes the following steps: acquiring the generated nuclear magnetic resonance image from the nuclear magnetic resonance imaging device 2 and displaying it on the display 13, or the output is stored in the memory 11 for the doctor to diagnose and treat Provide reference.
- the present invention also provides a computer-readable storage medium that stores a plurality of computer program instructions, and the computer program instructions are loaded by a processor of a computer device and execute each of the nuclear magnetic resonance imaging control methods of the present invention. step.
- the program can be stored in a computer-readable storage medium.
- the storage medium may include: read-only memory, random access memory, Disk or CD, etc.
- the nuclear magnetic resonance imaging control method, device and computer readable medium of the present invention pre-scan the object to be scanned, input the obtained scan data into the neural network model for information identification, and then determine the calibration parameters according to the output result of the neural network model , In order to complete the calibration of the MRI equipment, solve the problem of low calibration efficiency and need to be improved in the patient calibration process of MRI clinical scanning, and realize the improvement of the calibration efficiency and calibration accuracy during the patient calibration process.
- the effect of artifacts in the phase encoding direction from the movement of the moving reference organ on the region of interest is reduced, thereby improving the NMR
- the image quality of the resonance image is conducive to improving the accuracy of clinical diagnosis.
- the MRI control method, device and computer readable medium of the present invention perform pre-scanning of the object to be scanned, and input the obtained scan data into the neural network model for information identification, and then according to the neural network
- the output result of the model determines the calibration parameters to complete the calibration of the MRI equipment, which realizes the improvement of the calibration efficiency and accuracy during the patient calibration process, and at the same time, the phase encoding direction and the region of interest are preset at the scan level.
- Directional deflection presets the deflection angle to reduce the influence of artifacts in the phase-encoding direction from the movement of the moving reference organ on the region of interest, thereby improving the image quality of the MRI image, which is conducive to improving the accuracy of clinical diagnosis.
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Condensed Matter Physics & Semiconductors (AREA)
- High Energy & Nuclear Physics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Radiology & Medical Imaging (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
Description
Claims (10)
- 一种核磁共振成像控制装置,包括适于实现各种计算机程序指令的处理器以及适于存储多条计算机程序指令的存储器,其特征在于,该装置连接有磁共振成像设备,所述计算机程序指令由处理器加载并执行如下步骤:设置预扫描视野FOV,在预设的FOV下控制核磁共振成像设备激发扫描序列对目标扫描部位进行预扫描得到扫描数据;将扫描数据输入一个用于识别扫描部位信息的神经网络模型,并从神经网络模型输出扫描数据对应的扫描部位信息;根据扫描部位信息确定核磁共振成像的梯度编码方向以及梯度编码方向中的相位编码方向相对于预设基准方向旋转预设偏转角度;控制核磁共振成像设备激发扫描序列对目标扫描部位进行核磁共振扫描以获取多个回波信号,其中,在多个回波信号的采集过程中施加沿梯度编码方向的梯度场;将多个回波信号填充入K空间以获取目标扫描部位的K空间数据;根据K空间数据确定初始核磁共振图像,将初始核磁共振图像反向旋转预设偏转角度以生成目标扫描部位的核磁共振图像。
- 如权利要求1所述的核磁共振成像控制装置,其特征在于,所述控制核磁共振成像设备激发扫描序列对目标扫描部位进行核磁共振扫描以获取多个回波信号的步骤包括:控制核磁共振成像设备的射频发射线圈向扫描部位发射射频脉冲以激发扫描部位的核自旋;控制核磁共振成像设备的梯度线圈产生梯度场,梯度场对扫描部位激发的核自旋进行编码以产生回波信号,梯度场的相位编码方向相对于预设基准方向旋转预设偏转角度;通过核磁共振成像设备的射频接收线圈接收回波信号。
- 如权利要求2所述的核磁共振成像控制装置,其特征在于,所述预设偏转角度由运动型参考目标对象的预设参考方向与扫描部位的感兴趣区的预设目标方向在扫描层面上的夹角所确定。
- 如权利要求1所述的核磁共振成像控制装置,其特征在于,所述计算机程序指令由处理器加载还执行如下步骤:预先训练所述神经网络模型,该神经网络模型训练的具体步骤包括:将样本图像输入至待训练的信息识别模型中,得到与样本图像对应的当前输出识别结果;根据当前输出识别结果和样本图像的扫描部位和组织的识别结果的误差,确定所述神经网络模型的损失函数是否收敛;当所述损失函数收敛时,所述神经网络模型训练结束。
- 如权利要求1至4任一项所述的核磁共振成像控制装置,其特征在于,所述计算机程序指令由处理器加载还执行如下步骤:从核磁共振成像设备获取生成的核磁共振图像,并显示在该装置的显示器上或者存储在存储器中。
- 一种核磁共振成像控制方法,应用于计算机装置中,该计算机装置连接有核磁共振成像设备,其特征在于,该方法包括如下步骤:设置预扫描视野FOV,在预设的FOV下控制核磁共振成像设备激发扫描序列对目标扫描部位进行预扫描得到扫描数据;将扫描数据输入一个用于识别扫描部位信息的神经网络模型,并从神经网络模型输出扫描数据对应的扫描部位信息;根据扫描部位信息确定核磁共振成像的梯度编码方向以及梯度编码方向中的相位编码方向相对于预设基准方向旋转预设偏转角度;控制核磁共振成像设备激发扫描序列对目标扫描部位进行核磁共振扫描以获取多个回波信号,其中,在多个回波信号的采集过程中施加沿梯度编码方向的梯度场;将多个回波信号填充入K空间以获取目标扫描部位的K空间数据;根据K空间数据确定初始核磁共振图像,将初始核磁共振图像反向旋转预设偏转角度以生成目标扫描部位的核磁共振图像。
- 如权利要求6所述的核磁共振成像控制方法,其特征在于,所述控制核磁共振成像设备激发扫描序列对目标扫描部位进行核磁共振扫描以获取多个回波信号的步骤包括:控制核磁共振成像设备的射频发射线圈向扫描部位发射射频脉冲以激发扫描部位的核自旋;控制核磁共振成像设备的梯度线圈产生梯度场,梯度场对扫描部位激发的核自旋进行编码以产生回波信号,梯度场的相位编码方向相对于预设基准方向旋转预设偏转角度;通过核磁共振成像设备的射频接收线圈接收回波信号。
- 如权利要求7所述的核磁共振成像控制方法,其特征在于,所述预设偏转角度由运动型参考目标对象的预设参考方向与扫描部位的感兴趣区的预设目标方向在扫描层面上的夹角所确定。
- 如权利要求6所述的核磁共振成像控制方法,其特征在于,该方法还包括预先训练所述神经网络模型的步骤,该神经网络模型训练的具体步骤包括:将样本图像输入至待训练的信息识别模型中,得到与样本图像对应的当前输出识别结果;根据当前输出识别结果和样本图像的扫描部位和组织的识别结果的误差,确定所述神经网络模型的损失函数是否收敛;当所述损失函数收敛时,所述神经网络模型训练结束。
- 一种计算机可读存储介质,该计算机可读存储介质存储多条计算机程序指令,应用于计算机装置中,该计算机装置连接有核磁共振成像设备,其特征在于,所述计算机程序指令由计算机装置的处理器执行并实现如权利要求6至9任一项所述基于核磁共振成像控制方法。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910659519.3A CN112285622A (zh) | 2019-07-22 | 2019-07-22 | 核磁共振成像控制方法、装置及计算机可读存储介质 |
CN201910659519.3 | 2019-07-22 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2021012972A1 true WO2021012972A1 (zh) | 2021-01-28 |
Family
ID=74193199
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2020/101520 WO2021012972A1 (zh) | 2019-07-22 | 2020-07-11 | 核磁共振成像控制方法、装置及计算机可读存储介质 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN112285622A (zh) |
WO (1) | WO2021012972A1 (zh) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2023028786A1 (zh) * | 2021-08-30 | 2023-03-09 | 深圳先进技术研究院 | 一种设计非线性梯度线圈的方法以及非线性空间编码方法 |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106373167A (zh) * | 2016-11-15 | 2017-02-01 | 西安交通大学 | 一种基于深度神经网络的压缩传感核磁共振成像方法 |
CN106842084A (zh) * | 2016-12-30 | 2017-06-13 | 上海联影医疗科技有限公司 | 一种磁共振成像方法及装置 |
WO2017106469A1 (en) * | 2015-12-15 | 2017-06-22 | The Regents Of The University Of California | Systems and methods for analyzing perfusion-weighted medical imaging using deep neural networks |
CN108010100A (zh) * | 2017-12-07 | 2018-05-08 | 厦门大学 | 一种基于残差网络的单扫描磁共振定量t2成像重建方法 |
CN108132274A (zh) * | 2017-12-21 | 2018-06-08 | 厦门大学 | 不均匀磁场下回波平面成像无参考扫描图像畸变矫正方法 |
CN109658471A (zh) * | 2018-12-20 | 2019-04-19 | 上海联影医疗科技有限公司 | 一种医学图像重建方法和系统 |
-
2019
- 2019-07-22 CN CN201910659519.3A patent/CN112285622A/zh active Pending
-
2020
- 2020-07-11 WO PCT/CN2020/101520 patent/WO2021012972A1/zh active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2017106469A1 (en) * | 2015-12-15 | 2017-06-22 | The Regents Of The University Of California | Systems and methods for analyzing perfusion-weighted medical imaging using deep neural networks |
CN106373167A (zh) * | 2016-11-15 | 2017-02-01 | 西安交通大学 | 一种基于深度神经网络的压缩传感核磁共振成像方法 |
CN106842084A (zh) * | 2016-12-30 | 2017-06-13 | 上海联影医疗科技有限公司 | 一种磁共振成像方法及装置 |
CN108010100A (zh) * | 2017-12-07 | 2018-05-08 | 厦门大学 | 一种基于残差网络的单扫描磁共振定量t2成像重建方法 |
CN108132274A (zh) * | 2017-12-21 | 2018-06-08 | 厦门大学 | 不均匀磁场下回波平面成像无参考扫描图像畸变矫正方法 |
CN109658471A (zh) * | 2018-12-20 | 2019-04-19 | 上海联影医疗科技有限公司 | 一种医学图像重建方法和系统 |
Also Published As
Publication number | Publication date |
---|---|
CN112285622A (zh) | 2021-01-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101659578B1 (ko) | 자기 공명 영상 처리 방법 및 장치 | |
CN110031786B (zh) | 磁共振图像重建方法、磁共振成像方法、设备及介质 | |
US8290565B2 (en) | System and method for cyclic motion encoding for enhanced visualization of slip interfaces with MRI | |
US9626777B2 (en) | Method and apparatus to generate image data | |
EP1182613B1 (en) | Diagnostic imaging | |
US8861819B2 (en) | Apparatus and method for correcting artifacts of functional image acquired by magnetic resonance imaging | |
US9712789B2 (en) | Method and apparatus to generate image data | |
US20110228998A1 (en) | System and method for automatic computation of mr imaging scan parameters | |
US20120245453A1 (en) | Respiratory interval-based correlation and processing of dynamic imaging data | |
CN106716167A (zh) | 用于评估结构空间频率的基于选择性采样磁共振的方法 | |
US11796617B2 (en) | System and method for reconstruction of magnetic resonance images acquired with partial Fourier acquisition | |
CN109917315A (zh) | 磁共振成像扫描方法、装置、计算机设备和存储介质 | |
WO2014154544A1 (en) | Real-time motion correction for mri using fat navigators | |
US20230204701A1 (en) | Systems and methods for magnetic resonance imaging | |
CN112133410A (zh) | 使用机器学习的mri图像重建 | |
US11071469B2 (en) | Magnetic resonance method and apparatus for determining a characteristic of an organ | |
CN105030238B (zh) | 在多个桌台位置处执行磁共振检查 | |
US10362961B2 (en) | System and method for neutral contrast magnetic resonance imaging of calcifications | |
WO2021012972A1 (zh) | 核磁共振成像控制方法、装置及计算机可读存储介质 | |
US20180100906A1 (en) | Magnetic resonance imaging apparatus and method of operating the same | |
CA3109456A1 (en) | System, method and computer-accessible medium for tissue fingerprinting | |
KR102342813B1 (ko) | 볼륨 선택적 3차원 자기 공명 영상 생성 장치 및 방법 | |
Preiswerk et al. | Hybrid utrasound and MRI acquisitions for high-speed imaging of respiratory organ motion | |
US10466324B2 (en) | Method and magnetic resonance apparatus determining markings on a quantitative image data | |
US12000919B2 (en) | Systems and methods for magnetic resonance imaging |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20844197 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20844197 Country of ref document: EP Kind code of ref document: A1 |
|
32PN | Ep: public notification in the ep bulletin as address of the adressee cannot be established |
Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 01/09/2022) |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20844197 Country of ref document: EP Kind code of ref document: A1 |