Echo spacing shuffling for echo planar imaging

Abstract

The invention provides echo spacing shuffling for echo planar imaging. An EPI magnetic resonance imaging method is described which incorporates a variable density undersampling scheme. The method comprises the steps of generating a HF pulse (HF), applying switched frequency encoding readout gradients (GR) with variable time intervals (EST), and simultaneously applying intermittent marker low amplitude phase encoding gradients (GP) with variable integration values of the phase encoding gradients (GP). The above steps are performed such that k-space is at least partially undersampled, the time intervals (EST) of one readout gradient (GR) are varied in dependence of the integration of the phase encoding gradients (GP) such that a ratio between the variable time intervals (EST) of the readout gradients (GR) and the integration of the corresponding phase encoding gradients (GP) is kept constant and above or at a predetermined value which is related to a predetermined standard of image quality. Furthermore, a MR control sequence determination device (40) is described. Moreover, a MR imaging system (1) is described.

Description

Technical Field

[0001] This invention relates to an EPI magnetic resonance imaging method incorporating a variable density undersampling scheme. Furthermore, this invention relates to an MR control sequence determination apparatus. Additionally, this invention relates to an MR imaging system. Background Technology

[0002] With the help of modern imaging methods, two-dimensional or three-dimensional image data is usually generated. This two-dimensional or three-dimensional image data can be used to visualize the patient, such as a person or animal, and can also be used for other applications.

[0003] Magnetic resonance imaging (MRI) is used to achieve a special type of medical imaging.

[0004] Magnetic resonance imaging (MRI) is a medical imaging technique that provides diagnostic information about biological tissues. Its diagnostic relevance stems from the informational power and versatility of this imaging modality, providing differentiated and precise structural information in a non-invasive manner. MRI is based on the controlled manipulation of nuclear spins within the patient's body and the subsequent detection of the nuclear spin response. Spatial correlation coding of the spin response allows for the reconstruction of the patient's structural and functional composition for diagnostic reading by radiologists.

[0005] In a magnetic resonance imaging (MRI) system, the body being examined is typically exposed to a high fundamental magnetic field, such as 1.5 Tesla, 3 Tesla, or 7 Tesla, via a fundamental field magnet system. After the fundamental field is applied, the atomic nuclei in the object being examined align themselves with their non-zero nuclear magnetic dipole moments (often referred to as spins) along the fundamental field. This collective behavior of the spin system is described macroscopically as "magnetization." Macroscopic magnetization is the vector sum of all the microscopic magnetic moments of the object at a given location. In addition to the fundamental field, a gradient system is used to apply a magnetic field gradient, by which the magnetic resonance frequency (Larmor frequency) at the corresponding location is determined. A high-frequency excitation signal (RF / HF pulse) is then transmitted via a high-frequency transmission system using a suitable antenna arrangement. This causes the spin flip angle of certain atomic nuclei excited by this high-frequency field (i.e., at the Larmor frequency present at the corresponding location) around a defined region to tilt relative to the magnetic field lines of the fundamental magnetic field. If such an RF / HF pulse is applied to the already excited spins, they can be flipped to different angular positions, or even folded back to their initial state parallel to the fundamental magnetic field. During the relaxation of excited nuclear spins, the high-frequency signal of the so-called magnetic resonance signal is resonantly emitted, received by means of a suitable receiving antenna (also known as a magnetic resonance coil or receiving coil), then demodulated and digitized, and then further processed as so-called "raw data". The acquisition of the magnetic resonance signal occurs in the spatial frequency space, the so-called "k-space", where, during measurements such as slices, the signal traverses the k-space over time along a "gradient trajectory" (also known as a "k-space trajectory") defined by the switching of gradient pulses. Furthermore, the RF / HF pulses must be emitted at appropriate times. From the raw data acquired in this way, after additional processing steps, which typically also depend on the acquisition method, the desired image data can ultimately be reconstructed using a two-dimensional Fourier transform.

[0006] Alternatively, the three-dimensional volume can now be excited and read in a defined manner, and the raw data can be sequentially classified into a three-dimensional k-space after further processing steps. The three-dimensional image data volume can then be reconstructed using a three-dimensional Fourier transform.

[0007] Typically, during measurements, a magnetic resonance imaging (MR) system is controlled using predetermined pulse sequences (i.e., sequences of RF / HF pulses and gradient pulses defined by different directions and readout windows), during which the receiving antenna is switched to receive and the magnetic resonance signal is received and processed. These sequences are pre-parameterized for desired examinations, such as calculated image contrast, with the aid of a so-called measurement protocol. The measurement protocol may also include additional control data for the measurements. Numerous magnetic resonance sequencing techniques exist, from which pulse sequences can be constructed. One of the major challenges facing the future development of magnetic resonance imaging (MR imaging) is accelerating magnetic resonance sequencing techniques without making extensive compromises in resolution, contrast, and sensitivity to artifacts.

[0008] Current clinical MR imaging is almost entirely based on so-called Cartesian or rectangular (straight line) imaging, where the scanned k-space points (i.e., the scan points in k-space that record the raw data) are located on grid points of a rectangular grid or pattern.

[0009] In doing so, so-called parallel imaging methods have successfully and significantly accelerated clinical MR imaging. In the case of a particular variant of parallel MR imaging (known as compressed sensing), data acquisition is shortened because some grid lines, which are actually necessary for reconstructing non-convolutional images, are not acquired in k-space. These missing lines are later replaced during image reconstruction in k-space, or convolution artifacts caused by undersampling are removed in image space. A prerequisite for using parallel imaging methods is receiving high-frequency signals using several receiving coils (antennas), where the spatial sensitivity of each receiving coil must be known. The spatial sensitivity of the receiving coils is calculated with the help of so-called coil calibration data. The coil calibration data must typically be adequately sampled. Since sensitivity usually varies slowly in space, this is usually sufficient if the coil calibration data has low spatial resolution. Typically, the coil calibration data must be remeasured for each patient.

[0010] One of the most important parallel imaging methods is the so-called GRAPPA method, which is described, for example, in the following literature: Marc Griswold et al., “Generalized Autocalibrating Partially Parallel Acquisitions (GRAPPA)”, Magnetic Resonance in Medicine, 2002, Vol. 47, pp. 1202-1210.

[0011] For general parallel imaging, or particularly compressed sensing, variable density undersampling schemes are beneficial for improving image quality in accelerated MRI. Using this method, the central portion of k-space is sampled more densely, while more lines or points on the periphery are ignored. Combining compressed sensing with fast acquisition techniques based on echo-plane imaging (EPI) has potential benefits. However, standard EPI is incompatible with variable density undersampling because it disrupts the phase evolution of non-resonant signals, leading to image artifacts.

[0012] Spatial encoding using EPI is achieved by sampling the signal during fast gradient switching, which generates a zigzag trajectory in k-space. This trajectory is along the readout direction (k... x (abbreviated as RO direction) and phase encoding direction (k y There is a difference between the k-space and the PE direction. Along the RO direction, high gradient intensities are used for rapid sampling in k-space, while along the PE direction, small gradient blips are applied at regular intervals corresponding to the low sampling rate in k-space. Strong sensitivity differences can occur at tissue interfaces, such as the boundary between water and air in the frontal cortex of the human brain. This has the effect of contributing to the background magnetic field gradient, which is comparable in magnitude to the PE blip gradient. This disrupts the k-space trajectory of signals emanating from this tissue region, resulting in spatial distortion in the final image.

[0013] To reduce these distortions, parallel imaging can be applied, which uses conventional undersampling to shorten the echo train, thereby reducing the time required to traverse the k-space in the PE direction. This effectively increases the magnitude of the signature PE gradient relative to the background gradient and reduces their impact. Similarly, multi-shot, piecewise k-space techniques can be used to reduce the echo train length, although at the cost of an overall increase in scan time.

[0014] The above EPI method suffers from the following problem: it relies on periodic sampling in k-space to maintain image quality, which is incompatible with the requirement of CS for variable-density undersampling. This is because variable sampling leads to non-uniform step sizes in k-space that do not match the constant phase evolution of the non-resonant signal, which is particularly problematic in regions with background magnetic field gradients. In practice, this produces spatial distortion as a function of spatial frequency, resulting in trailing and loss of detail in the final image.

[0015] One possible approach is to always keep the minimum amplitude of the flag PE gradient above a certain threshold so that no distortion occurs during acquisition. Above this threshold, a variable density undersampling mode can be achieved. However, due to the need for a low threshold, the echo train length cannot be longer than approximately 11 echoes. This reduces the overall acquisition speed, thus making this method unsuitable for applications where imaging speed is critical, such as functional MRI (fMRI) or diffusion-weighted imaging (DWI).

[0016] Therefore, there is a problem in developing a fast MR imaging method with sufficient image quality. Summary of the Invention

[0017] This objective is achieved by the EPI magnetic resonance imaging method combining a variable density undersampling scheme according to the present invention, the MR control sequence determination device according to the present invention, and the MR imaging system according to the present invention.

[0018] The EPI magnetic resonance imaging method according to the present invention, incorporating a variable density undersampling scheme, includes the following steps: generating an HF pulse; applying a switched frequency-coded readout gradient at variable time intervals; and simultaneously applying an intermittent, low-amplitude phase-coded gradient with a variable integral value of a phase-coded gradient. The variable time interval means that the time intervals of at least two of the playout readout gradients include different values. The variable integral value of the phase-coded gradient refers to the fact that the integral values ​​of at least two of the playout phase-coded gradients are different from each other. Performing the above steps results in the k-space being at least partially undersampled; and the time interval of a readout gradient varies according to the integral value of the phase-coded gradient, such that the ratio between the variable time interval of the readout gradient and the corresponding integral value of the phase-coded gradient is kept constant and above or at a predetermined value, which is related to a predetermined standard of image quality. The time interval of a readout gradient is also referred to as the echo interval. The integral value of the phase-coded gradient is proportional to a local undersampling factor in the k-space. The ratio between the variable time interval of the readout gradient and the corresponding integral value of the phase-coded gradient is referred to as the effective echo interval. The effective echo interval can also be defined as the ratio between the echo interval of the readout gradient and the undersampling factor. The change in echo interval corresponding to the undersampling factor is also called echo interval shuffle.

[0019] Advantageously, the method according to the invention allows for the combination of an undersampling scheme with an EPI acquisition scheme that uses a long echo train to acquire large segments of the k-space at each excitation. This combination further reduces scan time. Since the effective echo interval is kept above a predetermined value, image quality is maintained at a predetermined level, but the undersampling scheme varies throughout the k-space. Therefore, any variable-density undersampling scheme combined with an EPI pulse sequence can be used without introducing image artifacts of a higher amount than predetermined.

[0020] The MR control sequence determination apparatus according to the present invention comprises: an HF pulse generation unit for emitting HF pulses; and a gradient generation unit for generating switched frequency-coded readout gradients at variable time intervals and simultaneously generating intermittent flag low-amplitude phase-coded gradients with variable amplitude or variable integral values ​​of phase-coded gradients. The gradient generation unit is designed to emit gradients such that the k-space is at least partially undersampled, and the time interval of a readout gradient varies according to the integral value of the phase-coded gradient, such that the ratio between the variable time interval of the readout gradient and the corresponding integral value of the phase-coded gradient is kept constant and above or at a predetermined value, which is related to a predetermined standard of image quality. The gradient generation unit shares the advantages of the EPI magnetic resonance imaging method according to the present invention.

[0021] The MR imaging system includes a scanning unit and an MR control sequence determination device according to the present invention. The MR imaging system shares the advantages of the MR control sequence determination device.

[0022] Some components of the MR control sequence determination apparatus according to the present invention can be designed largely as software components. This is particularly true for portions of the gradient generation unit and the HF pulse generation unit. However, in principle, some of these components can also be implemented in the form of software-supported hardware such as an FPGA, especially when particularly fast computations are involved. Similarly, the required interfaces—for example, if it is merely a matter of transferring data from other software components—can be designed as software interfaces. However, they can also be designed as hardware-based interfaces controlled by suitable software.

[0023] The largely software-based implementation has the following advantages: the computer unit or control unit of an existing MR imaging system can be easily modified through software updates to operate in accordance with the invention. In this regard, this objective is also achieved by a corresponding computer program product having a computer program that can be directly loaded into the memory device of the computer unit or control unit of the MR imaging system. This computer program includes program segments so that, when executed in the computer unit or control unit of the MR imaging system, all steps are performed to execute the method according to the invention.

[0024] In addition to computer programs, such computer program products may optionally include additional components such as documentation and / or additional components such as hardware components for using the software, such as hardware keys (dongles, etc.).

[0025] A computer-readable medium, such as a memory stick, hard disk, or some other removable or permanently mounted data carrier, stores program segments of a computer program that can be read and executed. This computer program can be used to transfer to the storage device of a computer unit within an MR imaging system and / or for storage on the computer unit of the MR imaging system. For example, the computer unit may have one or more cooperating microprocessors for this purpose.

[0026] Furthermore, the particularly advantageous configurations and developments of the present invention arise from the technical solutions of the invention and the following description and drawings, thereby allowing an independent claim in one claim class to be developed in a manner similar to a dependent claim in another claim class and its descriptive portion.

[0027] In a preferred variant of the EPI magnetic resonance imaging method according to the invention, the amplitude of the phase-encoded gradient is variable, and the time interval of a readout gradient varies according to the amplitude of the corresponding phase-encoded gradient, such that the ratio between the variable time interval of the readout gradient and the amplitude of the corresponding phase-encoded gradient is kept constant and above or at a predetermined value, which is related to a predetermined standard of image quality. In the aforementioned variant, the phase-encoded gradient includes a shape, wherein the amplitude is proportional to the integral value of the phase-encoded gradient. For example, this is the case for a rectangular or triangular phase-encoded gradient.

[0028] In a variant of the EPI magnetic resonance imaging method according to the invention, the ratio between the variable time interval of the readout gradient and the integral value and / or amplitude of the corresponding phase-encoded gradient is kept constant and above or at a predetermined value throughout the sampling process in k-space. Advantageously, the level of distortion is kept constant throughout k-space to maintain image quality.

[0029] In a preferred variant of the EPI magnetic resonance imaging method according to the invention, the undersampling scheme includes a variable density scheme, which is used to acquire raw data for parallel image reconstruction. The combination of parallel imaging and EPI enables further acceleration of EPI acquisition.

[0030] In a highly preferred variant of the EPI magnetic resonance imaging method according to the invention, the undersampling scheme includes a variable density scheme used to acquire the raw data for compressed sensing reconstruction. The combination of compressed sensing and EPI also enables an acceleration of the EPI acquisition process. Furthermore, such a combination allows for sampling of an extended region of k-space within a short time, which is crucial for time-correlated applications such as functional magnetic resonance imaging or diffusion-weighted magnetic resonance imaging methods. Attached Figure Description

[0031] The invention is described in more detail below with reference to the accompanying drawings and exemplary embodiments. In the various drawings, the same components have the same reference numerals.

[0032] Figure 1 Some pulse sequence diagrams of different EPI pulse sequences are shown;

[0033] Figure 2 Some k-space trajectories of pulse sequences for a flag phase-coded gradient EPI scheme combined with an undersampling scheme are shown;

[0034] Figure 3 A flowchart of an EPI magnetic resonance imaging method combined with a variable density undersampling scheme according to an embodiment of the present invention is shown;

[0035] Figure 4 A schematic diagram of an MR control sequence determination apparatus according to an embodiment of the present invention is shown;

[0036] Figure 5 A schematic diagram of an MR imaging system according to an embodiment of the present invention is shown;

[0037] Figure 6 A pulse sequence diagram of an EPI pulse sequence combined with compressed sensing is shown; and

[0038] Figure 7 It shows the relationship with Figure 6 The k-space trajectory corresponding to the pulse sequence shown. Detailed Implementation

[0039] Figure 1 The diagram illustrates three different pulse sequence diagrams according to the prior art, along with the corresponding k-space trajectories of the three different EPI pulse sequences. The first pulse sequence represents an EPI sequence with a continuous phase encoding gradient GP, which in... Figure 1 The upper part is shown. The corresponding k-space trajectory on the right side of the pulse sequence diagram shows a linear variation in frequency and encoded phase. In contrast, Figure 1The second pulse sequence diagram shown in the middle position includes a pulse sequence with a so-called marker gradient GP, which has the effect of decoupling frequency and phase changes, as shown in the k-space diagram on the right side of the second pulse sequence diagram. The third pulse sequence diagram shows some alternating phase-encoded gradients GP and readout gradients GR, which result in a spiral trajectory in k-space, as shown on the right side of the corresponding pulse sequence diagram.

[0040] Figure 2 Some k-space trajectories of pulse sequences from a flag phase-coded gradient (EPI) scheme combined with undersampling are shown. On the upper left, the conventional undersampling EPI scheme is represented, where dashed lines represent ignored readout trajectory lines, and solid lines represent sampled trajectory lines from the EPI scheme. A problem with compressed sensing schemes with conventional undersampling is the appearance of duplication artifacts, which degrade image quality.

[0041] exist Figure 2 The right side shows an undersampling EPI scheme with variable undersampling, where the number of ignored readout traces varies from 2 to 3. The problem with such variable undersampling is that it results in an uneven step size in k-space that doesn't match the constant phase evolution of the non-resonant signal, which is important for regions with background magnetic field gradients. This effect produces spatial distortion as a function of spatial frequency, leading to trailing and loss of detail in the final image.

[0042] Figure 3 A flowchart 300 of an EPI magnetic resonance imaging method combined with a variable density undersampling scheme according to an embodiment of the present invention is shown.

[0043] In step 3.I, the HF pulse HF is emitted. In step 3.II, the switched frequency-coded readout gradient GR is emitted at variable time intervals, and in step 3.III, an intermittent flag low-amplitude phase-coded gradient GP is emitted at variable amplitude, such that the k-space is at least partially undersampled, and the time interval of one readout gradient GR varies according to the integral value of the phase-coded gradient GP. The time interval of one readout gradient GR is changed such that the ratio between the variable time interval of the readout gradient GR and the integral value of the corresponding phase-coded gradient GP is kept constant and above or at a predetermined value, which is related to a predetermined standard of image quality. In step 3.IV, the gradient echo is read out.

[0044] exist Figure 4 The diagram shows an MR control sequence determination device 40 according to an embodiment of the present invention.

[0045] The MR control sequence determination apparatus 40 includes an HF pulse generation unit 41 for generating HF pulses. Furthermore, the MR control sequence determination apparatus 40 also includes a phase-encoded gradient generation unit 42a for generating a phase-encoded pulse sequence GP and a readout gradient generation unit 42b for generating a readout pulse sequence GR.

[0046] The readout gradient generation unit 42b generates a switched frequency-coded readout gradient within a variable time interval, while the phase-coded gradient generation unit 42a simultaneously generates an intermittent low-amplitude phase-coded gradient GP with a variable amplitude A, such that the k-space is at least partially undersampled, and the time interval EST of a readout gradient GR varies according to the integral value of the phase-coded gradient GP, such that the ratio between the variable time interval EST of the readout gradient GR and the integral value of the corresponding phase-coded gradient GP is kept constant and above or at a predetermined value, which is related to a predetermined standard of image quality.

[0047] exist Figure 5 The image schematically illustrates a magnetic resonance system 50 (hereinafter simply referred to as "MR system") according to an embodiment of the present invention. In one aspect, the MR system 50 includes an actual magnetic resonance scanner 52 having an examination room 53 or a patient access passage, into which a patient O or test subject can enter the examination room 53 or patient access passage on a bed 58, where, for example, an organ to be imaged is present within the patient O or test subject.

[0048] The magnetic resonance scanner 52 is conventionally equipped with a basic field magnet system 54, a gradient system 56, and an RF transmitting antenna system 55 and an RF receiving antenna system 57. In the exemplary embodiment shown, the HF transmitting antenna system 55 is a whole-body coil permanently mounted in the magnetic resonance scanner 52, while the HF receiving antenna system 57 consists of local coils arranged on the patient or test subject (in... Figure 5 (This is represented by only a single local coil). However, in principle, the whole-body coil 55 can also be used as an RF receiving antenna system, and the local coil 57 can also be used as an RF transmitting antenna system, as long as these coils can be switched to different operating modes respectively.

[0049] The MR system 50 also includes a central control unit 63 for controlling the MR system 50. This central control unit 63 includes a sequence control unit 64 for pulse sequence control. This is used to control the time series of high-frequency pulses (RF pulses / HF pulses) and gradient pulses GR, GP according to a selected imaging sequence PS. For example, such an imaging sequence can be specified within a measurement or control protocol P. Different control protocols P for different measurements are typically stored in a memory 69 and can be selected by the operator (and changed if needed), and then used to perform the measurement. The sequence control unit 64 also includes a control sequence determination device 40 according to the invention, which has… Figure 4 The structure is shown. The control sequence determination device 40 generates control data SD for the sequence that enables the excitation module and the readout module, and the sequence control unit 64 outputs the control data SD for pulse sequence control.

[0050] To output individual RF pulses, the central control unit 63 has a high-frequency transmitter 65 that generates, amplifies, and feeds the RF pulses to the RF transmitting antenna system 55 via a suitable interface (not shown in detail). To control the gradient coils of the gradient system 56, the central control unit 63 has a gradient system interface 66. The sequence control unit 64 communicates in a suitable manner, for example, by transmitting sequence control data SD, wherein the high-frequency transmitter 65 and the gradient system interface 66 are used to transmit the pulse sequence PS in the order generated by the control sequence determination device 40. The control unit 63 also has a high-frequency receiver 67 (also communicating with the sequence control unit 64 in a suitable manner) to coordinately acquire the magnetic resonance signals received by the RF transmitting antenna system 57. The reconstruction unit 68 accepts the acquired data as raw data RD after demodulation and digitization and reconstructs MR image data from it. This image data BD can then be stored, for example, in a memory 69.

[0051] The central control unit 63 can be operated via a terminal 61 having an input unit 60 and a display unit 59, and therefore the entire MR system 50 can also be operated by an operator via the terminal 61. MR images can also be displayed on the display unit 59, and measurements can be planned and started by means of the input unit 60, possibly in conjunction with the display unit 59. In particular, suitable control protocols with appropriate measurement sequences can be selected as described above, and these suitable control protocols can be modified if necessary.

[0052] The MR system 50 according to the invention, and in particular the central control unit 63, may also have a number of other components, which are not shown in detail herein but are generally present on devices such as network interfaces to connect the entire system to a network and to exchange raw data RD and / or image data BD or parameter cards, as well as other data, such as patient-related data or control protocols.

[0053] For those skilled in the art, how to obtain suitable raw data RD by irradiating with RF / HF pulses and generating a gradient field, and how to reconstruct MR images BD from them, are known in principle, and therefore will not be described in more detail here. Similarly, for those skilled in the art, the wide variety of measurement sequences used to generate dynamic or static images, such as different EPI sequences, are generally known.

[0054] exist Figure 6 The diagram shows a pulse sequence diagram combining compressed sensing EPI pulse sequences. In the top row, a time-dependent HF pulse HF is shown, which is emitted in conjunction with the readout pulse GR shown in the third row. A first echo is generated by emitting a flag phase-coded gradient GP with a first low amplitude A, as shown in the second row, followed by a readout gradient GR with a short echo interval time proportional to the low amplitude of the phase-coded gradient GP. Next, a flag phase-coded gradient GP with an amplitude twice as high as the first amplitude of the flag phase-coded gradient GP is emitted. Then, the readout gradient GR is emitted with a longer echo interval time proportional to the amplitude or integral value of the assigned phase-coded gradient GP. Next, the flag phase-coded gradient GP is emitted with a low amplitude, followed by a short echo interval time EST. Next, the flag phase-coded gradient GP is emitted with a high amplitude three times higher than the previous low amplitude, followed by a long echo interval time EST three times longer than the previous readout gradient GR's echo interval time EST. Finally, a marker phase-coding gradient GP, twice as high as the low-amplitude phase-coding gradient GP, is emitted, along with a corresponding readout gradient GR, twice as long as the short readout gradient GR. Therefore, the effective echo interval, which is the ratio between the amplitude or integral value of the phase-coding gradient GP and the echo interval time EST of the subsequent readout gradient GR, is constant. This constant effective echo interval produces a predetermined image quality.

[0055] exist Figure 7 In the middle, it is shown that... Figure 6 The pulse sequence shown corresponds to the k-space trajectory. For example, from... Figure 7 It can be seen that undersampling is achieved by ignoring the dashed lines in k-space. Furthermore, the trajectory in k-space... yThe direction, i.e., the length of a portion of the phase encoding direction, is related to the k-space trajectory in k. x The length of the subsequent portion in the direction is proportional. Due to undersampling variations, the trajectory in k... x The length in the direction also changes, making the trajectory in k x The length of a portion of the trajectory in the direction of k y The ratio between the lengths of the front portions in the direction is constant. This results in a constant effective echo interval. It must be mentioned that the effective echo interval can also be constant and higher than a predetermined value, resulting in image quality exceeding the minimum requirements. In fact, a higher effective echo interval leads to a longer echo interval time or a reduced undersampling factor, which means that the imaging process requires more time and / or acquires more raw data.

[0056] Finally, it should be reiterated that the methods and apparatus described in detail above are merely exemplary embodiments, and those skilled in the art can modify them in various ways without departing from the scope of the invention. Furthermore, the use of the indefinite article "a" or "an" does not preclude the possibility that the relevant feature may appear more than once. Similarly, the term "unit" does not preclude the fact that the relevant component is composed of several interacting sub-components, which may also be spatially distributed.

Claims

1. An EPI magnetic resonance imaging method, said method in combination with a variable density undersampling scheme and comprising the following steps: - Generate HF pulses (HF). - Apply frequency-coded readout gradients (GR) with variable time intervals (EST), and - Simultaneously apply an intermittent flag low-amplitude phase-coded gradient (GP) with a variable integral value, such that - The k-space is at least partially undersampled, and - The time interval (EST) of a readout gradient (GR) varies according to the integral value of the phase-coded gradient (GP), such that the ratio between the variable time interval (EST) of the readout gradient (GR) and the integral value of the corresponding phase-coded gradient (GP) is kept constant and above or at a predetermined value throughout the sampling process in k-space, the predetermined value being related to a predetermined standard of image quality, wherein the integral value of the corresponding phase-coded gradient is proportional to a local undersampling factor in k-space.

2. The EPI magnetic resonance imaging method according to claim 1, wherein, - The magnitude of the phase-encoded gradient (GP) is variable, and - The time interval (EST) of a readout gradient (GR) varies according to the magnitude (A) of the phase-encoded gradient (GP), such that the ratio between the variable time interval (EST) of the readout gradient (GR) and the magnitude (A) of the corresponding phase-encoded gradient (GP) is kept constant and above or at a predetermined value throughout the sampling process in k-space, the predetermined value being related to a predetermined standard of image quality.

3. The EPI magnetic resonance imaging method according to claim 1 or 2, wherein, The variable density undersampling scheme includes a variable density scheme along the phase encoding direction, which is used to acquire raw data for image reconstruction based on a parallel imaging scheme.

4. The EPI magnetic resonance imaging method according to claim 1 or 2, wherein, The variable density undersampling scheme includes a variable density scheme along the phase encoding direction, which is used to obtain raw data for image reconstruction based on a compressed sensing scheme.

5. The EPI magnetic resonance imaging method according to claim 1 or 2, including the functional magnetic resonance imaging method.

6. The method according to claim 5, comprising diffusion-weighted magnetic resonance imaging.

7. An MR control sequence determination device (40), comprising: - HF pulse generation unit (41), the HF pulse generation unit is used to generate HF pulses. - Gradient generation units (42a, 42b), the gradient generation units being used for: - Generate switching frequency-coded readout gradients (GR) using variable time intervals (EST) and - Simultaneously generate intermittent flag low-amplitude phase-coded gradients (GP) with variable integral values, such that - The k-space is at least partially undersampled, and - The time interval (EST) of a readout gradient (GR) varies according to the integral value of the phase-coded gradient (GP), such that the ratio between the variable time interval (EST) of the readout gradient (GR) and the integral value of the corresponding phase-coded gradient (GP) is kept constant and above or at a predetermined value throughout the sampling process in k-space, the predetermined value being related to a predetermined standard of image quality, wherein the integral value of the corresponding phase-coded gradient is proportional to a local undersampling factor in k-space.

8. An MR imaging system (50), comprising: - Scanning unit (52) - The MR control sequence determination device (40) according to claim 7.

9. A computer program product having a computer program that can be directly loaded into the storage unit of a control unit (63) of an MR imaging system (50), the computer program having program segments to perform all steps of the method according to any one of claims 1 to 6 when the computer program is executed in the control unit (63).

10. A computer-readable medium storing a program segment executable by a computer unit so as to perform all steps of the method according to any one of claims 1 to 6 when the program segment is executed by the computer unit.

Images