(19)
(11)EP 3 093 680 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
27.11.2019 Bulletin 2019/48

(21)Application number: 15167569.1

(22)Date of filing:  13.05.2015
(51)International Patent Classification (IPC): 
G01R 33/3415(2006.01)
G01R 33/565(2006.01)

(54)

PHASE CORRECTION METHOD FOR MRI

PHASENKORREKTURVERFAHREN FÜR MAGNETRESONANZTOMOGRAPHIE

PROCÉDÉ DE CORRECTION DE PHASE POUR L'IRM


(84)Designated Contracting States:
AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

(43)Date of publication of application:
16.11.2016 Bulletin 2016/46

(73)Proprietor: Medizinische Universität Wien
1090 Vienna (AT)

(72)Inventor:
  • Robinson, Simon Daniel
    1160 Wien (AT)

(74)Representative: Weiser, Andreas 
Patentanwalt Kopfgasse 7
1130 Wien
1130 Wien (AT)


(56)References cited: : 
EP-A1- 1 310 211
US-B1- 6 483 308
US-A1- 2012 321 162
  
  • VAN DER WEERD L ET AL: "Evaluation of algorithms for analysis of NMR relaxation decay curves", MAGNETIC RESONANCE IMAGING, ELSEVIER SCIENCE, TARRYTOWN, NY, US, vol. 18, no. 9, 1 November 2000 (2000-11-01), pages 1151-1157, XP002448420, ISSN: 0730-725X, DOI: 10.1016/S0730-725X(00)00200-9
  • ZHOU K ET AL: "Coil Combination Method for Multiple-Echo Sequences and PSF Mapping", INTERNATIONAL SOCIETY FOR MAGNETIC RESONANCE IN MEDICINE. SCIENTIFIC MEETING AND EXHIBITION. PROCEEDINGS, INTERNATIONAL SOCIETY FOR MAGNETIC RESONANCE IN MEDICINE, US, vol. 17, 4 April 2009 (2009-04-04), page 2777, XP002657873, ISSN: 1524-6965
  • SIMON ROBINSON ET AL: "Combining Phase Images From Multi-Channel RF Coils Using 3D Phase Offset Maps Derived From a Dual-Echo Scan", MAGNETIC RESONANCE IN MEDICINE, JOHN WILEY & SONS, INC, US, vol. 65, no. 6, 19 January 2011 (2011-01-19), pages 1638-1648, XP002657573, ISSN: 0740-3194, DOI: 10.1002/MRM.22753
  • JIE LUO ET AL: "Gradient Echo Plural Contrast Imaging - Signal model and derived contrasts: T2*, T1, Phase, SWI, T1f, FST2*and T2*-SWI", NEUROIMAGE, vol. 60, no. 2, 27 January 2012 (2012-01-27), pages 1073-1082, XP055030807, ISSN: 1053-8119, DOI: 10.1016/j.neuroimage.2012.01.108
  • NEWBOULD R D ET AL: "Phase-preserving multi-coil combination with improved intensity modulation", PROCEEDINGS OF THE INTERNATIONAL SOCIETY FOR MAGNETIC RESONANCE IN MEDICINE, 16TH SCIENTIFIC MEETING AND EXHIBITION, TORONTO, CANADA, 3-9 MAY 2008,, vol. 16, 19 April 2008 (2008-04-19), page 1280, XP007915541,
  • ROBINSON ET AL: "COMbining Phased array data using Offsets from a Short Echo-time Reference scan (COMPOSER)", PROCEEDINGS OF THE INTERNATIONAL SOCIETY FOR MAGNETIC RESONANCE IN MEDICINE, 23ND ANNUAL MEETING AND EXHIBITION, TORONTO, CANADA, 30 MAY - 5 JUNE 2015, vol. 23, 15 May 2015 (2015-05-15), page 3308, XP040668984,
  • HAMMOND ET AL: "Development of a robust method for generating 7.0 T multichannel phase images of the brain with application to normal volunteers and patients with neurological diseases", NEUROIMAGE, ACADEMIC PRESS, ORLANDO, FL, US, vol. 39, no. 4, 7 November 2007 (2007-11-07), pages 1682-1692, XP022498949, ISSN: 1053-8119, DOI: 10.1016/J.NEUROIMAGE.2007.10.037
  • ROEMER P B ET AL: "THE NMR PHASED ARRAY", MAGNETIC RESONANCE IN MEDICINE, JOHN WILEY & SONS, INC, US, vol. 16, no. 2, 1 November 1990 (1990-11-01), pages 192-225, XP000175903, ISSN: 0740-3194
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description


[0001] The present invention relates to a method for Magnetic Resonance Imaging (MRI) to depict a 3-dimensional object by an image having pixels representing volume elements of the object.

[0002] MRI is used in radiology to visualize details of structures in a patient's body. For aligning the magnetic spin of the nuclei, mostly of protons in water molecules in the body tissue, the patient is placed inside a powerful static magnetic field. Excited by an electro-magnetic radio-frequency pulse from a transmitter coil the nuclei resonating at this frequency deflect and then gradually relax towards the static field while emitting detectable electro-magnetic radiation, which can be captured as an "echo" at a certain time after excitation (the "echo time") by a receiver coil. Relaxation times and the resonance frequency of the nuclei depend both on local properties of the tissue material, which represents the underlying principle allowing visualization of these properties. The characteristics of the image are also influenced by proton density and magnetic field strength. By superposing lateral and longitudinal gradient fields and adjusting the excitation frequency, certain volumetric regions ("volume elements") can be measured by both selective excitation and frequency analysis of the captured echo signals. Fourier-transforming the latter generates complex representations of the frequency values and their phases, which can be made readable.

[0003] When the excitation is performed on 2-dimensional slices through the object, they are selected by adjusting the gradients in the magnetic field strength, thereafter resonant frequency values are captured and processed, forming 2-dimensional images, a number of which can be merged to form a 3-dimensional representation of the object. Alternatively, in 3-dimensional imaging, a large volume of tissue is excited and spatially encoded using frequency encoding in one direction and phase encoding in both of the remaining two orthogonal directions.

[0004] In the past, magnitude values have been used primarily from the complex representations. Nevertheless, phase values allow for the extraction of additional information about local properties inside the tissue, where they specifically benefit from strong susceptibility effects at high magnetic field strengths. For example, phase information is used in neuroimaging in phase-contrast angiography, Susceptibility-Weighted Imaging (SWI), susceptibility mapping - also known as Quantitative Susceptibility Mapping (QSM), Susceptibility Tensor Imaging, to depict iron accumulation in neurodegenerative disorders and to map in vivo conductivity. It can also be used to monitor temperature and encode flow velocity.

[0005] However, each phase value acquired by a receiver coil of the MRI machine is subject to a time-independent offset, often referred to as "phase offset". The phase offset comprises spatially constant components, e.g. due to the cable length from a receiver coil to a receiver, as well as spatially variable components, e.g. due to the path lengths of the excitation and echo signals from particular locations in the object to the receiver coil in question, reflecting inhomogeneities in the magnetic field.

[0006] It has been an aim of research to eliminate the effects of the phase offset, e.g. in order to facilitate combining multiple phase images acquired with a plurality of receiver coils arranged in an array around the object, and thereby increase the quality of an acquired image in terms of its signal-to-noise ratio. Several approaches have been presented in the past: One of these approaches, presented by Hammond, K.E. et al., "Development of a robust method for generating 7.0 T multichannel phase images of the brain with application to normal volunteers and patients with neurological diseases", NeuroImage 2008, 39; pp. 1682 - 1692, suggests to estimate a spatially constant phase offset by setting the phase values to zero in all coils at the centre of an image. This method, while being easy to apply, results in areas of poor phase matching. An alternative solution is to refer the phase values of each receiver coil to a "virtual reference coil" which is the result of a two-step procedure. In the first step, a combined image (the Virtual Reference Coil, or VRC, image) is generated using an image-based constant (as in the method of Hammond et al.). In the second step, the phase image from each coil is referenced to the VRC image. While the matching of phase values of different receiver coils is very good in this case, local field variations arising from magnetic susceptibility are not reflected. Yet another solution, proposed by Roemer, P.B. et al., "The NMR phased array", Magn Reson Med 1990, 16; pp. 192 - 225, uses an additional body coil or other homogeneous volume reference coil - i.e. a coil which is separate from said receiver coils and has to be sensitive over (at least) all the tissue over which the receiver coils, arranged in an array around the object, are sensitive - for referencing and using a phase offset separately measured by means of the body coil for each receiver coil; however, such an additional body coil is not commonly available in ultra-high field scanners and requires extra space and control.

[0007] Even if only magnitude values shall be used, discarding the phase information introduces additional noise when a relaxation decay of each pixel of an object is determined, as van der Weerd, L. et al., "Evaluation of Algorithms for Analysis of NMR Relaxation Decay Curves", Magnetic Resonance Imaging 2000, 18; pp. 1151 - 1157, showed. In case of severe phase distortion, this document suggests an algorithm to separately correct the phase values of even and odd echoes by respectively subtracting the phase value of the first echo image. Thereby, a non-zero base-line in the decay curves can be avoided when acquiring an echo train of, e.g., forty-six equally spaced echoes for each pixel.

[0008] It is an object of the present invention to provide a method for magnetic resonance imaging which does not rely on an additional volume coil (e.g. body coil) or coarse estimations and yields an accurate image of an object at low computing time.

[0009] This object is achieved with a method according to independent claim 1 for Magnetic Resonance Imaging to depict a 3-dimensional object by an image having pixels representing volume elements of the object, comprising:

immobilising the object and acquiring a reference image of the object with a receiver coil at a first echo time immediately following an excitation by a transmitter coil, wherein said reference image is complex-valued, representing each volume element by a pixel with a reference magnitude value and a reference phase value;

keeping the object immobilised and acquiring a target image of the object with said receiver coil at a pre-selected second echo time, longer than said first echo time, following the same or another excitation by said transmitter coil, wherein said target image is complex-valued, representing each volume element by a pixel with a target magnitude value and a target phase value;

subtracting, pixel by pixel, the reference phase value from the target phase value to obtain a corrected phase value for each pixel; and

obtaining said image from said target magnitude values and said corrected phase values,

wherein the aforementioned steps are applied to each of a plurality of receiver coils arranged around said immobilised object to obtain a respective plurality of said images, followed by the step of calculating a combined phase image, pixel by pixel, according to

with

∠(·)
denoting the four-quadrant tangent inverse operator,
mT,p
being the target magnitude value of a pixel, representing a volume element of the object, in a target image acquired with the pth receiver coil,
ϑT,p
being the target phase value of said pixel,
ϑR,p
being the reference phase value of a pixel, representing said volume element, in a reference image acquired with the pth receiver coil, and
ϑS
being the phase value of a pixel, representing said volume element, in the combined phase image.



[0010] The method of the invention is particularly efficient and easy to implement, without requiring a volume reference coil (e.g. body coil), and accurate in the resulting image. An image obtained by the method is virtually without phase offset and can therefore be compared to or combined with other images obtained likewise from the same object using other receiver coils. The excitation can either be a single excitation pulse or a series of pulses which can be generated either from a single transmitter coil or an array of transmitter coils ("parallel transmit"). Thereby, the present method can be applied when parallel imaging is used to increase the signal-to-noise ratio and/or to reduce the acquisition time with a plurality of receiver coils arranged as phased array coil elements around the object. The method can be combined with image reconstruction techniques both in the image domain, i.e. after Fourier-transformation, and in the frequency domain, i.e. before Fourier-transformation, and is specifically beneficial for MRI at high magnetic field strengths.

[0011] By applying the method to a plurality of receiver coils accordingly, the resulting offset-compensated images can be easily combined to obtain the combined phase image for further processing or evaluation. The combined phase image is more accurate and robust and the phase values thereof represent the phase values of the volume elements of the object better than an image from a single receiver coil would generally do. If desired, the combined phase image can subsequently be unwrapped.

[0012] In a preferred embodiment of the invention said first echo time is less than 1 ms. Thereby, there is very little reduction in the contrast in the image relating to magnetic susceptibility effects. It is specifically preferred when said first echo time is less than 100 µs. This leads to an even better correction of the phase offset.

[0013] To save time during acquisition, it is further preferred that the reference image is acquired at a lower pixel resolution than the target image and, prior to said subtracting, is upscaled to the pixel resolution of the target image. Saving time during acquisition of the reference image is particularly desirable to keep the examination time and the discomfort of immobilisation short for the patient.

[0014] It is advantageous when the method is applied to each of a plurality of receiver coils arranged around said immobilised object to obtain a respective plurality of said images, followed by the step of calculating a combined magnitude image, pixel by pixel, according to

with
|·|
denoting the magnitude operator,
mT,p
being the target magnitude value of a pixel, representing a volume element of the object, in a target image acquired with the pth receiver coil,
ϑT,p
being the target phase value of said pixel,
ϑR,p
being the reference phase value of a pixel, representing said volume element, in a reference image acquired with the pth receiver coil, and
mS
being the magnitude value of a pixel, representing said volume element, in the combined magnitude image.


[0015] Thereby, random background noise is eliminated or at least significantly reduced; consequently, the signal-to-noise ratio of the combined magnetic image is considerably increased as compared to the signal-to-noise ratio of a single one of the images.

[0016] In an additional or alternative embodiment, the method is advantageously applied to each of a plurality of receiver coils arranged around said immobilised object to obtain a respective plurality of said images, followed by the step of calculating a combined magnitude image, pixel by pixel, according to

with
|·|
denoting the magnitude operator,
mT,p
being the target magnitude value of a pixel, representing a volume element of the object, in a target image acquired with the pth receiver coil,
ϑT,p
being the target phase value of said pixel,
ϑR,p
being the reference phase value of a pixel, representing said volume element, in a reference image acquired with the pth receiver coil, and
mW
being the magnitude value of a pixel, representing said volume element, in the combined magnitude image.


[0017] Likewise, this results in an effective reduction of background noise; moreover, by introducing a weighting by the magnitude values and applying a root-sum of squares thereto the resulting signal-to-noise ratio in the combined image is particularly high. For decreasing the acquisition and examination time, it is favourable when the respective reference or target images acquired with said plurality of receiver coils are all acquired following one and the same excitation.

[0018] The invention will now be described in further details by means of exemplary embodiments thereof under reference to the enclosed drawings, in which:

Fig. 1 shows the generation of a complex-valued image in Magnetic Resonance Imaging according to the state of the art;

Fig. 2 shows a flow chart of different embodiments of the method of the invention for Magnetic Resonance Imaging;

Fig. 3 depicts a timing diagram of an exemplary acquisition cycle used in the method of Fig. 2; and

Figs. 4a to 4d depict exemplary reference (Fig. 4a) and target images (Fig. 4b) acquired and complex-valued images (Fig. 4c) obtained according to the method of Fig. 2, and a magnitude image and phase image (Fig. 4d) calculated according to the method of Fig. 2.



[0019] Magnetic Resonance Imaging (MRI) is used in radiology to visualize soft tissues, non-invasively and in vivo. The process of generating an image of a patient usually consists of the following steps: creating a bulk (longitudinal) magnetisation in the tissue by placing and immobilising the patient inside a powerful static magnetic field; creating regional variation in this magnetic field, and thereby in the resonant frequency and phase of the nuclei, with three comparatively small, linear perpendicular magnetic fields ("gradients"); disturbing the magnetisation with one or more pulses of radio-frequency (RF) electromagnetic radiation ("excitation") applied at the resonant frequency by one or more transmitter coils, tipping the magnetisation into the transverse plane (which is perpendicular to the static magnetic field); and acquiring the RF signals emitted by the tissues as the magnetisation relaxes to the longitudinal direction, by one or more receiver coils.

[0020] In 2-dimensional tomographic imaging, space encoding of the signal works as follows. The first gradient field ("slice select") is applied during RF excitation, so that only spins in a narrow section of tissue are excited. The second ("readout") is applied while the signal is being acquired, so that spins along the readout axis are encoded by their resonant frequency. A number of such excitation-readout steps are acquired with differing applications of the third ("phase-encode") gradient, which encodes the signal along that gradient direction according to a dephasing rate. In 3-dimensional imaging, slice encoding is replaced by a second loop of phase-encoding steps in the slice gradient direction.

[0021] The RF signals emitted by the patient are captured as "echoes" at a certain time after excitation, by one or more receiver coils. Fourier-transforming the acquired MR-signals generates images of the patient, which consist of a large number of pixels representing volume elements, reflecting the local proton density and magnetic properties of the tissue. The acquired MR signals are complex-valued; images of the patient, as the Fourier-transform of the acquired signal, are therefore likewise complex-valued. That is, image signals consist of a magnitude value and a phase value and can be represented in conventional complex number notation.

[0022] Some MRI methods use only the magnitude of the MR-signal. Nevertheless, the phase value contains additional information, which can be clinically useful. While the magnitude value of the signal decays exponentially with echo time, the phase value evolves linearly, and reflects local deviation from the main magnetic field strength. The sensitivity of phase to local magnetic field also allows local iron (which is highly paramagnetic) to be imaged. Phase values can be used, in combination with magnitude values, e.g. to depict veins, due to the iron content of the deoxyhemoglobin iron, in a technique known as Susceptibility-Weighted Imaging. These techniques benefit from high static magnetic field, which provides enhanced magnetic susceptibility effects and higher quality images due to increased signal-to-noise ratio (SNR).

[0023] As shown in Fig. 1, captured signals of an acquired complex-valued signal image 1 of an object to be investigated cannot be interpreted straightforwardly. Fourier-transforming image 1 results in a complex-valued image 2 which can be separated on a pixel-by-pixel basis into a magnitude image M and a phase image Θ.

[0024] However, the phase image Θ suffers from a conceptual ambiguity: As adding 2π to the phase of a signal results in the same measured phase value, the encoding range in captured phase values is effectively limited to 2π radians. Variations in phase values of an object when passing through 2π lead to discontinuities in the phase image Θ known as "phase wraps" 3, which distort the readability and obscure interesting phase features. Different algorithms are known in the art to remove such phase wraps 3 from a phase image Θ.

[0025] Moreover, the phase values of the phase image Θ contain a time-independent phase offset, which, inter alia, depends on the position of the receiver coil of the MRI machine relative to the object to be examined and, to a certain extent, on the individual volume element to be examined. Phase images acquired by different receiver coils which generally are arranged as phased array coil elements around the 3-dimensional object can therefore not be combined with ease.

[0026] With reference to Figs. 2 to 4, different embodiments of a method for MRI accommodating these phase offsets shall now be described.

[0027] According to Fig. 2 in a first step 4 of said method a first or reference image SR of the object, here: the reference image SR,p of an exemplary receiver coil from a multitude of receiver coils - with indices p = 1, 2, 3,... - of the MRI machine (see Fig. 4a), is acquired at a first or reference echo time TER (Fig. 3). Said reference echo time TER immediately follows the respective excitation of the object by the transmitter coil of the MRI machine. In fact, said reference echo time TER may be the shortest echo time selectable for said receiver coil; it can be less than 1 ms or even less than 100 µs. The shortest echo time selectable, inter alia, depends on the MRI machine and the measurement sequence applied, e.g. on the sequence for space encoding, or on whether or not the receiver coil is also used as a transmitter coil, etc.

[0028] As can be seen from the schematic representation of Fig. 3, acquiring each reference image SR or SR,p, respectively, takes a certain, finite acquisition period 5; similarly, the excitation is conducted by a pulse 6 of finite length. The excitation could alternatively be conducted by a predefined series of pulses, each pulse 6 for acquiring one or more volume elements step-by-step. Any excitation, whether a single pulse or a series of pulses, may be generated by a single or a plurality of transmitter coils, as known in the art.

[0029] As shown in Fig. 4a, said reference images SR,1, SR,2, ..., SR,p are - as a Fourier-transformation of complex measured data - complex-valued, similarly to the image 2 of Fig. 1. Each of the complex-valued reference images SR,p can therefore be separated into a reference magnitude image MR,p and a reference phase image ΘR,p, in which the volume elements of the object are represented by complex-valued pixels comprising each a reference magnitude value mR,p and a reference phase value ϑR,p, respectively (shown without index for the respective pixel in Fig. 4a).

[0030] The phase value ϑR,p acquired in a receiver coil at an echo time TER depends both on the local deviation from the static magnetic field ΔB0 and the phase offset ϑο,p for that receiver coil according to ϑR,p = 2πγ·ΔB0·TER + ϑo,p (neglecting phase wraps.) Hence, in the limit TER → 0, a phase value ϑR,p approximates the phase offset ϑo,p.

[0031] Reverting to Fig. 2, in a step 7 - which can be performed using the same excitation pulse 6 or pulses as in step 4, or using a different excitation pulse 6 or pulses - a second or "target" image ST of the object (here: the target image ST,p of the receiver coil with index p, see Fig. 4b), is acquired at a pre-selected second or "target" echo time TET over a finite acquisition period 8 (Fig. 3), while keeping the object immobilised with respect to step 4. The target echo time TET is usually significantly longer than said reference echo time TER.

[0032] The reference echo time TER and/or the target echo time TET may follow the same excitation for a multitude of receiver coils; alternatively, for each receiver coil different reference and/or target echo times TER,p, TET,P, following the same or a separate excitation for each coil could be used.

[0033] The target images ST,p are complex-valued, representing the volume elements of the object selected by pixels comprising each a target magnitude value mT,p and a target phase value ϑT,p, respectively (pixel index not shown in Fig. 4b).

[0034] As shown in Figs. 4a and 4b, both the reference images SR,p and the target images ST,p can be separated into reference magnitude images MR,p and reference phase images ΘR,p, as well as target magnitude images MT,p and target phase images ΘT,p, respectively.

[0035] According to an optional embodiment of the method, in step 4 the reference images SR,p may be acquired at a lower pixel resolution than the target images ST,p in step 7. In this case, the reference images SR,P are each upscaled in a step 9 to the pixel resolution of the respective target image ST,p. Such an upscaling step 9 can be based on generally known numeric algorithms for interpolation and/or extrapolation.

[0036] In a step 10 following the steps 4 and 7 (and step 9, where applicable) the reference phase values ϑR,p of a reference image SR,p are subtracted from the target phase values ϑT,p of the respective target image ST,p, pixel by pixel, to obtain a corrected phase value ϑK,p for each pixel, and a complex-valued image SK,P is composed from said target magnitude values mT,p and said corrected phase values ϑK,p, e.g. according to

with
mT,p
being the target magnitude value of a pixel, representing a volume element of the object, in a target image acquired with the pth receiver coil,
ϑT,p
being the target phase value of said pixel,
ϑR,p
being the reference phase value of a pixel, representing said volume element, in a reference image acquired with the pth receiver coil,
ϑK,p
being the corrected phase value of a pixel, representing said volume element, in the image SK,p acquired the pth receiver coil, and
sK,p
being the complex value of said pixel in the image SK,p.


[0037] As shown in Figs. 4a to 4c, method steps 4, (9), 7 and 10 described so far are applied to each of a plurality of receiver coils in the same way, yielding a respective plurality of said images SK,p. The images SK,p may each be separated into a target magnitude image MT,p and a corrected phase image ΘK,p, which are formed of the target magnitude values mT,p and the corrected phase values ϑK,p, respectively, and which could be evaluated separately. In a preferred embodiment of the method, however, the target magnitude images MT,p and corrected phase images ΘK,p acquired with different receiver coils are evaluated in combination, as will now be detailed in the following.

[0038] A combined phase image ΘS is calculated, pixel by pixel, in a following (optional) step 11 according to

with
(·)
denoting an angle operator, which, e.g., can be implemented as a four-quadrant tangent inverse operation (generally known as "atan2"-function in trigonometry), and
ϑS
being the phase value of a pixel, representing a respective volume element, in the combined phase image ΘS.


[0039] In addition thereto, in an optional step 12, which may be executed prior to, in parallel with, or after step 11, a combined magnitude image MS can be calculated, pixel by pixel, according to:

and/or, see optional step 13 in Fig. 2, according to:

with
|·|
denoting the magnitude operator (generally known from absolute value operations),
mS
being the magnitude value of a pixel, representing a respective volume element, in the combined magnitude image MS, and
mW
being the magnitude value, based on a root-sum of squares calculation in eq. 4, of a pixel, representing a respective volume element, in the combined magnitude image MS.


[0040] Step 13, when applied, can be executed prior to, in parallel with, or after step 11 and step 12.

[0041] As the example of Fig. 4d shows, the combined phase image ΘS is phase offset-compensated and yields phase values ϑS for each pixel which represent the respective volume element more reliably than the target phase values ϑT,p of a single target phase image ΘT,p; and the combined magnitude image MS - whether calculated according to eq. 3 or eq. 4 - shows a substantial increase in SNR as compared to any single target magnitude image MT,p. Thereby, the structures in a patient's body are depicted in significantly more detail.

[0042] The invention is not limited to the embodiments described in detail above, but encompasses all variants and modifications thereof which will become apparent to the person skilled in the art from the present disclosure and which fall into the scope of the appended claims.


Claims

1. A method for Magnetic Resonance Imaging to depict a 3-dimensional object by an image having pixels representing volume elements of the object, comprising:

immobilising the object and acquiring (4) a reference image (SR,p) of the object with a receiver coil at a first echo time (TER) immediately following an excitation (6) by a transmitter coil, wherein said reference image (SR,p) is complex-valued, representing each volume element by a pixel with a reference magnitude value (mR,p) and a reference phase value (ϑR,p);

keeping the object immobilised and acquiring (7) a target image (ST,p) of the object with said receiver coil at a pre-selected second echo time (TET), longer than said first echo time (TER), following the same or another excitation (6) by said transmitter coil, wherein said target image (ST,p) is complex-valued, representing each volume element by a pixel with a target magnitude value (mT,p) and a target phase value (ϑT,p);

subtracting (10), pixel by pixel, the reference phase value (ϑR,p) from the target phase value (ϑT,p) to obtain a corrected phase value (ϑK,p) for each pixel; and

obtaining (10) said image (SK,p) from said target magnitude values (mT,p) and said corrected phase values (ϑK,p),

characterised in that

the aforementioned steps (4, 7, 10) are applied to each of a plurality of receiver coils arranged around said immobilised object to obtain a respective plurality of said images (SK,p), followed by the step (11) of calculating a combined phase image (ΘS), pixel by pixel, according to

with

∠(·) denoting the four-quadrant tangent inverse operator,

mT,p being the target magnitude value of a pixel, representing a volume element of the object, in a target image (ST,p) acquired with the pth receiver coil,

ϑT,p being the target phase value of said pixel,

ϑR,p being the reference phase value of a pixel, representing said volume element, in a reference image (SR,p) acquired with the pth receiver coil, and

ϑS being the phase value of a pixel, representing said volume element, in the combined phase image (ΘS).


 
2. The method of claim 1, characterised in that said first echo time (TER) is less than 1 ms.
 
3. The method of claim 2, characterised in that said first echo time (TER) is less than 100 µs.
 
4. The method of any one of claims 1 to 3, characterised in that the reference image (SR,p) is acquired at a lower pixel resolution than the target image (ST,p) and, prior to said subtracting, is upscaled (9) to the pixel resolution of the target image (ST,p).
 
5. The method of any one of claims 1 to 4, followed by the step (12) of calculating a combined magnitude image (MS), pixel by pixel, according to

with

|·| denoting the magnitude operator,

mT,p being the target magnitude value of a pixel, representing a volume element of the object, in a target image (ST,p) acquired with the pth receiver coil,

ϑT,p being the target phase value of said pixel,

ϑR,p being the reference phase value of a pixel, representing said volume element, in a reference image (SR,p) acquired with the pth receiver coil, and

mS being the magnitude value of a pixel, representing said volume element, in the combined magnitude image (MS).


 
6. The method of any one of claims 1 to 4, followed by the step (13) of calculating a combined magnitude image (MS), pixel by pixel, according to

with

|·| denoting the magnitude operator,

mT,p being the target magnitude value of a pixel, representing a volume element of the object, in a target image (ST,p) acquired with the pth receiver coil,

ϑT,p being the target phase value of said pixel,

ϑR,p being the reference phase value of a pixel, representing said volume element, in a reference image (SR,p) acquired with the pth receiver coil, and

mW being the magnitude value of a pixel, representing said volume element, in the combined magnitude image (MS).


 
7. The method of any one of claims 1 to 6, characterised in that the respective reference or target images (SR,p, ST,p) acquired with said plurality of receiver coils are all acquired following one and the same excitation (6).
 


Ansprüche

1. Ein Verfahren für Magnetresonanztomographie, um ein dreidimensionales Objekt durch ein Bild mit Pixeln darzustellen, welche Volumselemente des Objekts repräsentieren, umfassend:

Ruhighalten des Objekts und Erfassen (4) eines Referenzbildes (SR,p) des Objekts mit einer Empfangsspule zu einer ersten Echozeit (TER), welche unmittelbar auf eine Anregung (6) durch eine Sendespule folgt, wobei das genannte Referenzbild (SR,p) komplexwertig ist und jedes Volumselement durch ein Pixel mit einem Referenz-Amplitudenwert (mR,p) und einem Referenz-Phasenwert (ϑR,p) repräsentiert;

Beibehalten des Ruhighaltens des Objekts und Erfassen (7) eines Targetbildes (ST,p) des Objekts mit der genannten Empfangsspule nach einer vorgewählten zweiten Echozeit (TET), welche länger als die genannte erste Echozeit (TER) ist und auf die gleiche oder eine andere Anregung (6) durch genannte Sendespule folgt, wobei das genannte Targetbild (ST,p) komplexwertig ist und jedes Volumselement durch ein Pixel mit einem Target-Amplitudenwert (mT,p) und einem Target-Phasenwert (ϑT,p) repräsentiert;

pixelweises Subtrahieren (10) des Referenz-Phasenwertes (ϑR,p) vom Target-Phasenwert (ϑT,p), um einen korrigierten Phasenwert (ϑK,p) für jedes Pixel zu erhalten; und

Erhalten (10) des genannten Bildes (SK,p) aus den genannten Target-Amplitudenwerten (mT,p) und den genannten korrigierten Phasenwerten (ϑK,p),

dadurch gekennzeichnet, dass

die vorgenannten Schritte (4, 7, 10) auf jede einer Mehrzahl von Empfangsspulen angewendet werden, welche um das genannte ruhiggehaltene Objekt angeordnet sind, um eine entsprechende Mehrzahl der genannten Bilder (SK,p) zu erhalten, gefolgt vom Schritt (11) des pixelweisen Berechnens eines kombinierten Phasenbildes (ΘS) gemäß

wobei

∠(·) den inversen Tangens-Operator über vier Quadranten bezeichnet,

mT,p der Target-Amplitudenwert eines Pixels, welches ein Volumselement des Objekts repräsentiert, in einem mit der p-ten Empfangsspule erfassten Targetbild (ST,p) ist,

ϑT,p der Target-Phasenwert des genannten Pixels ist,

ϑR,p der Referenz-Phasenwert eines Pixels, welches das genannte Volumselement repräsentiert, in einem mit der p-ten Empfangsspule erfassten Referenzbild (SR,p) ist, und

ϑS der Phasenwert eines Pixels, welches das genannte Volumselement repräsentiert, in dem kombinierten Phasenbild (ΘS) ist.


 
2. Das Verfahren nach Anspruch 1, dadurch gekennzeichnet, dass die genannte erste Echozeit (TER) kleiner als 1 ms ist.
 
3. Das Verfahren nach Anspruch 2, dadurch gekennzeichnet, dass die genannte erste Echozeit (TER) kleiner als 100 µs ist.
 
4. Das Verfahren nach einem der Ansprüche 1 bis 3, dadurch gekennzeichnet, dass das Referenzbild (SR,p) in einer geringeren Pixelauflösung als das Targetbild (ST,p) erfasst wird und vor dem genannten Subtrahieren auf die Pixelauflösung des Targetbildes (ST,p) hochskaliert (9) wird.
 
5. Das Verfahren nach einem der Ansprüche 1 bis 4, gefolgt vom Schritt (12) des pixelweisen Berechnens eines kombinierten Amplitudenbildes (MS) gemäß

wobei

|·| den Betragsoperator bezeichnet,

mT,p der Target-Amplitudenwert eines Pixels, welches ein Volumselement des Objekts repräsentiert, in einem mit der p-ten Empfangsspule erfassten Targetbild (ST,p) ist,

ϑT,p der Target-Phasenwert des genannten Pixels ist,

ϑR,p der Referenz-Phasenwert eines Pixels, welches das genannte Volumselement repräsentiert, in einem mit der p-ten Empfangsspule erfassten Referenzbild (SR,p) ist, und

mS der Amplitudenwert eines Pixels, welches das genannte Volumselement repräsentiert, in dem kombinierten Amplitudenbild (MS) ist.


 
6. Das Verfahren nach einem der Ansprüche 1 bis 4, gefolgt vom Schritt (13) des pixelweisen Berechnens eines kombinierten Amplitudenbildes (MS) gemäß

wobei

|·| den Betragsoperator bezeichnet,

mT,p der Target-Amplitudenwert eines Pixels, welches ein Volumselement des Objekts repräsentiert, in einem mit der p-ten Empfangsspule erfassten Targetbild (ST,p) ist,

ϑT,p der Target-Phasenwert des genannten Pixels ist,

ϑR,p der Referenz-Phasenwert eines Pixels, welches das genannte Volumselement repräsentiert, in einem mit der p-ten Empfangsspule erfassten Referenzbild (SR,p) ist, und

mW der Amplitudenwert eines Pixels, welches das genannte Volumselement repräsentiert, in dem kombinierten Amplitudenbild (MS) ist.


 
7. Das Verfahren nach einem der Ansprüche 1 bis 6, dadurch gekennzeichnet, dass die jeweiligen Referenz- oder Targetbilder (SR,p, ST,p), welche mit der genannten Mehrzahl von Empfangsspulen erfasst werden, alle auf ein und dieselbe Anregung (6) folgend erfasst werden.
 


Revendications

1. Procédé d'imagerie par résonnance magnétique pour dépeindre un objet tridimensionnel par une image ayant des pixels représentant des éléments de volume de l'objet, comprenant :

l'immobilisation de l'objet et l'acquisition (4) d'une image de référence (SR,p) de l'objet avec une bobine réceptrice à un premier temps d'écho (TER) suivant immédiatement une excitation (6) par une bobine de transmission, où ladite image de référence (SR,p) a une valeur complexe, représentant chaque élément de volume par un pixel avec une valeur d'amplitude de référence (mR,p) et une valeur de phase de référence (ϑR,p) ;

la conservation de l'immobilisation de l'objet et l'acquisition (7) d'une image cible (ST,p) de l'objet avec ladite bobine réceptrice à un deuxième temps d'écho (TET) présélectionné, plus long que ledit premier temps d'écho (TER), suivant la même ou une autre excitation (6) par ladite bobine de transmission, où ladite image cible (ST,p) a une valeur complexe, représentant chaque élément de volume par un pixel avec une valeur d'amplitude cible (mT,p) et une valeur de phase cible (ϑT,p) ;

la soustraction (10), pixel par pixel, de la valeur de phase de référence (ϑR,p) de la valeur de phase cible (ϑT,p) afin d'obtenir une valeur de phase corrigée (ϑK,p) pour chaque pixel ; et

l'obtention (10) de ladite image (SK,p) à partir desdites valeurs d'amplitude cibles (mT,p) et desdites valeurs de phase corrigées (ϑk,p)

caractérisé en ce que

les étapes citées précédemment (4, 7, 10) sont appliquées à chaque bobine d'une pluralité de bobines réceptrices agencées autour dudit objet immobilisé pour obtenir une pluralité respective desdites images (SK,p), suivies par l'étape (11) de calcul d'une image de phase combinée (ΘS), pixel par pixel, selon

avec

(·) représentant l'opérateur tangente inverse en quatre quadrants,

mT,p étant la valeur d'amplitude cible d'un pixel, représentant un élément de volume de l'objet, dans une image cible (ST,p) acquise avec la pième bobine réceptrice,

ϑT,p étant la valeur de phase cible dudit pixel,

ϑR,p étant la valeur de phase de référence d'un pixel, représentant ledit élément de volume, dans une image de référence (SR,p) acquise avec la pième bobine réceptrice, et

ϑS étant la valeur de phase d'un pixel, représentant ledit élément de volume, dans l'image de phase combinée (ΘS).


 
2. Procédé selon la revendication 1, caractérisé en ce que ledit premier temps d'écho (TER) est inférieur à 1 ms.
 
3. Procédé selon la revendication 2, caractérisé en ce que ledit premier temps d'écho (TER) est inférieur à 100 µs.
 
4. Procédé selon l'une quelconque des revendications 1 à 3, caractérisé en ce que l'image de référence (SR,p) est acquise à une résolution de pixel inférieure que l'image cible (ST,p) et, avant ladite soustraction, est mise à l'échelle supérieure (9) à la résolution de pixel de l'image cible (ST,p).
 
5. Procédé selon l'une quelconque des revendications 1 à 4, suivi par l'étape (12) de calcul d'une image d'amplitude combinée (MS), pixel par pixel, selon

avec

|·| représentant l'opérateur d'amplitude,

mT,p étant la valeur d'amplitude cible d'un pixel, représentant un élément de volume de l'objet, dans une image cible (ST,p) acquise avec la pième bobine réceptrice,

ϑT,p étant la valeur de phase cible dudit pixel,

ϑR,p étant la valeur de phase de référence d'un pixel, représentant ledit élément de volume, dans une image de référence (SR,p) acquise avec la pième bobine réceptrice, et

mS étant la valeur d'amplitude d'un pixel, représentant ledit élément de volume, dans l'image d'amplitude combinée (MS).


 
6. Procédé selon l'une quelconque des revendications 1 à 4, suivi par l'étape (13) de calcul d'une image d'amplitude combinée (MS), pixel par pixel, selon

avec

|·| représentant l'opérateur d'amplitude,

mT,p étant la valeur d'amplitude cible d'un pixel, représentant ledit élément de volume de l'objet, dans une image cible (ST,p) acquise avec la pième bobine réceptrice,

ϑT,p étant la valeur de phase cible dudit pixel,

ϑR,p étant la valeur de phase de référence d'un pixel, représentant ledit élément de volume, dans une image de référence (SR,p) acquise avec la pième bobine réceptrice, et

mW étant la valeur d'amplitude d'un pixel, représentant ledit élément de volume, dans l'image d'amplitude combinée (MS).


 
7. Procédé selon l'une quelconque des revendications 1 à 6, caractérisé en ce que les images de référence ou cibles (SR,p, ST,p) respectives acquises avec ladite pluralité de bobines réceptrices sont toutes acquises suite à une et même excitation (6).
 




Drawing














Cited references

REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description