Method for automatically determining set of at least one optimal characteristic parameter of acquisition sequence by means of magnetic resonance

Abstract

The invention relates to a method for determining an optimal feature set (Tlopt) of at least one feature parameter of an acquisition sequence for acquiring an image of a region to be imaged by magnetic resonance, the acquisition sequence being configured to substantially cancel medical signals from at least one predetermined tissue, the method comprising steps implemented by a computer, according to the invention, an optimal feature image is selected from a plurality of feature images (TKi) of a feature region of the region to be imaged, said optimal feature image having a large number of pixels with intensities less than or equal to a given intensity threshold (SI), said feature images being generated by means of respective sets of respective acquisition sequences having at least one feature parameter (TIi), the optimal feature set (Tlopt) is a feature set (TIi) by means of which an acquisition sequence of optimal feature images is generated.

Description

Technical Field

[0001] The present invention relates to the field of magnetic resonance imaging (MRI), in particular to cardiac magnetic resonance imaging.

[0002] It relates to magnetic resonance imaging in which the acquisition sequence is configured to cancel the signal of at least one predetermined tissue, such as late gadolinium enhancement or LGE magnetic resonance imaging.

[0003] The reference technique for assessing local scarring and myocardial fibrosis is bright blood late gadolinium enhancement or BR-LGE imaging. In this type of imaging, an inversion recovery pulse is used to cancel the viable myocardial signal, which allows the scar to be visualized with a high contrast between healthy myocardial tissue and the scar. However, for myocardial scars adjacent to the cardiac blood chambers (right and left ventricles), the high intensity signal from the blood hinders clear visualization and precise demarcation of the scar, especially the subendocardial scar. To solve this problem, black blood LGE (BL-LGE) imaging techniques have been proposed. They make it possible to cancel the signals from healthy myocardium and blood simultaneously, providing high contrast between scar and blood and between scar and healthy myocardium.

[0004] Blood signal is eliminated to obtain "black blood" contrast by applying a radiofrequency (RF) inversion-recovery sequence to the patient's area to be imaged, which consists of a 180° pulse followed by a T1-rho preparation module and a readout module. The duration separating the 180° pulse and the readout module is called the inversion time.

[0005] An attempt is made to acquire signals with an acquisition sequence to generate an image of the area to be imaged so that the inversion time of the sequence is optimal so that at the moment of reading the longitudinal magnetization, the longitudinal magnetization of blood and healthy myocardium is canceled, which makes it possible to obtain an MRI image showing the best contrast between the scar and the adjacent blood and healthy myocardium. The choice of inversion time is fundamental because the contrast of the image has a direct impact on the visualization, detection and delimitation of the scar and therefore on the diagnosis and prognosis of the patient.

[0006] However, this inversion time depends on the patient, the static magnetic field and the interval between the acquisition sequence and the contrast agent injection. It is therefore necessary to determine the optimal inversion time for each patient before acquiring a medical image of the myocardium on which a specialist or detection program aims to detect or visualize and localize the myocardial scar. Background Art

[0007] Based on artificial intelligence techniques, methods for automatically determining an optimal inversion time from MRI images acquired by inversion-recovery type acquisition sequences having respective different inversion times in BR-LGE MRI imaging are known.

[0008] It is worth noting that methods based on analyzing spatial and temporal characteristics of different images by means of neural networks are known. A method based on the technique of segmentation by means of a convolutional neural network and then analyzing the pixel intensities in the region of interest obtained by segmentation is also known.

[0009] However, these methods are not straightforward to use in the clinic, as their installation on MRI imaging equipment is complex (specific computing libraries and software must be installed for these methods to work), and are currently still mainly used for research purposes in specialized settings.

[0010] Patent US9104783B2 discloses a system for automatically determining imaging parameters for BL-LGE imaging based on a model linking inversion time to an electrocardiogram (ECG). However, this method is completely dependent on ECG acquisition, the presence of blood flow, does not work in cases where blood is stagnant on the wall, and is limited to 2D applications.

[0011] It is an object of the present invention to limit at least one of the above-mentioned disadvantages. Summary of the invention

[0012] To this end, the present invention relates to a method for determining an optimal feature group of at least one characteristic parameter of a medical signal acquisition sequence, wherein the acquisition sequence is used to generate a medical image of a region to be imaged by magnetic resonance imaging, wherein the acquisition sequence is configured to substantially cancel out medical signals from at least one predetermined tissue, wherein the method comprises, implemented by a computer, selecting an optimal feature image from a plurality of feature images of a feature region of the region to be imaged generated by signals acquired by corresponding acquisition sequences with different corresponding groups of at least one characteristic parameter, wherein the optimal feature image has a large number of pixels whose intensities are less than or equal to a given intensity threshold, and wherein the optimal feature group is a feature group of the signal acquisition sequence that generates the optimal feature image.

[0013] In a specific embodiment, the medical signal acquisition sequence is a black blood late gadolinium enhancement acquisition sequence.

[0014] In certain embodiments, the characteristic region includes only portions of the myocardium and blood.

[0015] In a particular embodiment, the set of characteristics is reversal time.

[0016] In a particular embodiment, the method includes determining an intensity threshold, the step comprising:

[0017] - Calculate the histogram of the feature image of the feature area,

[0018] -Determines the intensity of the global maximum of the histogram.

[0019] In a particular embodiment, the method includes selecting a characteristic image of a characteristic region in an image of the region to be imaged.

[0020] In a specific embodiment, the method comprises the following sequence of steps:

[0021] - Acquire signals through corresponding acquisition sequences,

[0022] -Generate a feature image from the signal.

[0023] Advantageously, when the optimal inversion time is a maximum or minimum inversion time of the acquisition sequence, the sequence of steps and the selection of the optimal characteristic image are repeated with an inversion time different from that of the acquisition sequence.

[0024] The invention also relates to a method for generating a medical image of a region to be imaged, comprising determining an optimal set of at least one characteristic parameter by a method according to the invention. The method also comprises acquiring a medical signal by a medical acquisition sequence having an optimal set of characteristic parameters, and generating a medical image from the medical signal.

[0025] The invention relates to a processing system comprising a processor configured to implement the steps of the method according to the invention.

[0026] The invention also relates to a computer program product comprising instructions causing a computer to carry out the steps of the method according to the invention when the computer executes the program.

[0027] The present invention also relates to a computer-readable recording medium comprising instructions which, when executed by a computer, cause the computer to implement the method according to the present invention.

[0028] Furthermore, the invention relates to a magnetic resonance imaging system comprising a magnetic resonance system comprising a static magnetic field generator, a gradient generator, a radio frequency device and a processing and control system, the processing and control system being configured to implement the calculation steps of the method according to the invention and to control the magnetic resonance system such that the imaging system implements the method according to the invention.

[0029] Beneficial effects:

[0030] The method according to the invention makes it possible to determine the reversal time in a reliable, robust, computationally inexpensive, reproducible and simple manner.

[0031] It is easily integrated into existing MRI equipment without requiring major modifications to existing acquisition sequences.

[0032] Furthermore, this approach does not impose any additional workload on the handler.

[0033] Furthermore, the method is rapid (on the order of tens of milliseconds), which is advantageous because, given the kinetics of gadolinium elimination, the optimal inversion time may evolve during the examination.

[0034] The method according to the invention makes it possible to choose the spacing between inversion times or, more generally, between the spacing between parameter values ​​of different feature groups of different acquisition sequences, which makes it possible to obtain the optimal inversion time or the optimal group with good precision, which ultimately facilitates the correct detection and precise positioning of the features. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Other features and advantages of the present invention will become more apparent after reading the following detailed description with reference to the accompanying drawings, which illustrate:

[0036] Figure 1 is a schematic diagram of an MRI imaging system according to the present invention;

[0037] Figure 2 is a schematic diagram of the BL-LGE MRI acquisition sequence;

[0038] Figure 3 is a schematic diagram of steps of a method according to a first embodiment;

[0039] Figure 4 is a flow chart of the steps of the MRI image acquisition method according to the present invention;

[0040] Figure 5 is a flow chart of the steps of an exemplary embodiment of a global method according to the present invention. DETAILED DESCRIPTION

[0041] The present invention relates to the field of magnetic resonance imaging or MRI.

[0042] More specifically, the present invention relates to MRI imaging, where it is desirable to configure a medical signal acquisition sequence to obtain a substantially zero signal for one or more predetermined tissues. In other words, it is desirable to configure the medical signal acquisition sequence to generate a substantially black image of the predetermined tissues.

[0043] For example, it involves black blood MRI imaging.

[0044] The invention relates particularly to magnetic resonance imaging requiring the injection of a contrast agent such as gadolinium.

[0045] For example, it involves black blood late gadolinium enhancement (BL-LGE) imaging.

[0046] The invention is applicable to cardiac imaging, but is also applicable to angiography.

[0047] MRI System

[0048] Figure 1 A magnetic resonance imaging system according to the invention is schematically shown, comprising a set A of excitation and measurement devices and a processing and control system C.

[0049] In a manner known per se, the set A of excitation and measurement devices comprises an MRI imaging device B comprising a static magnetic field generator 1 comprising a main polarization magnet, a gradient generator 2 and a radiofrequency device 3 .

[0050] The static magnetic field generator 1 comprises a main polarizing magnet intended to generate a substantially uniform polarized static magnetic field in a polarizing region (usually a tunnel) intended to include a region of a patient to be imaged, such as the heart.

[0051] The gradient generator 2 comprises three gradient coils or solenoids arranged and configured to vary the magnetic field strength in the polarization zone along respective orthogonal axes fixed relative to the polarization zone, the orthogonal axes usually being referenced as x, y and z. The choice of the strengths circulating in these coils makes it possible to choose from several possible thicknesses and cross-sectional planes in which the magnetization of the region to be imaged of the patient received in the polarization zone will be measured.

[0052] The radio frequency device 3 comprises a coil or a solenoid and is capable of generating an MRI acquisition sequence comprising a 180° inversion pulse, a module for magnetization preparation of the region to be imaged and a module for reading signals from the region to be imaged.

[0053] Typically, an acquisition sequence consists of subsequences called modules, each of which includes at least one radio frequency pulse of predetermined and adjustable frequency, shape, duration, phase, amplitude.

[0054] The preparation module is configured to excite, ie change, the magnetization direction of tissue in the area to be imaged.

[0055] The reading module is configured to measure the magnetization of the area to be imaged generated by the preparation module.

[0056] The kit A advantageously comprises an electrocardiograph 4 for acquiring an electrocardiogram of the patient.

[0057] The processing and control system C includes a processing and control module 10, which is configured to generate controls for the MRI device B, in particular for the RF device 3, to generate an acquisition sequence and to generate a two-dimensional grayscale image of the area to be imaged from the signals measured by the RF magnetic field generator by reconstruction techniques known to those skilled in the art.

[0058] The resulting image is a grayscale matrix image. It consists of a set of pixels, each characterized by an intensity I that can take a set of N values ​​(N is an integer greater than 1) corresponding to N gray levels ranging from 0 to N-1. For example, this value can take 256 values ​​between 0 and 255, but N is not limited to 256.

[0059] The representation of the image generated by the processing and control module 10 is intended to be displayed on a screen of a human-machine interface 11 of the processing and control system C.

[0060] Acquisition sequence

[0061] Figure 2 An example of a black blood MRI acquisition sequence for myocardial images is shown. Figure 2 The upper part of shows a schematic diagram of an electrocardiogram (ECG) that the electrocardiograph 4 can acquire during a heartbeat and an acquisition sequence applied to the area to be imaged. Figure 2 The lower part of shows the variation of the longitudinal magnetization of the tissue of the area to be imaged as a function of time t, and the image I reconstructed from the signals acquired during the reading module starting at the time t=te corresponding to the elimination of the longitudinal magnetization of blood and healthy myocardium.

[0062] For BL-LGE MRI acquisitions, a gadolinium-based contrast product is advantageously injected intravenously into the patient 10 to 15 minutes before the acquisition sequence is applied in order to obtain images showing maximum contrast between scar and healthy tissue and blood. At the heart, the contrast agent is rapidly eliminated from healthy myocardium but accumulates in myocardial scar for a long time. The effect of gadolinium is to shorten the relaxation time T1 of the tissue in which it accumulates. As a result, the magnetization of the scar relaxes faster than that of blood and healthy myocardium after the pulse.

[0063] The goal is to generate an image with black blood contrast. In this type of image, the intensity of the pixel corresponding to blood is zero (black pixel) or substantially zero because it was acquired when the longitudinal magnetization of the blood was zero.

[0064] To generate such an image, the RF device 3 generates an inversion recovery acquisition sequence.

[0065] The acquisition sequence includes a preparation module, which includes Figure 2 The inversion pulse marked as 180° in the figure flips the longitudinal magnetization of the tissue in the imaging area in the opposite direction, that is, reverses the longitudinal magnetization. Figure 2 It is shown that under the influence of the inversion pulse, the magnetization of the area to be imaged changes from Mz to -Mz. Due to longitudinal relaxation, the longitudinal magnetization of the different tissues present in the area to be imaged increases and thus returns to its initial value, passing through the value zero. Of course, the relaxation dynamics of different tissues are different.

[0066] In a manner known per se, the preparation modules include, for example, a T1-rho adiabatic module (T1ρ) or a T2-weighted module of duration TSL (acronym for “Time of Spin Lock”) or a module of the MTC (acronym for “Magnetization Transfer Contrast”) type.

[0067] The preparation module (180°, T1ρ) is configured so that the longitudinal magnetization of the blood and the longitudinal magnetization of the healthy myocardium are canceled out at the same time te.

[0068] At the same time te, the longitudinal magnetization of the scar is significantly greater than zero. By acquiring the signal from the area to be imaged at this time te, an image with very high contrast between black pixels corresponding to blood and healthy myocardium and usually white pixels corresponding to the scar is obtained.

[0069] The inversion recovery sequence next comprises a read module LE comprising a 90° pulse and a read gradient applied at time te to read the transverse magnetization of the region to be imaged.

[0070] The inversion time TI is the time separating the 180° pulse and the readout module, i.e. the first instant of image acquisition in the Fourier domain. In order to obtain the best contrast between myocardial scar, blood and healthy myocardium, an attempt is made to start the readout sequence at the instant te when the longitudinal magnetization of blood and myocardium cancels out, in order to generate an image exhibiting the best contrast. The inversion time is optimal, such as Figure 2 TI opt shown.

[0071] Determine the optimal inversion time TI opt

[0072] Figure 3 The steps of the method according to the invention are shown.

[0073] The method according to the invention is a computer-implemented method for obtaining a plurality of feature images IK of feature regions of a region to be imaged. i Determine the optimal inversion time TI opt , the characteristic images are obtained by applying only at their respective inversion times TI i generated by different acquisition sequences.

[0074] from Figure 3 It can be seen that the feature image IK i With different contrasts (i = 1 to K, K is at least equal to 2, in Figure 3 In the non-limiting example, K is equal to 11).

[0075] The optimal inversion time TI opt is advantageously stored in a memory of system C and is used by the processing and control module 10 to generate a control for system A (in particular the RF device 3) so that it acquires so-called medical signals by a so-called medical acquisition sequence, wherein the inversion time TI is the optimal inversion time TI opt , so that processing and control equipment can generate images from these signals.

[0076] According to the present invention, Figure 3 As shown, it is used to automatically determine the optimal inversion time TI opt The method comprises implementing a plurality of feature images IK from i=1 to K by a computer i Select the best feature image IK opt , the best feature image has a larger number of pixels whose intensity is less than or equal to the given intensity threshold SI. For all images IK from i = 1 to K i , the intensity threshold SI used is the same.

[0077] Optimal inversion time TI opt is the inversion time TI of the acquisition sequence i , the optimal basic image IK is generated through this acquisition sequence opt .

[0078] We will now describe in more detail Figure 4 The best basic image IK shown in opt In a preferred embodiment of the selection step, the optimal basic image IK opt is the base image IK i The basic image IK has a large number of pixels whose intensity is less than or equal to the intensity threshold SI. i .

[0079] This step advantageously comprises the following steps, each of which is implemented by a computer through the processing and control module 10:

[0080] - For i=1 to K, calculate 21 the feature image IK i Histogram HIST i ,

[0081] - Determine 22 histograms HIST by conventional methods for detecting global maxima i The global maximum value M of

[0082] - The intensity SI of the global maximum M is determined 23.

[0083] The intensity threshold SI is the histogram HIST i The intensity of the global maximum M of .

[0084] For each of the N possible values ​​of intensity I, each histogram HIST i The number of pixels in the image that have this intensity I.

[0085] For the sake of clarity, Figure 3 The histogram in is local. Figure 3 Only the parts of these histograms relating to intensity values ​​(grayscale) between 0 and 60 are shown.

[0086] The method then comprises the following steps:

[0087] -For each feature image IK i , calculating 24 an image IK having an intensity less than or equal to an intensity threshold SI i The number of pixels NI i ,

[0088] - Image IK in i=1 to K i , determine 25 the maximum number of pixels NI with intensity less than or equal to SI i The best feature image IKopt.

[0089] Optimal inversion time TI opt is the inversion time of the acquisition sequence by which the signal used to generate the optimal characteristic image IKopt has the maximum number of pixels whose intensity is less than or equal to the intensity threshold SI.

[0090] In other words, the proposed method makes it possible to select the image with the most low-intensity pixels.

[0091] From the feature image IK i The fact that an intensity threshold SI is calculated makes it possible to adjust the threshold according to the patient and to select a reliable optimal inversion time.

[0092] Furthermore, it should be noted that in images comprising only myocardium and blood, usually only the scar will present high intensity on images obtained from an acquisition sequence with TIopt. It may happen that the images present high intensity artifacts. The intensity threshold SI used in the method according to the invention is not affected by these pixel values, which are different in the case where the threshold is equal to the average of the histogram maxima.

[0093] Optionally, the intensity threshold SI constitutes data used by the method according to the present invention. The intensity threshold SI is pre-stored in the memory of the system C, for example. Then, the method does not have steps 22 and 23.

[0094] Feature Image

[0095] Advantageously, it is desirable to configure the acquisition sequence to simultaneously acquire a substantially zero signal of at least one tissue of the characteristic region. In other words, at the moment of the read sequence, the relevant tissue(s) simultaneously exhibits a substantially zero longitudinal magnetization, i.e., the image of this / these tissues reconstructed from the signal from this / these tissues measured during the acquisition sequence is black or substantially black.

[0096] Advantageously, the characteristic region comprises the tissue(s).

[0097] When applied to BL-LGE cardiac imaging, the method according to the invention advantageously uses a characteristic image IK of a characteristic region including only a portion of the myocardium and blood. i .

[0098] The method according to the invention is very reliable. In fact, in this case, the acquisition sequence is configured to cancel the longitudinal magnetization of healthy myocardium and blood substantially simultaneously. The region of interest is the part of the myocardium affected by the scar. This region is in the minority and the image will be mainly black.

[0099] In angiography, we wish to visualize blood vessels and blood. The acquisition sequence is configured to substantially cancel the signal from the myocardium and other tissues such as back and chest muscles, as opposed to the signal from the blood. Advantageously, a contrast agent, such as gadolinium, is injected intravenously and the acquisition sequence is implemented in a non-delayed manner so that the blood appears white (bright) on the image.

[0100] Advantageously, the characteristic region includes only tissue for which it is desired to obtain a substantially zero signal through the acquisition sequence.

[0101] Advantageously, the characteristic region includes only tissue for which it is desired to acquire substantially zero signal through the acquisition sequence, as well as blood and most tissue for which it is desired to acquire substantially zero signal.

[0102] Determination of feature groups

[0103] The present invention more generally relates to a method for automatically determining an optimal feature set. The feature set comprises one or more parameters characterizing an MRI acquisition sequence. The object of the present invention is to determine an optimal feature set comprising corresponding values ​​of the feature parameters such that the contrast of an image acquired by the acquisition sequence characterized by the feature set is maximized.

[0104] Steps and references of the implemented method Figure 4 The steps described are the same.

[0105] The method then uses feature images of feature regions acquired by corresponding acquisition sequences which differ only in their feature sets, ie the parameter values ​​characterizing the sets.

[0106] The feature group can in particular include at least one parameter selected from: inversion time TI, a preparation module, parameters of the preparation module, such as the TSL duration of the module in T1ρ, the number of pulses of the preparation module, the time interval between two consecutive pulses of the preparation module, the intensity, phase shift, frequency, amplitude of each pulse, the shape, duration, phase of at least one or each of the pulses of the preparation module.

[0107] Each characteristic parameter can take different values. In other words, each characteristic parameter can be adjusted by the MRI system A, C.

[0108] The choice of the value of each characteristic parameter has an impact on the evolution of the longitudinal magnetization value of each tissue present in the image as a function of time and therefore also of the image contrast.

[0109] These parameters may or may not be independent of each other.

[0110] The optimal feature set is the inversion time of the signal acquisition sequence by which an image IK is generated with the maximum number of pixels whose intensity is less than or equal to the intensity threshold SI opt .

[0111] Global Methods

[0112] The invention also relates to a global method for generating a medical image of a region to be imaged, the steps of which are: Figure 5 Shown in.

[0113] The global approach includes methods for determining the best feature set, including, for example:

[0114] - a contrast product (e.g. gadolinium) may be injected intravenously111 into the patient,

[0115] - generated 112 by the processing and control module 10 to acquire signals of the area to be imaged by means of K respective acquisition sequences, said acquisition sequences differing only in their characteristic set of at least one characteristic parameter,

[0116] - the RF device 3 acquires 113 signals of the area to be imaged using a Ks acquisition sequence,

[0117] - generating 114 by the processing and control module 10 from the signals acquired during step 112 K two-dimensional global images IM of the area to be imaged i (i=1 to K),

[0118] - selected 115 by the processing and control module 10 as the respective global image IM i The K feature images IK of a part i (i=1 to K),

[0119] - by means of the previously described method for obtaining the characteristic image IK by means of the processing and control module 10 i Method for determining the best group, determining 116 the best group, such as the best inversion time TI opt .

[0120] The global method then includes the following steps:

[0121] - generating 120 an acquisition control by the processing and control module 10 for acquiring so-called medical signals to generate a set of at least one medical image of the area to be imaged from at least one acquisition sequence characterized by an optimal set of characteristics,

[0122] - acquiring 130 medical signals of the area to be imaged by the RF device 3 through acquisition sequence(s) characterized by the optimal set of characteristics,

[0123] - generating 140 a set of at least one medical image of the area to be imaged according to the medical signals of the area to be imaged acquired in step 130 by the processing and control module 10,

[0124] - Displaying 150 a representation of the medical image via the human-machine interface 11 .

[0125] Advantageously, during step 130, signals are acquired, which makes it possible to generate several medical images according to different cross-sectional planes during step 140. To this end, a gradient generator generates different gradients during the acquisition sequence.

[0126] The invention also relates to a system A, C configured to implement the method according to the invention.

[0127] Advantageously, in the non-limiting case of BL-LGE imaging, the acquisition step 113 is carried out between 10 and 15 minutes after the injection of gadolinium. It is triggered and controlled by the control generated during step 112.

[0128] In step 113, signals of the region to be imaged are acquired by respective acquisition sequences, which differ from each other only in their respective feature sets. In addition, other acquisition conditions, namely the magnetic field generated by the static magnetic field generator 1 and the gradient generated by the gradient generator, are the same during these sequences.

[0129] The acquisition sequence is advantageously synchronized with the heart rhythm, using the signals measured by the electrocardiograph 4, by the processing and control module 10, so as to be carried out during corresponding consecutive heart beats. The acquisition sequence is performed at each heart beat.

[0130] Therefore, the global image IM i It is a cardiac image generated based on the same cross-sectional plane of the area to be imaged.

[0131] In cardiac imaging, the section plane is, for example, the small axis section plane. This section plane has the advantage of allowing visualization of the entire myocardium in cross section, thus facilitating the detection and visualization of myocardial scars.

[0132] Other cross-sectional planes may be advantageous for other types of detection, such as for functional imaging.

[0133] Advantageously, in the case of BL-LGE cardiac imaging, when the static magnetic field is 1.5 Tesla, the acquisition sequence comprises a corresponding inversion time ranging from 60 ms to 160 ms. In fact, the applicant has found through tests performed on different patients that for this type of imaging, the inversion time is typically between 60 ms and 160 ms.

[0134] Typically, the corresponding reversal time may be between 0 seconds and 300 ms.

[0135] For example, the acquisition sequence includes different inversion times ranging from 60 ms to 160 ms and separated by a spacing interval.

[0136] For example, the interval is equal to 10 ms.

[0137] Optionally, the spacing is less than or greater than 10 ms. The choice of spacing depends on the compromise between acquisition time (the smaller the spacing, the greater the number of images) and the accuracy with which the optimal inversion time is desired to be defined, knowing that below a certain spacing the gain in accuracy of the optimal TI and the contrast of the medical image are imperceptible to the naked eye for medical images used for diagnosis.

[0138] The spacing can be fixed or variable within the interval.

[0139] The step 114 of generating a two-dimensional global image is performed by MRI image reconstruction techniques known to those skilled in the art. It is worth noting that parallel reconstruction algorithms such as GRAPPA or the SENSE algorithm (acronym for “SENSitivity Encoding”) are known.

[0140] Selecting the Feature Image IK i , that is, image IM i Part of step 115 may be performed in different ways.

[0141] In a particular embodiment, the selection step 114 implemented by a computer (e.g., by system C) includes two steps:

[0142] -Select image IM on each image IMi i Image part of PO i , the image part PO i exist Figure 3and corresponds to the image IM delimited by the white box represented on the image IMi i Part of PO i ,

[0143] -For each image part PO of the image i , select as the image part PO i The feature image IK of a part i The feature image IK i exist Figure 3 The third row is represented by and corresponds to the image portion PO i The image PO is delimited by the white box shown above i part.

[0144] Image part PO i The options include, for example, obtaining by the processing and control module 10 a completely surrounding and delimiting image portion PO i The coordinates, size, and orientation of the first box of . This box is Figure 3 Medium Image i A white box is displayed on the

[0145] In cardiac imaging, the frame is advantageously centered on the heart, delimits a heart section in the cross-sectional plane, and contains the entire heart section located in the cross-sectional plane. Advantageously, the frame is a polygon, such as a rectangle, each side of which delimits a segment of the heart in the cross-sectional plane.

[0146] The frame is displayed on the screen of the HMI 11 by the operator, for example, using an element of the HMI (for example, a mouse of the HMI). i The movement of the element results in the movement of the cursor on the HMI screen.

[0147] For example, the box is a shim box, which is usually drawn by the user on one of the images to select the portion of the field in which the magnetic field must be homogenized. i .

[0148] The size and orientation coordinates of the frame in the image mark are calculated by the processing and control module 10 according to the position, orientation and size of the frame on the screen and stored in the memory of the system C. The processing module 10 selects the image IM according to the coordinates, size and orientation of the frame. i Part of PO i . Partial PO i Stored in the memory of system C.

[0149] Advantageously, the characteristic image IK i The choice is to select the image part PO i The central part is equal to the image part POi The size of is divided by a predetermined factor.

[0150] In cardiac imaging, this factor is advantageously defined so that the characteristic image IK i Includes only blood and part of the heart muscle.

[0151] For example, the factor is comprised between 2 and 3. For example, it is equal to 2.5. The applicant tested different values ​​of the factor on different patients to obtain the factor.

[0152] Optionally, K feature images IK i The selection 115 of (i=1 to K) comprises a segmentation step known to those skilled in the art implemented by the processing and control module 10 in order to select an image portion PO which completely surrounds and delimits a heart section in the section plane. i . Such a segmentation step is described, for example, in the following paper: “Automated deep learning-based inversion time selection for cardiac late gadolinium enhancement imaging, ISMRM 2020, Seung Su Yoon, Michaela Schmidt, Bernd J Wintersperger, Teodora Chitiboi, Puneet Sharma, Christoph Tillmanns, Andreas Maier, and Jens Wetzl”, where in the case of BL-LGE imaging the algorithm must be trained on images acquired by a BL-LGE acquisition sequence, and in the following paper “Dark blood ischemic LGE segmentation using a deep learning approach”, C. Torlasco et al., European Heart Journal - Cardiovascular Imaging, Volume 22, Supplement_2, June 2021.

[0153] Optionally, selection 115 includes in the image IM i The center position of the heart section is acquired in the cross-sectional plane of one of the images, and the processing and control module 10 determines the image IM i A portion of the image located at the center of the image and a predetermined size stored in the memory of system C.

[0154] The selection step 115 causes the image IM to be selected in the cardiac imaging. i It is possible to include only a portion of the blood and myocardium in the image. It does not require the implementation of an accurate heart segmentation algorithm.

[0155] Optionally, the feature image IK i is the global image IM i Or image part PO i .

[0156] As previously described, the determination of the optimal group or more specifically the optimal inversion time TI is implemented by the processing and control module 10opt Step 116.

[0157] Advantageously, when the optimal inversion time TI opt It is used to generate feature image IK i The sequence of steps 112 to 116 of the method is repeated with an inversion sequence having respective acquisition times, the minimum inversion time of which is used to generate the image IK. i The maximum inversion time of the acquisition sequence, or the maximum inversion time is used to generate the image IK i This makes it possible to find the truly optimal inversion time.

[0158] Hardware

[0159] From a hardware perspective, Figure 1 The shown processing and control system C may be seen as a computer interacting with a computer program product.

[0160] System C is a computer, such as a microcomputer, a computer network, an electronic component, a tablet computer, a smartphone, or a personal digital assistant (PDA).

[0161] System C comprises a processing and control module 10. The processing and control module comprises, for example, one or more processors capable of interpreting instructions in the form of a computer program, a programmable logic circuit, such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a programmable logic device (PLD) and a programmable logic array (PLA), a system on a chip (SOC), a circuit board in which the steps of the method according to the invention are implemented in hardware elements. The steps of the method according to the invention can be performed by a processor in a simultaneous or sequential manner and / or by one or more processors.

[0162] The processing and control module 10 includes data processing circuitry for performing calculations, a memory operatively coupled to the data processing circuitry, a computer readable medium, and optionally a reader adapted to read the computer readable medium.

[0163] System C also includes an IMH 11, which includes input devices, output devices, and communication devices.

[0164] Each function of the system C is executed by causing the data processing circuit to read a predetermined program on hardware such as the memory of the processing and control module 10, so that the data processing circuit performs calculations, controls communications performed by the communication device, and reads data from and / or writes data to the memory of the computing module and the computer-readable medium.

[0165] The method is performed on a single computer or on a system distributed among several computers (particularly by using cloud computing).

[0166] The memory is a computer readable recording medium and may be configured with, for example, at least one of the following elements: ROM (read-only memory), EPROM (erasable programmable read-only memory), EEPROM (electrically erasable programmable read-only memory), RAM (random access memory) and other suitable storage media. The memory may include an operating system and load the program according to the present invention. It includes registers suitable for recording parameter variables created and modified during the execution of the above program.

[0167] The program product may include a computer-readable recording medium, which is a tangible device, which itself is not a transient signal, and which may be configured with, for example, at least one of the following elements: a removable medium, such as but not limited to a magneto-optical disk (e.g., a compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a removable disk, a hard disk drive, a smart card, a flash memory device (e.g., a card, a key), a magnetic stripe, a database, a server, and another suitable storage medium.

[0168] Alternatively, the program instructions are obtained from an external source and downloaded via a network. This is particularly true for application programs. In this case, the computer program product includes a computer-readable data medium on which the program instructions are stored or a data medium signal on which the program instructions are encoded.

[0169] The invention relates to a computer program product comprising a computer-readable medium containing instructions which, when executed by a data processing circuit, cause a system C to implement the steps of the method according to the invention.

[0170] The form of the program instructions is, for example, source code form, computer executable form, or any form between source code and computer executable form, such as a form obtained by converting the source code by an interpreter, assembler, compiler, link editor, or locator. Alternatively, the program instructions are microcodes, firmware instructions, state definition data, configuration data of an integrated circuit (such as VHDL), or object code. The program instructions are written in any combination of one or more programming languages, such as object-oriented programming languages ​​(C++, JAVA, Python), procedural programming languages ​​(such as language C).

[0171] The user interface or HMI 11 comprises input devices and output devices. The user interface 11 comprises input devices for allowing a user to input data, or control e.g. a shim box, in order to be able to interact with the program according to the invention. The input devices comprise e.g. a keyboard or a pointing interface, such as a mouse, a light pen, a touch pad, a remote control, a voice recognition device, a tactile device.

[0172] Output devices are designed to present information to a user in a sensory or electrical manner, such as visually or auditorily. Output interfaces include, for example, graphical interfaces. For example, in the case of a touch tablet computer, an output interface may be an input device.

[0173] The display step is implemented by the output device.

[0174] The set of at least one communication device allows communication between elements of system C and optionally between at least one element of the system and a device external to system C (e.g., system A). The communication device may establish physical links between elements of system C and / or between elements of system C and a device external to system C, and / or establish long-range (wireless) communication links between elements of system C and / or between elements of system C and a device external to system C.

Claims

1. An optimal feature set (TI) for determining at least one feature parameter (TI) of a medical signal acquisition sequence opt ), wherein the medical signal acquisition sequence is used to generate a medical image of a region to be imaged by magnetic resonance, the acquisition sequence is configured to substantially cancel out medical signals from at least one predetermined tissue, the method comprising: implementing by a computer, in a plurality of characteristic images (IK) of a characteristic region of the region to be imaged generated according to signals acquired by a plurality of acquisition sequences i ), select the best feature image (IK opt ),in, Each of the plurality of acquisition sequences has at least one characteristic parameter (TI i ), the optimal feature image having a large number of pixels whose intensity is less than or equal to a given intensity threshold (SI), the optimal feature group (TI opt ) is the best feature image (IK opt ) of the signal acquisition sequence (TI i ).

2. The method according to claim 1, wherein: The medical signal acquisition sequence is a black blood late gadolinium enhancement acquisition sequence.

3. The method according to claim 2, wherein: The characteristic region only includes part of the myocardium and blood.

4. A method according to any one of the preceding claims, wherein: The feature set is reversal time.

5. The method according to any of the preceding claims, comprising determining the intensity threshold (SI), the determining the intensity threshold (SI) comprising: Calculate (21) the feature image (IK) of the feature area i ) histogram (HIST i ), Determine (22) the histogram (HIST i )’s global maximum (M).

6. A method according to any one of the preceding claims, comprising: i ) in the image to select the feature image (IK) of the feature area i ).

7. A method according to any one of the preceding claims, comprising the following sequence of steps: Acquire signals through corresponding acquisition sequences, Generate a characteristic image (IK) according to the signal i ).

8. The method according to claim 7, wherein: When the optimal inversion time is the maximum inversion time or the minimum inversion time of the acquisition sequence, the sequence of steps and the selection of the optimal characteristic image are repeated with an inversion time different from the inversion time of the acquisition sequence.

9. A method for generating a medical image of an area to be imaged, comprising determining an optimal group of at least one characteristic parameter by a method according to any one of claims 1 to 8, acquiring a medical signal by a medical acquisition sequence having an optimal characteristic group, and generating the medical image based on the medical signal.

10. A processing system (C), comprising a processor configured to implement the steps of the method according to any one of claims 1 to 6.

11. A computer program product comprising instructions which, when executed by a computer, cause the computer to carry out the steps of the method according to any one of claims 1 to 6. 12 . A computer-readable recording medium comprising instructions, which, when executed by a computer, cause the computer to implement the method according to claim 1 .

13. A magnetic resonance imaging system, comprising a magnetic resonance system (A), the magnetic resonance system comprising a static magnetic field generator (1), a gradient generator (2), a radio frequency device (3) and a processing and control system (C), the processing and control system being configured to implement the steps of the method according to any one of claims 1 and 6, and to control the magnetic resonance system (A) so that the imaging system implements the method according to any one of claims 7 to 9.

Images