WO2025026987A1 - Mri method for quantifying the concentration of a contrast agent and associated device - Google Patents

Abstract

The invention relates to a method for measuring a concentration of a contrast agent in at least one body area of a living human or animal individual, the method comprising: - a) acquiring radiofrequency signals emitted by the organ, subjected to fast spin echo magnetic resonance imaging; - b) obtaining images of the body area from the radiofrequency signals resulting from a); and - c) estimating a relaxation time T1 in different portions of the body area from the images resulting from b).

Description

MRI METHOD FOR QUANTIFYING THE CONCENTRATION OF A CONTRAST AGENT AND ASSOCIATED DEVICE

Description

TECHNICAL AREA

The technical field of the invention is the quantification of a contrast agent in a human or animal body.

PRIOR ART

Radiotherapy is a treatment commonly applied to cancer. Radiotherapy may be one of the only treatments applicable in certain types of cancer, particularly cancers that are difficult to operate and/or take the form of diffuse tumors. This is particularly the case for certain brain tumors.

Some nanoparticles have the ability to increase, by radiobiological effect, the damage induced to a tumor during exposure to ionizing electromagnetic radiation of type X or gamma. This is called a radiosensitizing effect or an increase in dose deposition leading to harmful biological effects on tumor cells. These nanoparticles can also have the properties of contrast agents for MRI (Magnetic Resonance Imaging). They are then called "theranostic" agents, due to their combined application in the field of therapy and diagnosis by MRI imaging.

Nanoparticles, referred to as AGulX (Activation and Guidance of Irradiation by X-ray), concentrate a large number of Gd ³⁺ ions in nanoparticles with a diameter of less than 10 nm. These particles consist of a polymer core, on the surface of which are chelated metal ions, the atomic number of which is high. It has been shown that this type of nanoparticle makes it possible to increase radiation-induced damage in brain tumors, in particular by radiobiological effect, while providing protection with respect to surrounding healthy tissues, due to the high atomic number. Such particles are for example described in WO2022/106788, WO2011135101, WO2018224684 or WO2019008040. They are currently administered intravenously to patients prior to radiotherapy as part of their treatment pathway.

où In order to guide radiotherapy, it is necessary to know the concentration in the tumor tissues to be treated of the previously administered theranostic contrast agent. For this purpose, MRI can be implemented, before and after administration. of the contrast agent, so as to obtain a spatial distribution of the longitudinal relaxation time Ti. The concentration of the contrast agent is estimated according to the expression:

Or

Ti, pre ^and Ti,p _0S t are ^the times 7 respectively before and after the administration of the contrast agent (unit: s); is a longitudinal relaxivity constant of the contrast agent (unit: mW ¹ s" 1), and whose value is assumed to be known

C is the concentration of the contrast agent (unit: mM)

However, most MRI modalities, allowing to obtain a mapping of the relaxation time Tl, implement inversion recovery or saturation recovery sequences, whose acquisition times are too long for these sequences to be applied, in vivo, in clinical routine.

Among other modalities, the VFA (Variable Flip Angle) sequence allows to obtain a 3D mapping of the relaxation time Tl. However, the VFA sequence requires multiple successive spoiled gradient echo acquisitions (group of fast gradient echo sequences including a phase-shifting gradient called spoiler), by varying the excitation angle. In addition, the VFA sequence is known to be sensitive to magnetic field inhomogeneities, which generates measurement errors.

Document US20130200900 describes for example the establishment of a 3D mapping of the relaxation time Tl in a breast, by implementing a VFA sequence. In this document, a phantom, comprising different concentrations of Gd, is placed around the breast.

Another disadvantage of the VFA sequence is a too long acquisition time, which results in an occupancy time of the imaging device too long for routine use.

Clinical MRI imaging protocols for patients with brain metastases are primarily intended to obtain good anatomical representation, and in particular good discrimination between metastases and healthy tissues, particularly in the context of diagnosis and possibly pre-therapeutic assessment. The MPRAGE (Magnetization-Prepared Rapid Gradient Echo) acquisition sequence is a potentially usable gradient echo sequence. However, the MPRAGE sequence can induce confusion between a brain metastasis and a blood vessel, which can generate false positives.

The publication Wright PJ. "Water proton Tl measurements in brain tissue at 7, 3, and 1.5T using IR-EPI, IR-TSE, and MPRAGE: results and optimization" describes a comparison of the Tl relaxation time obtained by different MRI modalities applied either to phantoms, of known composition, or to brain structures. The modalities compared are TSE and MPRAGE, with the EPI (Echo Planar Imaging) modality being considered as a reference modality.

Other commonly used sequences for MRI of brain metastases include the Tl-weighted 3D TSE (Turbo Spin Echo) sequence, which may be referred to as SPACE, CUBE, VISTA, or FASE3D mVox by leading MRI manufacturers. The Tl-weighted 3D TSE sequence is known to generate an accurate anatomical representation. In addition, unlike the MPRAGE sequence, the TSE sequence provides good discrimination between metastases and blood vessels. It is therefore particularly suitable for detecting metastases.

However, for a Tl-weighted 3D TSE sequence, there is no analytical model to obtain the Tl value from the intensity of an image of an anatomical section.

The inventors propose a method for establishing the concentration of a contrast agent by implementing a routine MRI sequence, of the fast spin echo type. The concentration is obtained by measuring the relaxation time Tl from the intensity of an image of a section. This involves going beyond the simple comparison of sequences.

PRESENTATION OF THE INVENTION

A first object of the invention is a method for measuring a concentration of a contrast agent in at least one body area of a living human or animal individual, the method comprising:

- a) acquisition of radiofrequency signals emitted by the organ, subjected to a magnetic resonance imaging modality;

- b) from the radiofrequency signals resulting from a), obtaining images of the body area; - c) from the images resulting from b), estimation of a relaxation time Tl in different parts of the body area; steps a) to b) being implemented before and after an administration of a contrast agent to the individual, so as to obtain an estimation of relaxation times Tl, in the body area, respectively before and after the administration of the contrast agent;

- d) from the relaxation times Tl estimated respectively before and after the administration of the contrast agent, determination of a concentration of the contrast agent of the different parts of the body area; the method being characterized in that

- in each step a), the magnetic resonance imaging modality is a Tl-weighted fast spin echo modality, for example Tl-weighted 3D TSE;

- each step c) is implemented using a calibration function, establishing a correspondence between the intensity of an image of the body area and the relaxation time Tl, the calibration function being obtained, prior to step c), during a calibration phase comprising:

• i) arrangement of a phantom, comprising different compartments respectively comprising different compositions whose relaxation times Tl are known and different from each other;

• ii) acquisition of radiofrequency signals emitted by the phantom, subjected to the magnetic resonance imaging modality;

• iii) from the radiofrequency signals acquired during ii), obtaining images of the different compartments of the phantom;

• iv) establishing the calibration function from the intensity of the signal in said compartments of the phantom by taking into account, for each compartment, a pair of values formed by the intensity of the signal of an image of the compartment and the relaxation time Tl of the composition of said compartment.

According to one possibility:

- in step a), the body area is placed in a tunnel, subjected to a magnetic field;

- during step i) of the calibration phase, the phantom is placed in the tunnel, in place of the body area. According to one possibility, different compartments of the phantom respectively comprise different chemical species or different concentrations of the same chemical species, so that the relaxation times Tl of said compartments are different from each other.

According to one possibility, different compartments of the phantom have the same concentration of the same chemical species.

According to one possibility, the calibration function is established by a regression, from the pairs (signal intensity of an image of the compartment - relaxation time Tl of the compartment) taken into account during step iv) of the calibration phase.

According to one possibility:

- steps a) to d) are implemented successively for different individuals;

- the calibration function implemented in each step c) is identical for the different individuals, and results from the same calibration phase.

The contrast agent may be a radiosensitizing agent. The method may comprise, following step d), a determination of an irradiation dose in different parts of the body area, as a function of the concentrations of the contrast agent resulting from step d).

Another object of the invention is a processing unit, configured to process an image acquired by a magnetic resonance imaging device, according to a fast spin echo acquisition modality, for example of the 3D TSE type, weighted in Tl, the processing unit being programmed to apply, to the acquired image, a calibration function resulting from an implementation of a calibration phase comprising:

• arrangement of a phantom in the magnetic resonance imaging device, the phantom comprising different compartments respectively comprising different compositions whose relaxation times Tl are known and different from each other;

• acquisition of radiofrequency signals emitted by the phantom, subjected to the fast spin echo type magnetic resonance imaging modality;

• from the radiofrequency signals acquired during the previous step, obtaining images of the different compartments of the phantom;

• establishing the calibration function from the intensity of the signal in said compartments of the phantom, taking into account, for each compartment, a pair of values formed by the signal intensity of an image of the compartment and the relaxation time Tl of the composition of said compartment.

FIGURES

Figure 1 shows a schematic of an MRI imaging device, allowing implementation of the invention.

Figure 2A shows a stack of plates forming a calibration phantom.

Figure 2B shows the plates arranged in an envelope, the envelope allowing manipulation of the phantom.

Figure 2C shows an example of a plate configuration.

Figure 2D shows an image of the compartments embedded in a plaque obtained by an MRI modality using a Tl-weighted 3D TSE sequence.

Figure 3 shows the establishment of a bijective calibration function to relate the relaxation time Tl to the intensity on an image resulting from a fast spin echo sequence, for example Tl-weighted 3D TSE.

Figure 4 shows schematically the main steps of implementing a method according to the invention.

PRESENTATION OF SPECIFIC EMBODIMENTS

Figure 1 shows a device 1 allowing an implementation of the invention. The device is an MRI imaging device, comprising a cylindrical magnet extending around a tunnel 2. The tunnel extends around a central axis A. The device is intended to produce images of sections of biological tissues, in particular of the body areas of a living human or animal individual.

In a known manner, under the effect of an intense magnetic field, produced by the magnet, and of radiofrequency signals emitted by transmitters, at the Larmor frequency of the hydrogen nucleus, the hydrogen nuclei of the body area emit a radiofrequency signal, the latter being measured by receiving antennas of the device 1. Depending on the measured radiofrequency signals, the device 1 forms anatomical sections of the body area examined. The sections are two-dimensional images, in which the intensity depends on the composition of the body area examined. The device 1 comprises a central unit 3, programmed to generate the images from detected radiofrequency signals. The central unit can be formed by a microprocessor or different microprocessors. The central unit is to be considered as the set of calculation means allowing the formation and processing of the images. As described in connection with the prior art, several MRI imaging modalities are available. Among these, the so-called Tl-weighted 3D TSE modality allows rapid obtaining of anatomical images of a body area. The obtained images can be displayed in sections of the body areas in any plane relative to the central axis A of the tunnel.

As mentioned in the prior art, when considering the Tl-weighted 3D TSE acquisition sequence, there is no analytical model for estimating the relaxation time Tl (longitudinal relaxation time). When the MRI is implemented after administration of the contrast agent, knowledge of the relaxation times Tl respectively before and after administration of the contrast agent allows an estimation of the concentration of the latter, and this at each point of each slice image. See equation (1) of the section describing the prior art.

The inventors chose to establish, experimentally, a calibration function making it possible to establish a relationship between the intensity, at each point of a section image, and the relaxation time Tl. To do this, they designed a phantom, shown in FIGS. 2A to 2C. The phantom 10 is intended to be placed in the tunnel, in place of the individual examined.

The phantom 10 is formed of three identical plates 11, 12, 13 stacked on top of each other. Each plate is composed of compartments 20, comprising compositions having relaxation times T1 that are known and different from each other. In this example, the compartments 20 comprise solutions of paramagnetic chemical species whose composition and concentration are known. In the example shown in FIGS. 2A to 2C, each plate comprises 48 different tubes, arranged in a regular matrix arrangement. Each tube forms a compartment containing a solution of known composition. The tubes of the same plate comprise either the same chemical species, with respectively different concentrations, or different chemical species. The arrangement of the plates is described in more detail in connection with FIG. 2C.

By chemical species we mean a molecule or a chemical element or an ionic compound.

The three plates 11, 12, 13 are stacked on top of each other, along a stacking axis Z. The assembly is held in a solid transparent cylindrical envelope 14, made of cast PMMA (poly methyl methacrylate acrylic), filled with 9 g/L salt water to simulate the electromagnetic properties of biological tissues. Each plate extends along a length L, parallel to a longitudinal axis X and a width l, parallel to a lateral axis Y. The Length and width are sized so that the surface area of each plate is comparable to a cross-section of the body area of the individual to be studied. In this case, the body area of interest is the brain. The length of each plate is between 10 cm and 15 cm. The same is true for the width.

The diameter of the envelope 14 is typically between 15 cm and 25 cm; Thus, the phantom 10 is sized so as to replace the organ to be imaged. When acquiring an MRI image, the targeted body area is surrounded by a receiving antenna. The receiving antenna delimits a volume, intended to be occupied by the imaged body area. The phantom is preferably sized so as to occupy at least 50% or even 80% of the volume delimited by the receiving antenna.

The various compartments 11, 12 and 13 and the envelope 20 are watertight. The phantom is thus easily transportable.

The phantom 10 is intended to be placed in the tunnel 2, in place of an individual being examined. The Z axis of the phantom then extends parallel to the central axis A of the tunnel 2.

Figure 2C shows a possible arrangement of the plates and compartments. Each compartment 20 can contain: either a CuSO ₄ solution, symbolized by the letter C; or a MnCL solution, symbolized by the letter M; or a NiCL solution, symbolized by the letter N.

In Figure 2C, we observe that the arrangement chosen here for each plate is symmetrical, the compartments of the plate containing the same chemical species being arranged symmetrically with respect to the center of the plate, symbolized by a cross.

The solutions of chemical species C, M, N were chosen for their excellent stability over time, as well as their paramagnetism at the origin of their MRI contrast and the low sensitivity of their paramagnetism to temperature variations. These three contrast agents make it possible to obtain solutions covering a wide range of relaxation times Tl.

For compartments containing CuSO ₄ , the CuSO ₄ concentration is identical. For example, it is 1 mM. Compartments with identical CuSO ₄ concentrations ensure the spatial homogeneity of the MRI measurement.

The compartments containing the NiCL solutions have variable NiCL concentrations, so as to obtain relaxation times Tl at least between 0.5 s and 2.5 s. These are time corresponding to the Tl range of biological tissues. In this example, NiCl ₂ concentrations varied between 0.03 and 4.73 mM. The compartments are arranged in concentration increments of 0.1 mM.

The compartments containing the MnCI ₂ solutions have varying MnCI ₂ concentrations, so as to obtain relaxation times Tl at least between 0.5 and 3 s. In this example, the MnCI ₂ concentrations varied between 0.002 and 0.472 mM. The compartments are arranged in concentration increments of 0.01 mM.

Generally speaking, each compartment of the phantom has a known composition, and therefore known relaxation times Tl.

Figure 2D is an example of a section of a phantom plate, obtained by a Tl-weighted 3D TSE sequence. It is observed that at the level of each compartment of the phantom, the signal intensity, i.e. the gray level, is homogeneous. The signal intensity can be correlated to the relaxation time Tl, the latter being known, because the composition of each compartment is known.

In Figure 2D, we observe that the intensity of the image, for each compartment, depends on the composition of each compartment: molecule used and concentration. For the same molecule, the higher the concentration, the shorter the time Tl, and the more intense the signal, on the image, is.

Applied to the entire phantom, the Tl-weighted 3D TSE sequence allows to obtain 144 different gray levels, each of which can be associated with a chemical species and a known concentration, therefore known Tl relaxation times. It is thus possible to establish, at the level of each compartment of the phantom, a pair (Tl relaxation time, intensity).

In Figure 3, each pair (relaxation time Tl, intensity) has been plotted. In Figure 3, the abscissa axis corresponds to the relaxation time Tl (unit: second). The ordinate axis corresponds to the measured gray level (arbitrary units). In Figure 3, the solid disks correspond to the compartments containing NiCI ₂ and the circles correspond to MnCI ₂ .

From each pair (T _lt intensity), we determined, by adjustment, a bijective function f, such that I = f T^. The inverse ^-1 of this function constitutes a calibration function of the device 1, allowing to estimate, at any point of an image, of coordinates (%, y), the relaxation time T _± from the intensity: T^Oy y) = ^-1 (/(x, y)). The function of calibration f ¹ lx, y)) is an experimentally determined function. The function is preferably continuous over the range of relaxation times Tl considered.

An important aspect of the invention is that it is not necessary to use the phantom during each use. The inventors have found that the same phantom can be used to establish a calibration function of an MRI device. The same calibration function can be used to quantify the relaxation time Tl of different individuals successively examined. Thus, it is not necessary to perform a calibration prior to the examination of each individual. The calibration can be renewed at regular intervals, for example every week or every month.

The invention makes it possible to estimate, from a routine MRI imaging modality, for example the Tl-weighted 3D TSE sequence, a spatial distribution of Tl relaxation times, and this before and after the administration of a contrast agent. This makes it possible to estimate, at each point of an image, a concentration of contrast agent by implementing the expression (1) described in the prior art.

As mentioned in the prior art, the contrast agents may be nanoparticles on the surface of which metal ions, for example Gd ^{3+ ,} are chelated. The nanoparticles passively but specifically target tumors by EPR (Enhanced Permeability and Retention) effect. When nanoparticles have a theranostic effect, knowledge of their concentration makes it possible to adjust the irradiation dose during radiotherapy treatment, allowing the implementation of new therapeutic strategies: (i) maintaining the radiotherapy dose of the treatment standard and using the radiosensitizing effect of the contrast agent with theranostic effect to improve the effectiveness of the treatment on the destruction of tumor cells, (ii) reducing the radiotherapy dose of the treatment standard and using the radiosensitizing effect of the contrast agent with theranostic effect to maintain the effectiveness of the treatment on the destruction of tumor cells while limiting the exposure of healthy tissues nearby (iii) allowing two different parts of the tumor area examined, having a different concentration of contrast agent, to be irradiated so as to deliver two different doses.

This can either help to better eradicate the tumor or reduce the impact of the treatment on healthy tissues and organs at risk located near the irradiated parts of the body area, to increase survival and quality of life. Figure 4 shows the main steps of implementing the invention.

Step 100: arrangement of an individual in tunnel 2 of a device according to the invention and implementation of an MRI modality according to a 3D TSE sequence weighted in Tl.

Step 110: from radiofrequency signals emitted by a body area of the individual, obtaining images of sections of the body area in an axial, coronal or sagittal plane. Step 110 is implemented by the processing unit 3. The images can be normalized to the values of the proton density, obtained experimentally or predefined.

Steps 100 and 110 are performed prior to and subsequent to an administration of a contrast agent. The administration of the contrast agent is not part of the invention.

Step 120: estimation of a relaxation time Tl at different points of each image resulting from step 110. The estimation can be adjusted by taking into account a healthy area of tissue, whose relaxation time Tl is known, for example gray matter or white matter.

Step 130: from the relaxation times Tl resulting from step 120, determination of a concentration of contrast agent at the different points of each image.

Step 120 implements the calibration function ^-1 previously described. The calibration function is implemented by the processing unit 3, in which the calibration function has previously been stored. The calibration function is established during a calibration phase which comprises the following steps

Step 90: Arrange the phantom in the device tunnel and implement an MRI modality according to a Tl-weighted fast spin echo imaging sequence, for example a Tl-weighted 3D TSE sequence.

Step 91: from radiofrequency signals emitted by each compartment of the phantom, obtaining images of sections of the phantom in an axial, coronal or sagittal plane.

Step 92: definition of couples (intensity - relaxation time Tl) from the images resulting from step 91

Step 93: determination, by adjustment or regression, of a calibration function linking the intensity to the relaxation time Tl. The calibration function is preferably continuous over a range of relaxation times Tl varying from 0.1 s to 2.5s or 3s. Steps 90 to 93 constitute a calibration phase. Preferably, the calibration phase is not repeated for each examination of a new individual. Thus, the same calibration function is used to interpret different images of different individuals.

Maps of concentrations of Gd ³⁺ chelated on AGulX nanoparticles as described in the prior art were compared, using a method as previously described, as well as on the basis of a determination of the relaxation time Tl of a VFA modality. Images were also acquired on individuals with brain metastases. Brain metastases were systematically identified using the invention, i.e. on the basis of an estimation of the concentration of contrast agent carried out from determinations of the relaxation time Tl on images acquired according to the Tl-weighted 3D TSE modality.

In some instances, estimates of contrast agent concentrations based on Tl relaxation times on images acquired using the VFA modality did not reveal the presence of metastases.

The duration of each Tl-weighted 3D TSE acquisition was 4 minutes 57 seconds, compared to 11 minutes and 24 seconds for the VFA modality. The thickness of each Tl-weighted 3D TSE slice was 1 mm, compared to 4 mm for the VFA modality. The inter-slice spacing was 0 mm for the Tl-weighted 3D TSE modality, compared to 2 mm for the VFA modality. Thus, the Tl-weighted 3D TSE sequence provides a Tl measurement with higher spatial resolution in less time.

The use of a Tl-weighted 3D TSE sequence for the estimation of Tl relaxation time before and after the administration of a contrast agent, followed by an estimation of the contrast agent concentration, therefore seems particularly relevant.

The impact on Tl relaxation time of the administration of a contrast agent on healthy tissues was also estimated. The Tl-weighted 3D TSE sequence was used: no significant increase was observed on white matter and gray matter. The VFA method was also used: increases of the order of 300 to 500 ms were measured on white matter and gray matter, while these are healthy tissues. It seems that the estimates of Tl relaxation times using the invention (TSE sequence + application of an empirical calibration function, obtained on a phantom) are more consistent than with the VFA modality. Although described in connection with radiosensitizing nanoparticles, the invention can be applied to Gd-based contrast agents of the gadoteric acid type such as gadopentetate dimeglumine, gadobutrol, gadopiclenol or contrast agents based on manganese complex, USPIO (Ultrasmall superparamagnetic iron oxide) and SPIO (superparamagnetic iron oxide), this list not being exhaustive.

Claims

1. Method for measuring a concentration of a contrast agent in at least one body area of a living human or animal individual, the method comprising: - a) acquisition of radiofrequency signals emitted by the organ, subjected to a magnetic resonance imaging modality; - b) from the radiofrequency signals resulting from a), obtaining images of the body area; - c) from the images resulting from b), estimation of a relaxation time Tl in different parts of the body area; - steps a) to b) being implemented before and after an administration of a contrast agent to the individual, so as to obtain, during step c) an estimate of relaxation time Tl, in the body area, respectively before and after the administration of the contrast agent, the administration of the contrast agent not being part of the measurement method; - d) from the relaxation times Tl estimated respectively before and after the administration of the contrast agent, determination of a concentration of the contrast agent of the different parts of the body area; the method being characterized in that - in each step a), the magnetic resonance imaging modality is a Tl-weighted fast spin echo modality; - step c) is implemented using a calibration function, establishing a correspondence between the intensity of an image of the body area and the relaxation time Tl, the calibration function being obtained, prior to step c), during a calibration phase comprising: • i) arrangement of a phantom (10), comprising different compartments (20) respectively comprising different compositions whose relaxation times Tl are known and different from each other; • ii) acquisition of radiofrequency signals emitted by the phantom, subjected to the magnetic resonance imaging modality; • iii) from the radiofrequency signals acquired during ii), obtaining images of the different compartments of the phantom; • iv) establishing the calibration function from the intensity of the signal in said compartments of the phantom taking into account, for each compartment, a pair of values formed by the signal intensity of an image of the compartment and the relaxation time Tl of the composition of said compartment. 2. Method according to claim 1, in which: - in step a), the body area is placed in a tunnel, subjected to a magnetic field; - during step i) of the calibration phase, the phantom is placed in the tunnel, in place of the body area. 3. Method according to any one of the preceding claims, in which different compartments of the phantom respectively comprise different chemical species or different concentrations of the same chemical species, so that the relaxation times Tl of said compartments are different from each other. 4. Method according to any one of the preceding claims, in which different compartments of the phantom comprise the same concentration of the same chemical species. 5. Method according to any one of the preceding claims, in which the calibration function is established by a regression, from the pairs (intensity of the signal of an image of a compartment - relaxation time Tl of the compartment) taken into account during step iv) of the calibration phase. 6. Method according to any one of the preceding claims, in which: - steps a) to d) are implemented successively for different individuals; - the calibration function implemented in each step c) is identical for the different individuals, and results from the same calibration phase. 7. Method according to any one of the preceding claims, in which the contrast agent is a radiosensitizing agent, the method comprising, following step d), a determination of an irradiation dose in different parts of the body area, as a function of the concentrations of the contrast agent resulting from step d). 8. Processing unit (3), configured to process an image acquired by a magnetic resonance imaging device (1), according to a fast spin echo acquisition modality, weighted in Tl, the processing unit being programmed to apply, to the image acquired, a calibration function resulting from the implementation of a calibration phase comprising: • arrangement of a phantom (10), comprising different compartments (20) respectively comprising different compositions whose relaxation times Tl are known and different from each other, the phantom being arranged in the magnetic resonance imaging device; • acquisition of radiofrequency signals emitted by the phantom, subjected to the fast spin echo type magnetic resonance imaging modality; • from the radiofrequency signals acquired during the previous step, obtaining images of the different compartments of the phantom; • establishing the calibration function from the intensity of the signal in said compartments of the phantom by taking into account, for each compartment, a pair of values formed by the intensity of the signal of an image of the compartment and the relaxation time Tl of the composition of said compartment.