A magnetic resonance temperature imaging method and system

Abstract

The present application discloses a magnetic resonance temperature imaging method, which comprises the following steps: obtaining a phase map using a gradient echo sequence containing at least two different echo times, obtaining a phase difference map corresponding to at least two different echo times according to the phase map, and in the process of obtaining the phase difference map, using at least one phase difference map corresponding to an echo time as a reference phase difference map to calibrate at least one phase difference map corresponding to a different echo time; obtaining a temperature difference map corresponding to at least two different echo times according to the phase difference map, and obtaining a temperature map corresponding to at least two different echo times by using the temperature difference map; and obtaining a weighted temperature map using the temperature map corresponding to at least two different echo times.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and more specifically, to a magnetic resonance temperature imaging method and system. Background Art

[0002] Magnetic Resonance Temperature Imaging (MRTI) can realize non-invasive, real-time, in vivo monitoring of the internal temperature distribution and changes of the subject, and is of great use in the monitoring process of minimally invasive and non-invasive thermal therapy, such as magnetic resonance interstitial thermal therapy and focused ultrasound therapy.

[0003] One of the current magnetic resonance temperature imaging methods is the temperature measurement method based on the displacement of the proton resonance frequency (PRF). In practice, it is found that the temperature measurement method based on the displacement of the proton resonance frequency is greatly affected by objective environmental factors such as the uniformity of the magnetic field of the magnetic resonance coil, the uneven distribution of the magnetic susceptibility of the tissue, and the tissue movement caused by breathing / blood flow pulsation. Common errors caused include phase unwrapping errors, errors caused by rapid changes in magnetic susceptibility, errors caused by movement, etc., which can easily cause the final temperature map to differ greatly from the actual temperature, which makes the temperature map lose its reference significance. Summary of the invention

[0004] In order to solve or alleviate one or more of the above-mentioned technical problems, the present application provides a magnetic resonance temperature imaging method and related devices to achieve the purpose of reducing the error of the temperature map finally obtained and improving the accuracy of the temperature map.

[0005] In a first aspect, the present invention provides a method for magnetic resonance temperature imaging, comprising the following steps:

[0006] A phase map is obtained using a gradient echo sequence having at least two different echo times, and a phase difference map corresponding to at least two different echo times is obtained according to the phase map. In the process of obtaining the phase difference map, a phase difference map corresponding to at least one echo time is used as a reference phase difference map to calibrate the phase difference map corresponding to at least one different echo time.

[0007] According to the phase difference map, a temperature difference map corresponding to at least two different echo times is obtained,

[0008] According to the temperature difference map, a temperature map corresponding to at least two different echo times is obtained,

[0009] A weighted temperature map is obtained using the temperature maps corresponding to at least two different echo times.

[0010] Optionally, the echo time corresponding to the reference phase difference map is smaller than the echo time corresponding to the phase difference map for calibration.

[0011] Further, obtaining the phase difference map includes the step of obtaining a calibrated phase difference map:

[0012] According to the relationship that the phase difference map is proportional to the echo time, an estimated value of the phase difference map to be calibrated is calculated based on the echo time and the reference phase difference map;

[0013] The estimated value is used to unwrap the phase difference map to be calibrated according to the phase periodicity to obtain a calibrated phase difference map.

[0014] Optionally, the method of the present invention further comprises the step of eliminating the phase drift caused by the magnetic resonance system. Further, the step of eliminating the phase drift caused by the magnetic resonance system is performed on a phase difference map or a temperature change map, specifically as follows: a number of areas with stable and unchanged physical temperature and uniform tissue are selected as thermal reference points, and correction is performed by subtracting the average phase difference of the thermal reference points from each phase difference image or subtracting the average temperature change of the thermal reference points from the temperature change map.

[0015] Optionally, the method of the present invention further comprises a step of magnetic susceptibility correction. Further, the step of magnetic susceptibility correction is performed on a phase difference map or a temperature difference map;

[0016] The steps for performing magnetic susceptibility correction on the temperature difference map include:

[0017] Obtain a first temperature map according to the reference phase difference map, and obtain a corresponding second temperature map according to the calibrated phase difference map;

[0018] Determine whether the absolute value of the difference between the temperature value corresponding to each pixel in the second temperature map and the temperature value corresponding to the corresponding pixel in the first temperature map exceeds a preset temperature threshold, and if so, correct the temperature value corresponding to the corresponding pixel in the second temperature map;

[0019] The steps for performing magnetic susceptibility correction on the phase difference map include:

[0020] It is determined whether the absolute value of the difference between the phase difference value corresponding to each pixel in the calibrated phase difference map and the phase difference value corresponding to the corresponding pixel in the reference phase difference map exceeds a preset phase difference threshold; if so, the phase difference corresponding to the corresponding pixel in the calibrated phase difference map is corrected.

[0021] Optionally, the method of the present invention further comprises a step of correcting the phase error caused by motion. Further, the step of correcting the phase error caused by motion is performed on a phase difference map or a temperature difference map;

[0022] The steps performed on the phase difference map to correct the phase error caused by motion include:

[0023] The phase error caused by motion is removed by linear least square fitting at each pixel of the phase difference map;

[0024] The steps to correct motion-induced phase errors on the temperature difference map include:

[0025] A linear least squares fit is used at each pixel in the temperature difference map to remove the error caused by motion.

[0026] The above optional steps of eliminating the phase drift caused by the magnetic resonance system, correcting the error caused by the magnetic susceptibility, and correcting the phase error caused by motion, selected individually, partially or in whole as part of the ablation calculation method of the present invention, all fall within the scope of the present invention.

[0027] A second aspect of the present invention provides a magnetic resonance temperature imaging system, which includes: a memory and a processor;

[0028] The memory is used to store program codes, the processor is used to call the program codes, and the program codes are used to execute the magnetic resonance temperature imaging method of the present invention.

[0029] Furthermore, the magnetic resonance temperature imaging method is not a retrospective algorithm or an iterative algorithm, has a small amount of computation, can provide an almost real-time temperature map, and has a high reference significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0031] Figure 1 Amplitude, phase and temperature diagrams obtained in ex vivo and in vivo environments;

[0032] Figure 2 A schematic flow chart of a magnetic resonance temperature imaging method provided for one embodiment of the present application;

[0033] Figure 3 The acquired phase map, phase difference map and temperature map provided for one embodiment of the present application;

[0034] Figure 4 A schematic flow chart of a magnetic resonance temperature imaging method provided for another embodiment of the present application;

[0035] Figure 5 A schematic diagram of an experimental device provided for one embodiment of the present application;

[0036] Figure 6 A partial enlargement of a temperature map of a laser-induced thermal therapy for an in vitro pork experiment provided by one embodiment of the present application;

[0037] Figure 7 In one embodiment of the present application, a tissue mimic ( Figure 7 (a)) or in vitro pork ( Figure 7 (b) Schematic diagram of temperature variation with time during the experiment;

[0038] Figure 8 A representative temperature diagram of dog 01 in an in vivo experiment provided as an example of the present application; DETAILED DESCRIPTION

[0039] Magnetic resonance temperature imaging can guide a variety of energy delivery treatment methods, such as laser interstitial thermal therapy, focused ultrasound therapy, radiofrequency ablation, etc., to monitor the temperature of the target tissue and the treatment effect. The inventors found through research that the main sources of errors in the acquired temperature map are phase errors caused by phase unwrapping misalignment, magnetic susceptibility errors, and phase errors caused by motion. As the input energy dose changes, magnetic susceptibility distortion will cause a decrease in image amplitude and corresponding errors in image phase, thereby destroying the reconstructed temperature map of the heating center and its surroundings. Errors in the reconstructed temperature map may lead to errors in the estimation of the ablation area, which may lead to changes in the treatment effect and thermal damage to critical tissues. Therefore, accurate temperature imaging is crucial to the effectiveness and safety of treatment, especially when applied to the ablation area of ​​​​brain tissue.

[0040] Thermometry based on the proton resonance frequency shift is based on the following principle: the resonance frequency of hydrogen protons changes with the temperature in water molecules. For water-containing tissues, the change of the local magnetic field with temperature can be described as:

[0041]

[0042] Wherein, α is the proton resonance frequency coefficient that changes with temperature, and in the present invention, it is 0.008-0.015ppm / °C. The corresponding resonance frequency change of water protons affected by temperature can be expressed as:

[0043] Δf=αγB 0 ·ΔT; (2)

[0044] Where ΔT represents the temperature change, Δf represents the resonant frequency change, γ represents the gyromagnetic ratio, and B 0 Represents the static magnetic field strength.

[0045] The change in resonance frequency due to temperature change can be observed in the phase of complex magnetic resonance imaging. For a given interval TE of the gradient echo sequence, the relative temperature change ΔT can be calculated based on the phase difference Δφ, which can be expressed as:

[0046]

[0047] The gradient echo sequence is a sequence used in the temperature measurement method based on the proton resonance frequency shift, including sequences corresponding to different echo times, for example, an echo sequence corresponding to two echo times. According to formula (3), the longer the echo time in the gradient echo sequence, the greater the phase difference may be caused by the same temperature change, and a higher temperature sensitivity can be obtained.

[0048] refer to Figure 1 , showing various common problems existing in the temperature map obtained by using an echo sequence with a single echo time in the prior art. As the echo time of the gradient echo sequence increases, both the phase contrast and the phase wrapping increase, which indicates that using a sequence with a longer echo time has a higher temperature sensitivity and more phase wrapping. Figure 1 , the amplitude (upper row) and phase (second row) of the first to fourth echoes obtained by the gradient echo sequence containing 4 different echo times used in the embodiment of the present application in (a) (ex vivo, pig brain) and (b) (in vivo). The temperature map (lower row) is calculated according to each TE (echo time) setting using a traditional PRF algorithm. More phase wrapping will appear at longer echo times, and the image contrast will also increase accordingly. Strong heating will cause signal loss due to changes in magnetic susceptibility, and will also translate into phase and temperature errors of pixels around the heating center. In vivo experiments, cerebrospinal fluid (CSF) movement will cause incorrect high temperatures in magnetic resonance temperature imaging, which is more obvious in the previous echo because the shorter TE has a lower tolerance to similar phase errors introduced.

[0049] Specifically, inter-scan motion is a problem in temperature maps measured by proton resonance frequency shift-based thermometry due to the motion of cerebrospinal fluid (CSF) in the brain. The amplitude and phase signals of CSF are often altered on pulsed gradient echo sequences by the normal dynamic motion of CSF, which can confound temperature estimates. CSF motion can also cause pixels in and around the ventricles to shift, resulting in errors in phase contrast maps. Figure 1 (b) In vivo temperature map shows false hyperthermia in the third ventricle due to CSF ​​movement. Temperature errors are more pronounced on pulse gradient echo sequences with shorter echo times because, according to equation (3), they are less tolerant of the magnitude of the phase shift introduced by CSF flow.

[0050] In fact, the local magnetic field of water protons should also take into account the magnetic susceptibility x 0 , formula (1) becomes:

[0051]

[0052] in, Represents the local magnetic field change caused by magnetic susceptibility.

[0053] Laser heating can cause significant magnetization artifacts in GRE imaging around the laser tip. Figure 1 , the heating center with rapid temperature changes (such as Figure 1 The arrow in (a) shows severe signal loss at longer echo time amplitudes. Intra-voxel spin dephasing is caused by local magnetic field inhomogeneities due to temperature and magnetic susceptibility variations.

[0054] Heating-induced magnetization artifacts, especially those in images corresponding to gradient echo sequences with longer echo times, are an important source of error. Figure 1 , in ex vivo or in vivo experiments, phase errors around the heating center translate into pseudo-hypothermia on MR thermography. Typically, in gradient echo MR imaging, it is recommended to use a delay time with the shortest possible echo time to minimize susceptibility artifacts. However, gradient pulse sequences with longer echo times can provide better temperature sensitivity and signal-to-noise ratio, which is a current dilemma.

[0055] In order to take into account temperature sensitivity, signal-to-noise ratio and low error, the present application embodiment provides a magnetic resonance temperature imaging method, comprising the following steps:

[0056] A phase map is obtained using a gradient echo sequence having at least two different echo times, and a phase difference map corresponding to at least two different echo times is obtained according to the phase map. In the process of obtaining the phase difference map, a phase difference map corresponding to at least one echo time is used as a reference phase difference map to calibrate the phase difference map corresponding to at least one different echo time.

[0057] According to the phase difference map, a temperature difference map corresponding to at least two different echo times is obtained,

[0058] According to the temperature difference map, a temperature map corresponding to at least two different echo times is obtained,

[0059] A weighted temperature map is obtained using the temperature maps corresponding to at least two different echo times.

[0060] The gradient echo sequence with at least two different echo times means that the gradient echo sequence has at least two different echo times, for example (6ms, 18ms), (, 6ms, 24ms), (12ms, 18ms), (6ms, 12ms, 24ms), (6ms, 18ms, 24ms), (6ms, 12ms, 18ms, 24ms), (6ms, 12ms, 18ms, 24ms), (6ms, 12ms, 12ms, 24ms), (6ms, 18ms, 18ms, 24ms), etc.

[0061] Furthermore, the magnetic resonance temperature imaging method is not a retrospective algorithm or an iterative algorithm, has a small amount of computation, can provide an almost real-time temperature map, and has a high reference significance.

[0062] A specific example of the magnetic resonance temperature imaging method of the present invention is as follows: Figure 2 As shown, including:

[0063] S101: obtaining a phase map using a gradient echo sequence having at least two different echo times, obtaining phase difference maps corresponding to at least two different echo times according to the phase map, and in the process of obtaining the phase difference map, using at least one phase difference map corresponding to an echo time as a reference phase difference map to calibrate at least one phase difference map corresponding to a different echo time;

[0064] In step S101, the value of the echo time in the gradient echo sequence can be determined according to actual needs. For example, for head imaging, the value range of the echo time of the optional gradient time sequence is 3 to 30 ms, and the specific values ​​of the echo time contained in the gradient echo sequence must be within this value range.

[0065] The gradient echo sequence information may be obtained by reading or receiving from a server or other storage device, or may be obtained in real time according to the setting of the staff. The present application does not limit the specific method of obtaining multiple gradient echo sequences, which depends on the actual situation.

[0066] refer to Figure 3 , specific reference Figure 3 (a), the phase diagram obtained is as follows Figure 3 (a) shown.

[0067] S102: Obtaining a temperature difference map corresponding to at least two different echo times according to the phase difference map,

[0068] S103, obtaining temperature maps corresponding to at least two different echo times according to the temperature difference map, see Figure 3 (d) , where the error due to magnetic susceptibility is shown.

[0069] S104: Obtain a weighted temperature map using the temperature maps corresponding to at least two different echo times.

[0070] The following describes a feasible implementation method of each step of the magnetic resonance temperature imaging method provided in an embodiment of the present application.

[0071] On the basis of the above embodiment, in the embodiment of the present application, the specific steps of obtaining the temperature difference map include:

[0072] The phase difference map at the time is obtained by subtracting the phase map at the reference time from the phase map at any of the different time points, wherein the reference time is any time before energy (such as heat energy, light energy, radiofrequency ablation, cryoablation) is transmitted to the target tissue, preferably a time shortly before energy transmission, such as a time when energy transmission is about to be performed;

[0073] Optionally, the echo time of the reference phase difference map is less than or equal to 18ms. Preferably, the echo time of the reference phase difference map does not exceed 17ms, 16ms, 15ms, 14ms, 13ms, 12ms, 11ms, 10ms, 9ms, 8ms or 7ms. More preferably, the echo time of the reference phase difference map does not exceed 6ms, 5ms or 4ms.

[0074] Using at least one phase difference map corresponding to an echo time as a reference phase difference map to calibrate at least one phase difference map corresponding to a different echo time comprises the following steps:

[0075] Using the reference phase difference map and the phase difference map corresponding to the echo time to be calibrated, according to the relationship that the phase difference is directly proportional to the echo time, based on the phase difference between the echo time and the reference phase difference map, an estimated value of the phase difference of the phase difference map corresponding to the echo time to be calibrated is calculated; then using the estimated value, the phase difference to be calibrated is unwrapped according to the phase periodicity to obtain the calibrated phase difference.

[0076] Based on the above embodiment, in another embodiment of the present application, the magnetic resonance temperature imaging method further includes:

[0077] S105 is a step of eliminating the phase drift caused by the magnetic resonance system, which is performed on the phase difference map or the temperature map.

[0078] The steps of eliminating the phase drift caused by the magnetic resonance system on the phase difference map include:

[0079] Selecting a plurality of thermal reference points (Region of Interest, ROI), and subtracting the average phase difference of the thermal reference points from each phase difference map;

[0080] The steps for eliminating the phase drift caused by the magnetic resonance system on the temperature difference map include:

[0081] The average temperature difference of any thermal reference point is subtracted from the temperature difference map to perform correction. There are many ways to make the correction, for example, the temperature value of the reference temperature map can be used to replace the temperature value of the temperature map to be calibrated, or the temperature value of the adjacent pixel in the second temperature map can be used to replace the temperature value of the pixel in the second temperature map, or an approximate temperature can be fitted based on the temperature value of the adjacent pixel and the temperature value of the first temperature map to replace the temperature value of the second temperature map.

[0082] Based on the above embodiment, in another embodiment of the present application, the magnetic resonance temperature imaging method further includes:

[0083] S106: a step of correcting the magnetic susceptibility, which is performed on the phase difference map or the temperature map;

[0084] The steps to perform a susceptibility correction on a temperature map include:

[0085] Obtain a first temperature map according to the reference phase difference map, and obtain a corresponding second temperature map according to the calibrated phase difference map;

[0086] Optionally, a preset area is determined in the first temperature map and each of the second temperature maps;

[0087] Determine whether the absolute value of the difference between the temperature value corresponding to each pixel in the second temperature map and the temperature value corresponding to the corresponding pixel in the preset area in the first temperature map exceeds a preset temperature threshold, and if so, correct the temperature value corresponding to the corresponding pixel in the second temperature map; there may be multiple methods for correction, for example, the temperature value of the first temperature map may be used to replace the temperature value of the second temperature map, or an approximate temperature may be fitted based on the temperature values ​​of adjacent pixels and the temperature values ​​of the first temperature map to replace the temperature value of the second temperature map;

[0088] The steps for performing magnetic susceptibility correction on the phase difference map include:

[0089] Determining a preset area in the reference phase difference map and the calibrated phase difference map;

[0090] It is determined whether the absolute value of the difference between the phase difference value corresponding to each pixel in the calibrated phase difference map and the phase difference value corresponding to the corresponding pixel in the preset area in the reference phase map exceeds a preset phase difference threshold. If so, the phase difference in the calibrated phase difference map is corrected.

[0091] On the basis of the above embodiment, in another embodiment of the present application, the magnetic resonance temperature imaging method further includes:

[0092] S107: a step of correcting a phase error caused by motion on a phase difference map or a temperature map;

[0093] The steps performed on the phase difference map to correct the phase error caused by motion include:

[0094] removing motion-induced phase errors by linear least squares fitting at each pixel using the reference phase difference map and the calibrated phase difference map;

[0095] The steps to correct motion-induced phase errors on the temperature map include:

[0096] Obtain a first temperature map according to the reference phase difference map, and obtain a corresponding second temperature map according to the phase difference map;

[0097] The motion-induced phase error is removed by linear least squares fitting at each pixel using the first temperature map and the second temperature map.

[0098] Still reference Figure 4 (c) Figure 4 (c) shows the phase difference (first row) and relative temperature change (second row) as a function of time without (left) and with (right) motion error correction. For shorter echo times, the phase error Δφ(x,y) bias Large temperature deviations are introduced but are correctly eliminated after a linear least squares fit.

[0099] As mentioned above, steps S105, S106 and S107 can be performed on the phase difference map or the temperature map. That is, the step of eliminating the phase drift caused by the magnetic resonance system is performed on the phase difference map and / or the temperature map, the step of correcting the magnetic susceptibility is performed on the phase difference map and / or the temperature map, and the step of correcting the phase error caused by the motion is performed on the phase difference map and / or the temperature map.

[0100] Still reference Figure 3 , Figure 3 (c) shows the corrected phase difference map, Figure 3 (d) represents the temperature diagram, Figure 3 (e) shows the image after magnetic susceptibility correction. Figure 3 (f) shows the image after motion error correction.

[0101] On the basis of the above embodiment, in another embodiment of the present application, the using the reference phase difference map and the calibrated phase difference map to calculate the temperature map at the moment includes:

[0102] S104: using the temperature maps corresponding to at least two different echo times to obtain a weighted temperature map, the temperature maps of different echo times include at least one reference phase difference map and a temperature map calculated from a calibrated phase difference map, and weighting the calculated temperature map to obtain a weighted temperature map, the weighting can be performed on the temperature difference map or the temperature map. Figure 2 The temperature map can be obtained by adding the temperature difference map to the reference temperature, so the two can be interchanged; weighting can be various weighting methods, such as average weighting, or for example, the temperature map can be a temperature map corresponding to a single echo time, that is, the weighting coefficient of the temperature map corresponding to the echo time is 1, and the weighting coefficients of the temperature maps corresponding to other echo times are 0.

[0103] After step S104, the following steps may also be included:

[0104] S108: performing multiple interpolation processes on the temperature map of the target to be measured, and calculating the ablation region boundary using the temperature map of the target to be measured after the interpolation process.

[0105] The purpose of performing multiple interpolation processes on the temperature map of the target to be measured is to obtain a smoother ablation area boundary. The specific number of interpolation processes may be 2 or 3 times.

[0106] In the process of calculating the ablation area boundary, the following formula is used:

[0107]

[0108] Among them, E a represents the activation energy, A is the frequency factor, R is the universal gas constant, T(τ) is the function of temperature (°C) and time τ, and t is the current time. Pixels with Ω values ​​exceeding 1 are considered to be ablated.

[0109] The magnetic resonance temperature imaging method provided in the embodiment of the present application is verified below with reference to specific examples.

[0110] The tissue mimic (gel phantom) was heated using a laser ablation system consisting of a 10 W, 980 nm diode laser and a cooled laser applicator system. Phase images were acquired on a 3T MR scanner (Ingenia, Philips Healthcare, Best, The Netherlands) using 16 receive coils using a multi-echo temporal gradient echo sequence: flip angle = 30°, TE = 6 / 12 / 18 / 24 ms, TR = 22 ms, matrix = 176 × 176, FOV = 200 x 200 mm 2 , slice thickness = 5 mm, 3 s / image.

[0111] like Figure 5As shown, two MR-compatible fiber optic temperature probes were also inserted into the tissue mimic, with the probe tips located close to the ablation fibers to obtain the gel temperature at each point. Since the fiber optic probes were affected by the ablation fibers during the heating process, the thermometers only monitored the cooling phase. Figure 5 middle. Specifically, Figure 5 Figure 2 shows an ablation fiber and two fiber-optic temperature probes inserted into a tissue mimic, with a gel-filled reference tube fixed around it as an insulating reference.

[0112] The ex vivo experiments on pork and porcine brain were performed using the same scanning parameters as the tissue simulant experiments. The tissues were placed in a plastic container filled with water to isolate the air. Four reference tubes filled with gel were fixed around the periphery as insulating references. Two experiments were performed for each type of tissue (gel, pork, porcine brain), one with several cycles of laser heating and the other with continuous heating and cooling. The root mean square error between the MR-measured temperature and the fiber-optic-measured temperature was calculated as a measure of temperature accuracy.

[0113] The in vivo experiments in Doberman Pinschers were approved by the Ethics Review Committee of Tsinghua University. Nine adult Doberman Pinschers received thermal therapy with laser interstitial thermal therapy. The heating process was monitored on a 3T MR scanner (Ingenia, Philips Healthcare, Best, The Netherlands) using a multi-echo temporal gradient echo sequence with a 32-line receiving head coil.

[0114] Still reference Figure 3 and Figure 4 , Figure 3 and Figure 4 An example of a magnetic resonance temperature imaging method provided by an embodiment of the present application is described. Figure 3 A specific example step of magnetic resonance temperature imaging is shown. Figure 3 (a), A time acquired during laser thermal therapy, where a coil combination phase image is first obtained by a multi-TE echo sequence; Figure 3 (b), then the phase difference image is obtained, and the white arrow indicates the phase wrapping that occurs on the phase difference image around the heating center; Figure 3 (c) shows the phase difference map after phase unwrapping and B0 drift correction. Figure 3 (d) According to the PRF shift method, Figure 3 (c) Calculated temperature map. The white arrows highlight the error caused by the magnetic susceptibility. Figure 3 (e) Temperature map after magnetic susceptibility correction. The white arrows show the error caused by residual CSF motion. Figure 3 (f) Motion-corrected temperature map.

[0115] Figure 4 In FIG. 1 , an exemplary method flow for a representative pixel includes: Step 1, obtaining a phase difference map (upper left) and a phase map obtained by unwrapping the reference phase difference map (upper right). As shown by the black arrow, under the condition of rapid temperature change, the phase difference map corresponding to the 18ms and 24ms echo times is wrapped. Step 2, obtaining a phase unwrapping map. Step 3, static magnetic field strength (B 0 ) Drift correction, B0 drift correction is to reduce system fluctuations, step 4, magnetic susceptibility-induced phase error correction, the temperature error caused by magnetic susceptibility changes (black arrows) at longer echo times (TE) is corrected using the shortest echo time (TE). Step 5 motion-induced phase error correction. The first and second lines are the phase difference over time and the corresponding temperature change over time. For multiple echo times, the phase error caused by motion (black arrows) is almost the same, thus resulting in more obvious temperature errors at shorter TEs. The result of correcting the motion error shows smoother phase difference and temperature curves.

[0116] Phantom and in vitro experimental results:

[0117] Figure 6 A representative temperature map of an in vitro pork experiment during laser-induced thermal therapy is shown. Six representative images are selected from 300 frames (3s / frame) acquired during the thermal cycle (#50 represents the 50th frame, #146 represents the 146th frame, and so on). The first and second rows are respectively the phase unwrapping method using the traditional phase unwrapping method and the phase unwrapping method based on multi-echo time proposed in the embodiment of the present application. Using the phase unwrapping method of the prior art, the pixels on the temperature map will be severely damaged due to the change in magnetic susceptibility caused by laser heat, and cannot be restored even if the laser is no longer used. The technical principle is as follows: the phase unwrapping method of the prior art is applied to the time dimension for phase jump detection. If the phase difference map of the current frame is incorrectly unwrapped, all subsequent frames will be affected. On the other hand, the magnetic resonance temperature imaging method proposed in the present invention is performed on the basis of the multi-echo dimension, thereby avoiding interference from previous frames. The third row is a single echo time temperature map after phase unwrapping and magnetic susceptibility correction, and the damaged pixels around the heating center have been correctly restored. The last row is the result of multi-echo time data combination using the magnetic resonance temperature imaging method provided in the embodiment of the present application. The resulting MRI showed a more uniform temperature at the hot spot.

[0118] Figure 7 It is shown that in tissue mimics ( Figure 7 (a)) or in vitro pork ( Figure 7 (b)) Temperature variation over time during the experiment, measured by two temperature measuring optical fibers (red lines) and calculated by the method provided in the present application (dashed black lines). Figure 7 (a)) or single heating ( Figure 7 (b)), the temperature-time behavior calculated by the proton resonance frequency (PRF) matches well with the temperature-time measured by the temperature measuring fiber during the cooling stage. Table 1 lists the root mean square error (RMSE) values ​​between the MR calculated values ​​and the temperature measuring fiber measured values, which represent the temperature accuracy of the proposed algorithm. Experiment 1 performed several laser cycle heating, while Experiment 2 was a continuous heating and cooling stage. The results show that in most cases, the RMSE of gel, pork or pig brain tissue is less than 0.5°C. Table 1. Comparison between the temperature measured by the optical fiber and the temperature calculated by MR using the method provided in an embodiment of the present application.

[0119]

[0120] Abbreviations: RMSE, root mean square error; Experiment, Experiment L (R), fiber optic temperature probe on the left (right).

[0121] Figure 8 Representative temperature maps of dog 01 in the in vivo experiment are shown. It should be noted that the ablation area is located close to the third ventricle and the lateral ventricle. 100 frames (3s / frame) of images acquired during laser ablation were selected and superimposed on the T2w magnetic resonance thermal imaging after ablation. From top to bottom are temperature maps calculated by the prior art algorithm based on single echo time (TE) data (TE=6ms and TE=24ms), and temperature maps calculated using the algorithm proposed in the present invention based on multiple (joint) TE echo sequences. The first row (TE=6ms) shows pseudo-high temperatures in the third lateral ventricle and the lateral ventricle, indicating that the short TE calculated temperature is severely affected by CSF flow artifacts. CSF-induced artifacts in the third ventricle (indicated by white arrows) are still present in the second row (TE=24ms), but are well suppressed by the proposed multiple (joint) TE echo sequence algorithm. The second row shows that a longer TE (TE = 24 ms) provides smoother boundaries and better temperature SNR compared to a shorter TE (TE = 6 ms), but as mentioned above, the pixels around the heating center are corrupted due to the change in magnetic susceptibility. On the other hand, our proposed method integrates the information of multiple echoes, so the obtained temperature map eliminates both the CSF-induced error and the magnetic susceptibility-induced error, showing a more uniform and symmetrical heating area.

[0122] The above experimental results show that the magnetic resonance temperature imaging method provided by the embodiment of the present application can be used to correct the error caused by the magnetic susceptibility in the proton resonance frequency temperature map caused by the heating laser itself. We first proposed to apply the multi-echo time gradient echo pulse sequence to magnetic resonance thermal imaging through the proton resonance frequency shift method. Instead of a single echo sequence, a multi-gradient echo sequence can provide more information without additional scanning time, and provides a new method for phase unwrapping and artifact removal.

[0123] Shorter echo times can tolerate magnetic susceptibility artifacts but are sensitive to noise, while longer echo times have better temperature sensitivity and signal-to-noise ratio, but are greatly affected by magnetic susceptibility artifacts. The magnetic resonance temperature imaging method proposed in the present invention combines the advantages of different echoes to obtain better temperature map measurement results. Moreover, the magnetic resonance temperature imaging method of the present invention can significantly improve the robustness and signal-to-noise ratio of magnetic resonance thermal imaging, thereby avoiding damage to healthy tissues due to misjudgment of low temperature.

[0124] The proposed method also has excellent CSF flow phantom rejection and can provide accurate temperature measurements in or around the ventricles. Compensation for errors caused by CSF movement is clinically important for laser-induced thermal therapy of periventricular brain lesions. Furthermore, the proposed algorithm is online compatible and does not require iterative calculations, making it well suited for magnetic resonance thermography where very close to real-time temperature maps are required.

[0125] The magnetic resonance temperature imaging system provided in the embodiment of the present application is described below. The magnetic resonance temperature imaging system described below can be referenced to the magnetic resonance temperature imaging method described above.

[0126] Accordingly, an embodiment of the present application also provides a magnetic resonance temperature imaging system, characterized in that it includes: a memory and a processor;

[0127] The memory is used to store program codes, the processor is used to call the program codes, and the program codes are used to execute the magnetic resonance temperature imaging method according to any one of claims 1 to 9.

[0128] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0129] The features described in the embodiments of the present specification may be replaced or combined with each other. Each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments may be referenced to each other.

[0130] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A magnetic resonance temperature imaging method, characterized in that: The following steps are involved: A phase map is obtained using a gradient echo sequence containing at least two different echo times, and a phase difference map corresponding to at least two different echo times is obtained according to the phase map. In the process of obtaining the phase difference map, a phase difference map corresponding to at least one echo time is used as a reference phase difference map to calibrate at least one phase difference map corresponding to a different echo time, and the calibration is performed as follows: according to the relationship that the phase difference map is proportional to the echo time, an estimated value of the phase difference map to be calibrated is calculated based on the echo time and the reference phase difference map; using the estimated value, the phase difference map to be calibrated is unwrapped according to the phase periodicity to obtain a calibrated phase difference map, According to the phase difference map, a temperature difference map corresponding to at least two different echo times is obtained, According to the temperature difference map, a temperature map corresponding to at least two different echo times is obtained, A weighted temperature map is obtained using the temperature maps corresponding to at least two different echo times.

2. The method according to claim 1, characterized in that The echo time corresponding to the reference phase difference map is smaller than the echo time corresponding to the calibrated phase difference map.

3. The method according to claim 1 or 2, characterized in that: Also included is the step of eliminating phase drift introduced by the magnetic resonance system.

4. The method according to claim 3, characterized in that The step of eliminating the phase drift caused by the magnetic resonance system is performed on a phase difference image or a temperature change image, and is specifically as follows: a number of areas with stable physical temperature and uniform tissue are selected as thermal reference points, and correction is performed by subtracting the average phase difference of the thermal reference points from each phase difference image or subtracting the average temperature change of the thermal reference points from the temperature change image.

5. The method according to claim 3, characterized in that: A magnetic susceptibility correction step is also included.

6. The method according to claim 5, characterized in that The step of correcting the magnetic susceptibility is performed on a phase difference map or a temperature difference map; The steps for performing magnetic susceptibility correction on the temperature difference map include: Obtain a first temperature map according to the reference phase difference map, and obtain a corresponding second temperature map according to the calibrated phase difference map; Determine whether the absolute value of the difference between the temperature value corresponding to each pixel in the second temperature map and the temperature value corresponding to the corresponding pixel in the first temperature map exceeds a preset temperature threshold, and if so, correct the temperature value corresponding to the corresponding pixel in the second temperature map; The steps for performing magnetic susceptibility correction on the phase difference map include: It is determined whether the absolute value of the difference between the phase difference value corresponding to each pixel in the calibrated phase difference map and the phase difference value corresponding to the corresponding pixel in the reference phase difference map exceeds a preset phase difference threshold; if so, the phase difference corresponding to the corresponding pixel in the calibrated phase difference map is corrected.

7. The method according to claim 5, characterized in that Also included is the step of correcting for motion induced phase errors.

8. The method according to claim 7, characterized in that The step of correcting the phase error caused by the motion is performed on a phase difference map or a temperature difference map; The steps performed on the phase difference map to correct the phase error caused by motion include: The phase error caused by motion is removed by linear least square fitting at each pixel of the phase difference map; The steps to correct motion-induced phase errors on the temperature difference map include: A linear least squares fit is used at each pixel in the temperature difference map to remove the error caused by motion.

9. The method according to any one of claims 1 or 2, characterized in that: A magnetic susceptibility correction step is also included.

10. The method according to claim 9, characterized in that The step of correcting the magnetic susceptibility is performed on a phase difference map or a temperature difference map; The steps for performing magnetic susceptibility correction on the temperature difference map include: Obtain a first temperature map according to the reference phase difference map, and obtain a corresponding second temperature map according to the calibrated phase difference map; Determine whether the absolute value of the difference between the temperature value corresponding to each pixel in the second temperature map and the temperature value corresponding to the corresponding pixel in the first temperature map exceeds a preset temperature threshold, and if so, correct the temperature value corresponding to the corresponding pixel in the second temperature map; The steps for performing magnetic susceptibility correction on the phase difference map include: It is determined whether the absolute value of the difference between the phase difference value corresponding to each pixel in the calibrated phase difference map and the phase difference value corresponding to the corresponding pixel in the reference phase difference map exceeds a preset phase difference threshold; if so, the phase difference corresponding to the corresponding pixel in the calibrated phase difference map is corrected.

11. The method according to claim 1 or 2, characterized in that: Also included is the step of correcting for motion induced phase errors.

12. The method according to claim 11, characterized in that The step of correcting the phase error caused by the motion is performed on a phase difference map or a temperature difference map; The steps performed on the phase difference map to correct the phase error caused by motion include: The phase error caused by motion is removed by linear least square fitting at each pixel of the phase difference map; The steps to correct motion-induced phase errors on the temperature difference map include: A linear least squares fit is used at each pixel in the temperature difference map to remove the error caused by motion.

13. A magnetic resonance temperature imaging system, characterized in that: include: Memory and processor; The memory is used to store program codes, the processor is used to call the program codes, and the program codes are used to execute the magnetic resonance temperature imaging method according to any one of claims 1 to 12.

Images