System and method for correcting measurement artifacts in magnetic resonance thermometry

Abstract

A system and method for performing magnetic resonance (MR) thermometry includes a magnetic resonance imaging (MRI) unit (100) and a controller (124) in communication with the MRI unit (100). The controller is configured to: cause the MRI unit (100) to acquire at least one baseline phase image of an imaging region (110) and at least one treatment phase image of the imaging region (110) after a temperature change of a subregion within the imaging region (110); electronically generate a thermal map based at least in part on the acquired baseline phase image and treatment phase image, the thermal map indicating the temperature change of the subregion pixel by pixel; computationally predict the temperature change of the subregion based at least in part on energy accumulated in the subregion during treatment without reference to the generated thermal map; and determine whether the thermal map is inaccurate based at least in part on the temperature change of the subregion indicated by the thermal map and the predicted temperature change of the subregion.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to and the benefit of U.S. Patent Application No. 16 / 202,558, filed on November 28, 2018, and the entire contents of that U.S. Patent Application are incorporated herein by reference. Technical Field

[0003] The present invention relates generally to magnetic resonance (MR) imaging and, more particularly, to techniques for MR thermography (or MR thermometry) and corrections in MR thermometry. Background Art

[0004] MR imaging of tissues in the body can be used for many medical procedures including diagnosis and surgery. In general, MR imaging begins by placing the subject in a relatively uniform static magnetic field. The static magnetic field causes the nuclear spins of hydrogen atoms to align and precess in the general direction of the magnetic field. Radio frequency (RF) magnetic field pulses are then superimposed on the static magnetic field to cause some of the aligned spins to alternate between temporarily high-energy non-aligned and aligned states, thereby inducing RF response signals known as MR echoes or MR response signals. It is known that different tissues in the subject generate different MR response signals, and this property can be used to create contrast in MR images. The RF receiver detects the duration, intensity, and source location of the MR response signal, and then processes the data to generate a tomographic image or a three-dimensional image.

[0005] MR imaging can also be effectively used during medical procedures to assist in positioning and guiding medical devices. For example, a medical procedure can be performed on a patient using a medical device while the patient is in an MRI machine. The medical device can be used to be inserted into the patient's body, or the medical device can be used externally but still has a therapeutic or diagnostic effect. For example, the medical instrument can be an ultrasound device that is arranged outside the patient's body and focuses ultrasound energy to ablate or destroy tissue or other material in or on the patient's body. The MRI machine preferably generates images at a high rate so that the position of the medical device relative to the patient (or the focus of the action of the medical device) can be monitored in real time (or substantially in real time).

[0006] MR imaging can also provide a non-invasive means of quantitatively monitoring body temperature. This is particularly useful in the above-mentioned MR-guided focused ultrasound (MRgFUS) treatment or other MR-guided thermal therapies, where the temperature of the treatment area should be continuously monitored during the above-mentioned treatments and therapies to assess the progress of the treatment and correct for local differences in heat conduction and energy absorption. Monitoring (e.g., measuring and / or mapping) temperature with MR imaging is generally referred to as MR thermometry or MR thermography.

[0007] Among the various methods available for MR thermometry, the proton resonance frequency (PRF) shift method is generally preferred because of its excellent linearity in temperature changes, close independence from tissue type, and high spatial and temporal resolution of the obtained temperature map. The PRF shift method is based on the following phenomenon: the MR resonance frequency of protons in water molecules varies linearly with temperature (with a proportionality constant that is advantageously relatively constant across tissue types). Since the frequency variation with temperature is small, only -0.01 ppm / °C for gravity water and about -0.0096 ppm / °C to -0.013 ppm / °C in tissue, a phase-sensitive imaging method is generally used to detect the PRF shift; in this phase-sensitive imaging method, imaging is performed twice: first, a baseline PRF phase image is acquired before the temperature change, and then a second phase image is acquired after the temperature change, thereby capturing small phase changes proportional to the temperature change.

[0008] For example, a phase image may be calculated from the MR image data, and a temperature difference map relative to a baseline image may be obtained by: (i) determining, on a pixel-by-pixel basis, a phase difference between a phase image corresponding to the baseline and a phase image corresponding to a subsequently acquired MR image; and (ii) converting the phase difference to a temperature difference based on the PRP temperature dependence while taking into account imaging parameters such as static magnetic field strength and echo time (TE). It should be understood that the determination of the phase difference involves more than a simple subtraction of a scalar quantity, although a subtraction step may be involved.

[0009] Unfortunately, changes in phase images are not uniquely caused by temperature changes. A variety of factors unrelated to temperature (e.g., changes in the local magnetic field due to nearby moving objects, changes in magnetic susceptibility within the patient's body due to breathing or other movement, and magnet or shim drift) may result in confusing phase shifts that invalidate phase-sensitive temperature measurements. For example, during an MRgFUS treatment procedure, one or more treatment devices may need to be repositioned and / or reoriented in or near the MR imaging area. Since treatment devices typically include metal components, their movement may disturb the local magnetic field and thereby significantly change the phase background. Non-metallic objects and their movement may also interfere with the local magnetic field. For example, the patient's breathing or rotational motion may have a similar effect on MR imaging data. In fact, magnetic field changes associated with patient motion and / or nearby objects may be severe enough to invalidate temperature measurements made using the above-described phase-sensitive approach.

[0010] In order to detect phase changes caused by factors unrelated to temperature, various conventional approaches create real-space pixel images of MR imaging data after acquiring the MR imaging data and identify artifacts that appear in the pixel images. Based on the detected artifacts, phase changes caused by factors unrelated to temperature are indirectly inferred. However, artifacts that have little effect on the pixel image may have a significant effect on the phase image. As a result, conventional approaches may still generate flawed thermal maps, which may compromise medical treatment.

[0011] Therefore, there is a need to accurately and reliably identify erroneous MR thermal images caused by factors unrelated to temperature to ensure effective and safe medical procedures. Summary of the invention

[0012] Various embodiments of the present invention provide systems and methods for detecting inaccurate temperature maps generated from MR imaging data. For ease of reference, the following description relates to MR imaging data acquired during ultrasound thermal therapy; however, it should be understood that the same approach is generally applicable to any MR-guided medical procedure (including diagnosis and surgery) that requires continuous temperature monitoring of a region of interest (e.g., for evaluating the progress of the procedure).

[0013] In some embodiments, prior to thermal treatment, MR raw imaging data is acquired; the raw data is then processed to identify the location and / or orientation of the target region and to generate a baseline phase image. MR imaging data may be acquired again during thermal treatment and processed to identify the location of the target tissue and to generate a treatment phase image. The treatment phase image may then be compared on a pixel-by-pixel basis with the baseline phase image acquired prior to treatment to calculate a phase difference between the two; based on the calculated phase difference, an MR thermal map may be created that indicates point-by-point changes in temperature measured by the MR imaging data. In a number of embodiments, some or all of the measured pixel-by-pixel temperature changes in the thermal map are compared (again on a pixel-by-pixel basis) with temperature changes predicted using a physical model; based on the deviation between the measured temperature changes and the predicted temperature changes, the accuracy of the acquired thermal map may be determined. For example, if the deviation exceeds a predetermined threshold amount (for a single pixel or over a region of pixels, such as on an aggregated basis), the acquired thermal map may be identified as inaccurate. As a result, the acquired thermal map may be discarded, and in some embodiments, ultrasound therapy may be suspended until an accurate thermal map is obtained, thereby avoiding damage to non-target tissue.

[0014] The physical model can predict the temperature change caused by thermal treatment in a computational manner based on, for example, the acoustic energy accumulated in the target area and / or non-target area, and tissue characteristics such as anatomical characteristics (e.g., type, nature, structure, thickness, density, etc.), and / or material characteristics of the target area and / or non-target area (e.g., sound velocity). The acoustic energy accumulated in the target area and / or non-target area can be estimated based on the ultrasound parameter values ​​that generate the focal zone at the target area and the tissue characteristics of the intervening tissue on the beam path between the transducer and the target area. In one embodiment, the tissue characteristics of the target tissue and non-target tissue (including the intervening tissue and the tissue surrounding the target area) are acquired using an imaging device, such as an MRI device, a computed tomography (CT) device, a positron emission tomography (PET) device, a single photon emission computed tomography (SPECT) device, or an ultrasound examination device. Additionally, the physical model may further take the form of (or include) differential equations (such as the Pennes model and biothermal equations) to simulate heat transfer in tissue, thereby predicting the temperature increase in the target area / non-target area during the time interval.

[0015] Alternatively or additionally, a statistical model may be used to predict the temperature increase caused by thermal treatment. For example, the statistical model may include historical data of the cumulative acoustic energy or temperature increase measured during previous thermal treatments performed on the same patient or on different patients. Based on retrospective studies, a statistical model may be established that relates the transmitted acoustic power and tissue characteristics to the cumulative acoustic energy or temperature changes at the target area / non-target area. The tissue characteristics of the current patient and the ultrasound parameter values ​​used in the current treatment may then be applied to the statistical model to predict the accumulated acoustic energy or temperature increase during treatment at a given time or within a time interval.

[0016] The predetermined threshold(s) used to decide whether the temperature increase in the thermal map is due to the tissue response to the thermal treatment or due to some unrelated artifact may be fixed or dynamically variable. For example, the size of the threshold may be positively correlated with the amount of acoustic energy transmitted to the target area, so that for smaller acoustic energies, the threshold is smaller; and for larger acoustic energies, the threshold is larger. As a result, at higher acoustic energies, a larger difference between the measured temperature and the predicted temperature is required to conclude that the measured thermal map is inaccurate. In one embodiment, the predetermined threshold is adjusted based on the noise level of the acquired MR imaging data. For example, MR imaging data with a smaller signal-to-noise ratio may correspond to a larger threshold than MR imaging data with a larger signal-to-noise ratio. Therefore, when the thermal map includes a higher noise level, the measured temperature in the thermal map may have a larger deviation from the predicted temperature, and then the measured thermal map is determined to be defective.

[0017] To detect inaccurate thermal maps, it may not be necessary to predict temperature changes during thermal treatment. In some embodiments, detection of inaccurate thermal maps is based solely on historical imaging data acquired during thermal treatment. For example, assuming that the transmitted ultrasound power remains constant during treatment, the accumulated energy (and resulting temperature) at the target area can be expected to gradually increase over time. Therefore, if a target area or non-target area in a particular temperature map shows a sudden increase or decrease in temperature (for example, compared to an average increase or decrease in the same area in several previous images), the temperature map may be incorrect at the specified area.

[0018] Thus, multiple embodiments provide a means for real-time monitoring of the in vivo temperature of target tissue and / or non-target tissue during thermal treatment and for detecting inaccurate thermal maps in real-time. This is particularly useful in MR-guided thermal therapy, whereby the temperature of target tissue and / or non-target tissue can be continuously monitored to assess the progress of thermal treatment and correct for local differences in heat conduction and energy absorption, thereby achieving the desired treatment effect on the target and avoiding damage to non-target tissue.

[0019] Thus, in one aspect, the present invention relates to a system for performing magnetic resonance (MR) thermometry. In various embodiments, the system comprises a magnetic resonance imaging (MRI) unit and a controller in communication with the MRI unit. The controller is configured to: (i) cause the MRI unit to acquire at least one baseline phase image of an imaging region and at least one treatment phase image of the imaging region after a temperature change of a sub-region within the imaging region; (ii) electronically generate a thermal map based at least in part on the acquired baseline phase image(s) and treatment phase image(s), the thermal map indicating the temperature change of the sub-region pixel by pixel; (iii) computationally predict the temperature change of the sub-region based at least in part on the energy accumulated in the sub-region during treatment without reference to the generated thermal map; and (iv) determine whether the thermal map is inaccurate based at least in part on the temperature change of the sub-region indicated by the thermal map and the predicted temperature change of the sub-region.

[0020] The controller may be further configured to compare the temperature change in the generated thermal map with the predicted temperature change to determine a deviation therebetween; and to compare the deviation with a predetermined threshold. In addition, the controller may be further configured to determine that the thermal map is inaccurate upon determining that the deviation exceeds the predetermined threshold (which may or may not be a fixed value). In some embodiments, the controller is further configured to adjust the predetermined threshold based at least in part on the energy accumulated in the sub-region, the noise level associated with the baseline phase image(s) and / or the treatment phase image(s), or the deviation between the temperature change in the generated thermal map and the predicted temperature change.

[0021] In one embodiment, a medical device is also included, the medical device being configured to cause a temperature change in the sub-region. For example, the medical device may include an ultrasound transducer having a plurality of transducer elements. The controller may also be configured to use a physical model to computationally predict the temperature change in the sub-region. In one embodiment, the physical model is based at least in part on the value of an ultrasound parameter (e.g., an amplitude, frequency, phase, direction, or activation time associated with the transducer element) used to generate a focal zone at the sub-region.

[0022] In a number of embodiments, the controller is further configured to use a physical model to computationally predict the temperature change of the sub-region without reference to the generated thermal map. The physical model is based at least in part on tissue properties (e.g., type, structure, thickness, density, speed of sound, heat absorption coefficient, perfusion coefficient, and / or metabolic heat generation rate) associated with at least one of the sub-region or a second sub-region different from the sub-region. In one embodiment, the controller is further configured to obtain the tissue properties based at least in part on imaging data acquired using the MRI unit. In addition, the physical model can be based on a biological heat transfer equation (e.g., Pennes equation).

[0023] The controller may also be configured to predict the temperature change of the sub-region using a statistical model. In addition, the system may further include a medical device configured to cause a temperature change in the sub-region, and the statistical model may then include historical data of temperature changes caused by previous activations of the medical device. In one embodiment, the controller is further configured to cause the MRI unit to acquire a reference library, the reference library comprising a plurality of baseline phase images of the imaging region, each of the baseline phase images corresponding to a phase background during a different phase of expected movement of the imaging region. The controller may then be further configured to identify a baseline phase image in the reference library that best matches the treatment phase image based on a similarity between the baseline phase image and the treatment phase image, and generate the thermal map based at least in part on the identified best matching baseline phase image.

[0024] In another aspect, the present invention relates to a method of performing MR thermometry. In various embodiments, the method comprises: acquiring at least one baseline phase image of an imaging region and at least one treatment phase image of the imaging region after a temperature change of a subregion within the imaging region; electronically generating a thermal map based at least in part on the acquired baseline phase image(s) and treatment phase image(s), the thermal map indicating the temperature change of the subregion pixel by pixel; computationally predicting the temperature change of the subregion based at least in part on energy accumulated in the subregion during treatment without reference to the generated thermal map; and determining whether the thermal map is inaccurate based at least in part on the temperature change of the subregion indicated by the thermal map and the predicted temperature change of the subregion.

[0025] Another aspect of the present invention relates to a system for performing MR thermometry. In various embodiments, the system includes: a magnetic resonance imaging (MRI) unit; and a controller in communication with the MRI unit; the controller being configured to: (i) cause the MRI unit to acquire one or more baseline phase images of an imaging region and a plurality of treatment phase images of the imaging region after a temperature change of at least a sub-region within the imaging region; (ii) electronically generate a plurality of thermal maps based at least in part on the acquired baseline phase image(s) and the treatment phase image, each of the thermal maps indicating, pixel by pixel, a temperature change of the sub-region associated with one of the treatment phase images; and (iii) determine whether one of the thermal maps is inaccurate based at least in part on a comparison between a temperature change associated with one of the thermal maps and a temperature change associated with at least one other of the thermal maps.

[0026] As used herein, the term "substantially" refers to ±10%, and in some embodiments ±5%. Throughout the specification, references to "an example," "an example," "an embodiment," or "an embodiment" refer to specific features, structures, or characteristics described in conjunction with the example being included in at least one example of the present technology. Therefore, the phrases "in an example," "in an example," "an embodiment," or "an embodiment" appearing in multiple places throughout the specification do not necessarily all refer to the same example. In addition, specific features, structures, routines, steps, or characteristics may be combined in any suitable manner in one or more examples of the technology. The titles provided herein are for convenience only and are not intended to limit or interpret the scope or meaning of the claimed technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In the accompanying drawings, like reference numerals generally refer to the same parts in different views. Moreover, the drawings are not necessarily drawn to scale, but emphasis is generally placed on illustrating the principles of the invention. In the following description, various embodiments of the invention will be described with reference to the following drawings, in which:

[0028] Figure 1 An exemplary MRI apparatus according to various embodiments of the present invention is shown;

[0029] Figure 2A and 2B is a flow chart illustrating an exemplary method for detecting an inaccurate MR thermogram in which a temperature increase is caused by a temperature-related factor according to various embodiments of the present invention;

[0030] FIG. 3A to FIG. 3D Depicted are exemplary temperature deviation graphs illustrating temperature differences between measured and predicted thermal maps at target areas and / or non-target areas according to various embodiments of the present invention.

[0031] Figure 4 is a flow chart illustrating an exemplary method of predicting temperature changes in target / non-target areas during a medical procedure according to various embodiments of the present invention;

[0032] Figure 5A shows the relationship between temperature change at a target area and application time of acoustic energy according to various embodiments of the present invention;

[0033] Figure 5B shows the relationship between the temperature change at the target area and the amplitude of the applied ultrasonic wave according to various embodiments of the present invention;

[0034] Figure 6is a flow chart illustrating another exemplary method for detecting an inaccurate MR thermal map according to various embodiments of the present invention, wherein the temperature increase is caused by factors unrelated to temperature; and

[0035] Figure 7 Various relationships between temperature changes at target areas and non-target areas and application time of acoustic energy according to various embodiments of the present invention are depicted. DETAILED DESCRIPTION

[0036] Figure 1 An exemplary MRI system is shown in which techniques for performing MR thermometry and for detecting measurement artifacts in MR thermometry according to various embodiments of the present invention may be implemented or used. The illustrated MRI system 100 includes an MRI machine 102. If an MR-guided procedure is being performed, a medical device (e.g., an ultrasound transducer) 104 may be disposed within the bore of the MRI machine 102. Because the components and operation of MRI machines are well known in the art, only some basic components that are helpful in understanding the system 100 and the operation of the system are described herein.

[0037] The MRI machine 102 generally includes a cylindrical electromagnet 106 that generates a static magnetic field within a bore 108 of the electromagnet 106. The electromagnet 106 generates a substantially uniform magnetic field within an imaging region 110 inside the magnet bore 108. The electromagnet 106 may be enclosed in a magnet housing 112. A support table 114 on which a patient 116 lies is disposed within the magnet bore 108. A region of interest 118 within the patient 116 may be identified and located within the imaging region 110 of the MRI machine 102.

[0038] A set of cylindrical magnetic field gradient coils 120 may also be disposed within the magnet bore 108. The gradient coils 120 also surround the patient 116. The gradient coils 120 may generate magnetic field gradients of predetermined magnitude at predetermined times and in three mutually orthogonal directions within the magnet bore 108. Using the field gradients, different spatial locations may be associated with different precession frequencies, thereby imparting spatial resolution to the MR image. An RF transmitter coil 122 surrounds the imaging region 110 and the region of interest 118. The RF transmitter coil 122 transmits RF energy in the form of a magnetic field into the imaging region 110, including into the region of interest 118.

[0039] The RF transmitter coil 122 may also receive MR response signals transmitted from the region of interest 118. As known to those of ordinary skill in the art, the MR response signals are amplified, modulated, and digitized into raw k-space data using a controller 124. The controller 124 also processes the raw k-space data into an array of image data using known computational methods including Fast Fourier Transform (FFT). The image data may then be displayed on a monitor 126, such as a computer CRT, LCD display, or other suitable display.

[0040] In a typical MR imaging procedure, the emission of RF excitation pulses, the application of field gradients in multiple directions, and the acquisition of RF response signals are performed in a predetermined sequence. For example, in some imaging sequences, a linear field gradient parallel to the static magnetic field is applied simultaneously with the excitation pulse to select slices within a three-dimensional tissue for imaging. Subsequently, a time-dependent gradient parallel to the imaging plane can be used to impart a position-dependent phase and frequency to the magnetization vector. Alternatively, an imaging sequence can be designed for a three-dimensional imaging region. Time sequences suitable for PRF temperature measurement include, for example, gradient echo (GRE) sequences and spin echo sequences.

[0041] The time-varying RF response signal is integrated over the entire (two-dimensional or three-dimensional) imaging area, and the time-varying RF response signal is sampled to generate a time sequence of response signals that constitute the raw image data. Each data point in the time sequence can be interpreted as a Fourier transform value of the local magnetization intensity associated with the position at a specific point in k-space, where k is a function of the time development of the gradient field. Thus, by acquiring the time sequence of the response signal and performing a Fourier transform on the time sequence, a real-space image of the tissue (i.e., an image showing the effect of the measured magnetization intensity on the properties of the tissue as a function of the spatial coordinates) can be reconstructed from the raw data. Computational methods (including, for example, fast Fourier transforms) for constructing real-space image data from raw data are generally known to those skilled in the art and can be easily implemented in hardware, software, or a combination of both hardware and software in the controller 124 without the need for experimentation.

[0042] In the presence of ultrasound-induced temperature changes, hot spots may appear in the phase of the image data because the resonant frequency of water protons decreases as temperature increases. Therefore, for the purpose of PRF temperature measurement, the controller 124 also includes the following functions: extracting phase information from the real-space image data, and calculating the real-space map of the temperature-induced phase shift based on images acquired before heating the target tissue and after heating the target tissue (or during heating the target tissue) (i.e., baseline images and treatment images). From the phase shift map, a temperature change map (in units of Δ°C) can be calculated by multiplying with a constant c, which is given by the following formula:

[0043]

[0044] where α is the applicable PRF variation coefficient (for hydrated tissue, α is -0.01 ppm / °C), γ is the proton gyromagnetic ratio, B0 is the primary magnetic field strength, and TE is the echo time of the GRE or other imaging sequence.

[0045] The medical device 104 may also be placed in or near the imaging area 110 of the MRI machine 102. Figure 1 In the example shown, the medical device 104 can be an ultrasound instrument used to ablate tissue such as fibroid tissue or cancerous tissue (or non-cancerous tissue), to destroy blockages within blood vessels, to open the blood-brain barrier, or to perform other treatments on tissue in or on the body of the patient 116. In fact, the medical device 104 can be any type of medical instrument, such as a needle, catheter, guide wire, radiation emitter, endoscope, laparoscope, or other instrument. In addition, the medical device 104 can be configured to be placed outside the patient 116 or inserted into the patient's body.

[0046] During MR thermography of region 110 (or any medical procedure involving MR temperature mapping), region of interest 118, which is typically a portion of a patient's body, may change its shape and / or position due to movement of the patient's body. Figure 1 In the embodiment of the present invention, the region of interest 118 is the patient's head, which may rotate slightly to the left or right during the thermal imaging process. If the region of interest 118 is a portion of the patient's abdominal region, its shape may shrink or expand with the patient's breathing cycle. Changes in the shape and / or position of the region of interest 118 may disturb the magnetic field, thereby changing the phase associated with the MR imaging data; as a result, the thermal map generated thereby may be inaccurate.

[0047] Similarly, in a medical procedure involving MR temperature mapping of region 110, medical device 104 may be repositioned and / or reoriented one or more times according to a dynamic protocol. Movement of medical device 104 resulting from repositioning and / or reorientation may change the magnetic field and thereby change the phase of the MR imaging data, which in turn may result in inaccurate thermal maps.

[0048] The present invention provides multiple approaches for detecting inaccurate MR thermal maps caused by factors unrelated to temperature (e.g., movement of the patient or movement of nearby objects) during a medical procedure (e.g., ultrasound treatment). These approaches generally involve: using MR thermometry to monitor the temperature at the region of interest 118 before and during the medical procedure; and predicting the temperature increase caused by the procedure by computational means. If the measured temperature increase (for a single pixel, or in a region with clustered pixels) exceeds the temperature increase predicted by computational means by a predetermined threshold amount, then the temperature map corresponding to such pixel or in such region in the thermal map acquired at a later time may be inaccurate, that is, the temperature increase for such pixel or in such region is caused by some unrelated artifact, rather than by a true tissue response to the medical procedure.

[0049] Figure 2A 2 is a flow chart illustrating an exemplary approach 200 for detecting inaccurate MR thermograms in which temperature increases during a medical procedure are (at least in part) derived from factors unrelated to temperature, according to various embodiments. In a first step 202, prior to a medical procedure (e.g., thermal therapy), an MR imaging sequence is performed to acquire response signals from an imaging region 110, which are subsequently converted to raw image data (i.e., "k-space data"). In a second step 204, the raw image data is converted (using a fast Fourier transform) to a real-space MR image of the imaging region; a PRF baseline phase image associated with the real-space image can then be generated, and a target tissue in the real-space image (the target tissue corresponding to the ROI 118) can be selected. In some embodiments, the selection is manual, i.e., based on user input (e.g., a line drawn with a mouse to circle a target in the image), while in other embodiments, the selection is automatically performed by a computer algorithm (e.g., a conventional algorithm that thresholds pixel values ​​using contrast between the target and surrounding tissue in the MR image). Optionally, steps 202, 204 may be repeated multiple times, for example, at different phases during a periodic motion cycle (eg, a cardiac or respiratory cycle), to create a reference library having a series of baseline reference images.

[0050] In a third step 206, one or more medical devices associated with the procedure (e.g., an ultrasonic transducer 104 for thermal therapy) may be activated to treat the target tissue. During treatment, raw image data of the target area is acquired using the MRI device 100 as described above (step 208). Again, the raw treatment image may be converted into a real space image and processed to identify the location of the target tissue and generate a PRF treatment phase image (step 210). In step 212, the PRF treatment phase image is compared with a PRF baseline phase image acquired before thermal therapy, and the comparison is performed on a pixel-by-pixel basis to calculate the phase difference between the PRF treatment phase image and the PRF baseline phase image; based on the calculated phase difference, an MR thermal map associated with the treatment image in the imaging area may be created. Optionally, steps 208 to 212 may be repeated for monitoring the in vivo temperature of the target tissue and / or non-target tissue during the medical procedure. This is particularly useful in MR-guided thermal therapy (e.g., MRgFUS therapy), in which the temperature of the target tissue and / or non-target tissue is continuously monitored in order to assess the progress of the thermal therapy and correct for local differences in heat conduction and energy absorption.

[0051] As described above, if a reference library of baseline images covering the expected range of motion is obtained, a reference baseline image in the reference library that best matches the acquired treatment image can be selected based on the similarity between the reference baseline image and the treatment image. The selected baseline image and treatment image are then processed to generate a thermal map that shows temperature changes in target / non-target areas. This approach is generally referred to as multi-baseline thermometry; an exemplary approach for performing multi-baseline thermometry is described in U.S. Patent No. 9,814,909, the entire disclosure of which is incorporated herein by reference.

[0052] To determine whether the acquired thermal map is inaccurate, in various embodiments, as further described below, the thermal map generated from the MRI measurements in step 212 may be compared to a thermal map predicted using a physical model (step 214). If the deviation between the measured thermal map and the predicted thermal map exceeds a predetermined threshold amount ΔT, the thermal map is compared to the predicted thermal map. th If the thermal map is not accurately determined (for a single pixel, or in an area where pixel values ​​are clustered), the thermal map is considered inaccurate (step 216). The inaccurate thermal map can then be discarded, and new MR imaging data can be acquired to generate a new thermal map (step 218). Additionally or alternatively, the medical device 104 can be paused until an accurate thermal map is generated, thereby avoiding damage to non-target tissue (step 220). In contrast, if the deviation between the measured thermal map and the predicted thermal map is equal to or below a predetermined threshold, the thermal map acquired in step 212 is considered accurate (step 222).

[0053] For example, refer to Figure 3A , the controller 124 can compare the phase difference between the phase image corresponding to the baseline image 302 and the phase image corresponding to the treatment image 304, and based on this, convert the phase difference into a temperature difference in the thermal map 306. In addition, in the thermal map 306, the target area 308 includes target pixels T1-T3, and the non-target area 310 surrounding the target area includes non-target pixels NT1-NT7. In some embodiments, the controller further creates a thermal map 312, which indicates the predicted temperature increase caused by the thermal treatment, as further described below. The difference between the measured thermal map 306 and the predicted thermal map 312 (as shown in the deviation map 314) can then be determined on a pixel-by-pixel basis, and the difference is compared with a predetermined threshold. For example, the predetermined threshold of the difference corresponding to each pixel T1-T3 in the target area and the predetermined threshold of the difference corresponding to each pixel NT1-NT7 in the non-target area are 0.5°C and 0.1°C, respectively, and the threshold of the deviation in the target area with clustered pixels and the threshold of the deviation in the non-target area with clustered pixels are 1.2°C and 0.5°C, respectively. refer to Figure 3B , because the deviation between the measured temperature and the predicted temperature of each pixel in the target pixels T1-T3 and the non-target pixels NT1-NT7 is less than the predetermined threshold, and the aggregated temperature deviation of the target pixels and the aggregated temperature deviation of the non-target pixels are 0.4°C and 0.1°C respectively (both are less than the predetermined threshold), the measured thermal map 306 is considered to be accurate.

[0054] In contrast, when the deviation between the measured heat map and the predicted heat map of individual pixels and / or aggregated pixels in the target area and / or non-target area exceeds a predetermined threshold, the heat map 306 is determined to be inaccurate. Figure 3C , because the temperature deviation at the target pixel T1 in the heat map 324 exceeds the predetermined threshold (0.5° C.), the heat map 306 is considered inaccurate. Figure 3D , the deviation between the measured thermal map and the predicted thermal map is depicted in the thermal map 334. Although each of the target pixels T1-T3 and non-target pixels NT1-NT7 does not exceed the predetermined thresholds (0.5°C and 0.1°C, respectively), the aggregate temperature difference in the target pixels is 1.5°C, which exceeds the predetermined aggregate threshold of 1.2°C; as a result, the thermal map 306 is considered to be inaccurate.

[0055] After generating the thermal map 306, the processing time for determining the temperature change, comparing the measured temperature change with the predicted value to determine the deviation between the two, and determining whether the deviation exceeds a predetermined threshold is relatively fast (compared to acquiring MR imaging data). Therefore, the above-described approach can advantageously determine the accuracy of a newly acquired thermal map in real time (or substantially real time) during a medical procedure.

[0056] Figure 2B is a flow chart illustrating another method 250 for detecting inaccurate MR thermal images caused by one or more factors unrelated to temperature during a medical procedure in accordance with various embodiments. Figure 2A In the method 200 shown in FIG. 250 , steps 202 to 212 are performed. However, in this method, two thermal images acquired at different times during the medical procedure are analyzed on a pixel-by-pixel basis to determine the temperature difference ΔT between the two thermal images. m (Step 252); then the calculated difference and the predicted temperature rise ΔT p Compare (step 254). If the calculated temperature difference ΔT m Exceeding the predicted temperature rise ΔT p exceeds a predetermined threshold value ΔT th If the calculated temperature difference between the two thermal maps does not deviate significantly from the predicted temperature increase (i.e., the calculated temperature difference is below a predetermined threshold amount for the clustered pixel or the single pixel, |ΔT m -ΔT p |≤ΔT th ), then the thermal map acquired at a later time is considered accurate (step 258).

[0057] In multiple embodiments, based on tissue characteristics of the target area and / or non-target area, and the energy accumulated in the target area and / or non-target area during a relevant time interval Δt (e.g., from the time of starting the thermal treatment to the time of acquiring the thermal map, or from t=t1 to t=t2), the temperature rise at a given time t=t1 during the medical procedure and the temperature rise between two thermal maps acquired at times t=t1 and t=t2 are predicted. Figure 4 Depicted are exemplary approaches for predicting the temperature rise caused by thermal treatment according to various embodiments. In a typical ultrasound treatment, when using the above approach (e.g., Figure 2AAfter determining the location and / or orientation of the target region (in step 204 in the example), ultrasound parameter values ​​(e.g., amplitude, frequency, phase, and / or orientation associated with a transducer element, or a time interval between consecutive sequences of sonication) may be calculated to create a focal zone at the target region (in step 406). This step typically involves applying a physical model and taking into account the geometry and location and orientation of the ultrasound transducer relative to the target region. In addition, tissue properties such as anatomical properties (e.g., type, nature, structure, thickness, density, etc.), and / or material properties of intervening tissue located in the beam path between the transducer and the target region (e.g., speed of sound) may be included in the physical model to predict and correct beam aberrations generated thereby. In one embodiment, a MRI apparatus such as 100 (e.g., FIG. 10A ) is used to perform a beam path analysis of the target region. Figure 1 The anatomical properties of the intervening tissue can be acquired by an imaging device such as a CT device, a PET device, a SPECT device or an ultrasound device. For example, based on the acquired image, a tissue model that characterizes the material properties of the intervening tissue can be established. The tissue model generally includes a plurality of tissue types or layers (for example, for focusing ultrasound into the skull, a layer of cortical bone, bone marrow and soft brain tissue), and the tissue model characterizes the anatomical and / or material properties of each of the plurality of tissue types or layers. The tissue model can take the form of a 3D table of cells corresponding to voxels representing target tissue and / or non-target tissue; the cells have properties whose values ​​represent the properties of the tissue (for example, the speed of sound), which are related to the aberrations that occur when the beam passes through the tissue. The voxels are obtained by tomography by an imaging device, and the tissue type represented by each voxel can be automatically determined by conventional tissue analysis software. The cells of the tissue model can be filled using a lookup table of determined tissue types and tissue parameters (for example, by the speed of sound of the tissue type). Further details regarding the creation of tissue models that identify the acoustic velocity, thermal sensitivity, and / or thermal energy tolerance of various tissues can be found in U.S. Patent Publication No. 2012 / 0029396, the entire disclosure of which is incorporated herein by reference.

[0058] The acoustic power of the beam in the focal zone is (at least partially) absorbed by the target tissue, thereby generating heat and raising the temperature of the tissue to the point where cells are denatured and / or ablated. The degree of ultrasound absorption over the propagation length in the tissue is a function of frequency and is given by:

[0059] P t =P0×(1-10- 2αfz )10 -2∝f

[0060] Where P0 represents the initial acoustic power of the ultrasonic beam emitted from the transducer; f represents the emission frequency of the ultrasonic wave (in MHz); α represents the absorption coefficient in the relevant frequency range (in cm -1 MHz -1 z represents the focal length, ie, the distance the ultrasound beam propagates through the tissue before reaching the target, in cm; and P t represents the acoustic power at the target area. Therefore, in various embodiments, the controller 124 processes the acquired images to further characterize the anatomical properties and / or material properties of the target tissue and / or non-target tissue, and includes these anatomical properties and / or material properties in the tissue model (in step 408). For example, the 3D table of cells in the tissue model may further include the following attributes, the values ​​of which represent the absorption coefficients associated with the target tissue / non-target tissue.

[0061] Therefore, based on the anatomical properties and / or material properties of the target tissue / non-target tissue represented by the tissue model and the adopted ultrasound parameter values, the physical model can predict the ultrasound beam path, the propagation of the induced effect in the tissue, the ultrasound energy transmitted to the target area and / or non-target area during the time interval Δt, and the conversion of ultrasound energy or pressure into heat at the target area and / or non-target area (in step 410). In some embodiments, the computational physics model further takes the form of (or includes) differential equations (such as the Pennes model and the biothermal equation) to simulate heat transfer in the tissue, thereby predicting the temperature increase in the target area / non-target area during the time interval Δt (step 412).

[0062] In general, the Pennes model is based on the following assumptions: The heat transfer rate between blood and tissues h b and blood perfusion rate W b (kg / (sm 3 ) and arterial blood temperature T a Proportional to the product of the difference between the local tissue temperature T(x, y, z): h b =W b C b (T a -T), where C b is the specific heat of blood in J / (K kg). The Pennes equation adds the contribution of heat transfer due to thermal conduction in the tissue and takes into account the rate Q m (in J / (sm 3 ), so the Pennes equation expresses the thermal energy balance of the perfused tissue in the following form:

[0063]

[0064] where ρ, C, and k are the density, heat capacity, and thermal conductivity of the tissue, respectively (in J / (sm K)), and Q ext Represents the thermal power extracted from each unit volume of tissue in thermal treatment. Therefore, by numerically solving the Pennes equation using any of a variety of methods known to those skilled in the art (such as, finite difference method and finite element method), a temperature map at a given time point can be calculated. Therefore, a thermal map can be determined that indicates the temperature change at a given time after the application of thermal treatment or the temperature change between two times t=t1 and t=t2. For example, U.S. Patent Publication Nos. 2012 / 0071746 and 2015 / 0359603 provide methods for computationally predicting the temperature increase during ultrasound treatment, and the entire disclosure of the patents is incorporated herein by reference.

[0065] Alternatively or additionally, a statistical model may be used to predict the temperature changes resulting from thermal treatment. For example, the statistical model may include historical data of the accumulated acoustic energy or temperature increase during treatment intervals Δt previously performed on the same patient or on different patients. In one embodiment, MR images acquired of the same type of target tissue and / or non-target tissue in previous thermal treatments are retrospectively studied to determine the heat absorbed in the target tissue / non-target tissue. In addition, the ultrasound parameter values ​​used for previous treatments are analyzed to determine the acoustic power transmitted to the target tissue / non-target tissue. Based on these retrospective studies, a statistical model that relates the transmitted acoustic power to the accumulated acoustic energy or temperature increase at the target area / non-target area can be directly established. The ultrasound parameter values ​​used in the current treatment can then be applied to the statistical model to predict the accumulated acoustic energy or temperature increase during the treatment interval Δt. For example, referring to Figure 5A , retrospective studies can illustrate that the temperature increase at the target area with type A tissue is positively correlated with the application time of the acoustic energy. The statistical model may therefore include a regression 502 performed on the measured temperature relative to the duration of the ultrasound application time; the regression 502 may then be applied to predict the temperature of the type A target tissue at ultrasound application times t=t1 and t=t2. Subsequently, the temperature increase ΔT in the target tissue from t=t0 (i.e., the time of starting ultrasound treatment) to t=t1 (or from t=t1 to t=t2) may be calculated. Similarly, referring to Figure 5B , retrospective studies may indicate that the temperature rise ΔT at the target area within the time interval Δt is positively correlated with the amplitude of the ultrasound waves. By performing a regression 504 of the measured temperature rise against the ultrasound amplitude, the temperature rise at the target tissue may be calculated based on the amplitude of the currently applied ultrasound treatment.

[0066] It should be noted that the approaches described herein for predicting the accumulated energy and / or temperature increase at the target area / non-target area are merely exemplary, and any suitable approach for predicting the accumulated energy and / or temperature increase during thermal therapy may be used in methods 200, 250 to detect inaccurate MR thermal maps as described above, and therefore the suitable approach falls within the scope of the present invention.

[0067] In addition, the predetermined threshold(s) used to determine whether the temperature rise in the thermal map is due to a tissue response to thermal treatment or due to some unrelated artifact (described in steps 216, 222, 256, 258) may be fixed or dynamically changing. Typically, the threshold(s) may represent a significant clinical effect on the target tissue / non-target tissue due to a medical procedure. As used herein, "significant clinical effect" refers to an undesirable (sometimes, lack of expectation) effect on the tissue that is considered significant by a clinician, such as damage to the tissue or the onset of other clinically adverse effects, whether temporary or permanent. In some embodiments, the threshold is determined based on the type, material properties and / or location of the target tissue / non-target tissue. For example, since the target tissue will be ablated in ultrasound treatment, the threshold for the temperature rise corresponding to the target tissue may be greater than the threshold for the temperature rise corresponding to the non-target tissue. In addition, if the non-target tissue adjacent to the target area is a sensitive and / or important organ, the risk of damaging the non-target organ is high, and the need for protecting sensitive / important non-target organs is increased. Thus, in this case, the predetermined threshold corresponding to the temperature increase in the non-target tissue may be less than the threshold in the case where non-sensitive and / or clinically insignificant non-target tissue surrounds the target area. Thus, in one embodiment, the threshold is predetermined by the controller 124, for example, based on the anatomical properties of the target tissue / non-target tissue acquired using the imaging device as described above, and / or a tissue model characterizing the material properties of the tissue / non-target tissue.

[0068] In some embodiments, the magnitude of the threshold is positively correlated with the amount of acoustic energy delivered to the target area, so that for relatively small acoustic energies, the threshold is small, and for relatively large acoustic energies, the threshold is large (e.g., increased by 10% or 20%). For example, during thermal therapy, the acoustic energy delivered to the target can be increased from E1 to E2 (E2 = E1 + ΔE); the threshold associated with each pixel in the target area can be dynamically increased from T1°C to T2°C (T2 = T1 + ΔT). As a result, at higher acoustic energies, a larger difference between the measured temperature and the predicted temperature is required to determine that the measured thermal map is inaccurate.

[0069] The threshold(s) may be adjusted based on other parameters related to the temperature measurement. For example, an MR response signal received in steps 202, 208 having a smaller signal-to-noise ratio (i.e., a higher noise level) may correspond to a larger threshold than an MR response signal having a larger signal-to-noise ratio (i.e., a lower noise level). Thus, if the thermal map has a higher noise level, a larger difference between the measured temperature and the predicted temperature is required to determine that the measured thermal map is defective. In some embodiments, the threshold(s) may be dynamically changed based on the difference between the measured temperature and the predicted temperature. For example, each measured thermal map in the target area has a temperature difference relative to the predicted thermal map; for example, the threshold may be statistically defined based on the average temperature difference, for example, 1 / 2 standard deviation or 1 standard deviation relative to the average temperature difference of the entire measured thermal map.

[0070] Figure 6 600 is a flow chart illustrating another exemplary approach 600 for detecting inaccurate MR thermograms in which an increase in temperature during a medical procedure is caused by factors unrelated to temperature, according to various embodiments. Figure 2A and Figure 2B In the method 200, 250 of FIG. 1 , steps 202-212 are performed in the pathway 600 to generate a thermal map indicating temperature changes caused by the treatment. However, the pathway 600 does not require a prediction of temperature rise as required in the method 200, 250. Instead, an incorrect thermal map is detected based on historical imaging data acquired during the medical procedure. For example, referring to Figure 7 , assuming that the transmitted ultrasound power remains constant during treatment, the accumulated energy (and therefore the temperature) at the target area can be expected to gradually increase over time (as shown by line 702); similarly, the temperature at the non-target area can remain constant (e.g., line 704), or increase over time, but with a slower rate of increase than the rate of temperature increase in the target area (e.g., line 706). Therefore, if the temperature of a target area or non-target area in a particular temperature map suddenly increases or decreases (e.g., compared to the average increase or average decrease in the same area in the previous few images), it indicates that the temperature map is incorrect at the specified area. For example, at time t=t3, the temperature at non-target area B shows a sudden increase 708, indicating that the thermal map at non-target area B is flawed. Similarly, a sudden decrease 710 in the temperature in the target area at time t=t5 indicates that the thermal map in the target area acquired at time t=t5 is inaccurate. Again, as Figure 7 As shown, the temperature evolution at the target tissue / non-target tissue during treatment is monitored based on single pixel and / or aggregate pixel values ​​associated with the target area / non-target area. Figure 6 After acquiring multiple thermal maps during thermal treatment, the controller may compare temperature changes indicated in a newly acquired thermal map with temperature changes measured in previously acquired thermal maps (step 602). If the newly acquired thermal map shows abrupt temperature changes, this indicates that the thermal map may be inaccurate (step 604). If no abrupt temperature changes are detected, this indicates that the newly acquired thermal map is accurate (step 606).

[0071] Although the present invention has been described with reference to monitoring the temperature at the target area and / or non-target area during a medical procedure (e.g., ultrasound thermal treatment) using MR thermometry, this arrangement is not intended to limit the scope of the present invention. For example, a temperature sensor can be implemented to measure the temperature during treatment. In addition, it should be understood that the features of the multiple embodiments described herein are not necessarily mutually exclusive and can exist in various combinations and arrangements without departing from the spirit and scope of the present invention, even if such combinations or arrangements are not expressed herein. In fact, without departing from the spirit and scope of the present invention, a person of ordinary skill in the art will think of various changes, modifications and other embodiments of the contents described herein.

[0072] Generally, the functionality for performing MR thermometry and detecting inaccurate thermal maps in MR thermometry can be constructed in one or more modules implemented in the form of hardware, software, or a combination of hardware and software, regardless of whether the functionality is integrated in the controller 124 of the imaging device (e.g., the MRI device 100), and / or provided by a separate external controller or one or more other computing entities, and the functionality includes: for example, as described above, before or during a medical procedure, analyzing imaging data of a target area and / or a non-target area acquired using one or more imaging modalities (e.g., MR imaging); determining a target position; generating a baseline phase image based on the imaging data; generating an MR thermal map; calculating a temperature difference between two thermal maps; and establishing a computational physics model and / or a statistical model. to predict a temperature rise during treatment; compare the measured temperature (or temperature change) with the predicted temperature (or temperature rise); determine whether a thermal map acquired at a later time is inaccurate based on the comparison and / or historical imaging data; calculate ultrasound parameter values ​​for generating a focal zone at a target area; activate a medical device (e.g., an ultrasound transducer) based on the determined parameter values; obtain anatomical properties and / or material properties of target tissue and / or non-target tissue; computationally predict an ultrasound beam path; computationally predict the propagation of induced effects through tissue; computationally predict the ultrasound energy delivered to the target area and / or non-target area within a time interval; and computationally predict the conversion of ultrasound energy or pressure into heat at the target area and / or non-target area. The controller 124 may include one or more modules implemented in hardware, software, or a combination of both hardware and software. For embodiments in which the functionality is provided as one or more software programs, the programs may be written in any of a number of high-level languages, such as PYTHON, FORTRAN, PASCAL, JAVA, C, C++, C#, BASIC, various scripting languages, and / or HTML. Additionally, the software may be implemented in assembly language for a microprocessor resident on the target computer; for example, if the software is configured to run on an IBM PC or PC clone, the software may be implemented in Intel 80x86 assembly language. The software may be embodied on an article of manufacture, including, but not limited to, a floppy disk, a flash disk, a hard disk, an optical disk, a tape, a PROM, an EPROM, an EEPROM, a field programmable gate array, or a CD-ROM. Embodiments using hardware loops may be implemented using, for example, one or more FPGA processors, CPLD processors, or ASIC processors.

[0073] In addition, the term "controller" used in this document broadly includes all necessary hardware components and / or software modules for performing any functions described above; the controller may include multiple hardware components and / or software modules, and the functions may be extended between different components and / or modules.

[0074] The terms and expressions employed herein are used as descriptive terms and expressions, rather than limiting terms and expressions, and when such terms and expressions are used, it is not intended to exclude any equivalents of the features shown and described or parts thereof. In addition, certain embodiments of the present invention have been described, and it will be apparent to those of ordinary skill in the art that other embodiments incorporating the concepts disclosed herein may be used without departing from the spirit and scope of the present invention. Therefore, the described embodiments should be considered in all respects to be illustrative only and not restrictive.

Claims

1. A system for performing magnetic resonance (MR) thermometry, the system comprising: Magnetic Resonance Imaging (MRI) Unit; as well as a controller in communication with the magnetic resonance imaging unit and configured to: (i) causing the magnetic resonance imaging unit to acquire at least one baseline phase image of an imaging region and at least one treatment phase image of the imaging region after a temperature change of a sub-region within the imaging region; (ii) electronically generating a thermal map based at least in part on the acquired baseline phase image and the treatment phase image, the thermal map indicating temperature changes of the sub-region on a pixel-by-pixel basis; (iii) computationally predicting a change in temperature of the sub-region based at least in part on energy accumulated in the sub-region during treatment without reference to the generated thermal map; and (iv) determining whether the thermal map is inaccurate based at least in part on whether a deviation between a temperature change of the sub-region indicated by the thermal map and a predicted temperature change of the sub-region exceeds a predetermined threshold, wherein the value of the predetermined threshold is based on the amount of acoustic energy delivered to the sub-region, and the predetermined threshold is positively correlated with the energy accumulated in the sub-region during processing.

2. The system of claim 1, wherein: The controller is also configured to: comparing the temperature changes in the generated heat map with the predicted temperature changes to determine deviations therebetween; and The deviation is compared to a predetermined threshold.

3. The system of claim 2, wherein: The controller is further configured to determine that the thermal map is inaccurate upon determining that the deviation exceeds the predetermined threshold.

4. The system of claim 3, wherein: The predetermined threshold is a fixed value.

5. The system of claim 3, wherein: The controller is also configured to adjust the predetermined threshold based at least in part on energy accumulated in the sub-region, a noise level associated with the baseline phase image and / or the treatment phase image, or a deviation between a temperature change in the generated thermal map and a predicted temperature change.

6. The system of claim 1, further comprising a medical device configured to induce a temperature change in the sub-region.

7. The system of claim 6, wherein: The medical device includes an ultrasound transducer including a plurality of transducer elements, and the controller is further configured to computationally predict a temperature change of the sub-region using a physical model.

8. The system of claim 7, wherein: The physical model is based at least in part on values ​​of ultrasound parameters used to generate a focal zone at the sub-region.

9. The system of claim 8, wherein: The ultrasound parameters include at least one of amplitude, frequency, phase, direction, or activation time associated with each of the transducer elements.

10. The system of claim 1, wherein: The controller is further configured to computationally predict temperature changes of the sub-region using a physical model without reference to the generated thermal map.

11. The system of claim 10, wherein: The physical model is based at least in part on tissue characteristics associated with at least one of the subregion or a second subregion different from the subregion.

12. The system of claim 11, wherein: The controller is further configured to acquire the tissue characteristic based at least in part on imaging data acquired using the magnetic resonance imaging unit.

13. The system of claim 11, wherein: The tissue characteristic includes at least one of type, structure, thickness, density, acoustic velocity, thermal absorption coefficient, perfusion coefficient, or metabolic heat generation rate.

14. The system of claim 10, wherein: The physical model is based on biological heat transfer equations.

15. The system of claim 14, wherein: The biological heat transfer equations include the Pennes equation.

16. The system of claim 1, wherein: The controller is further configured to predict temperature changes of the sub-regions using a statistical model.

17. The system of claim 16, further comprising a medical device configured to cause a temperature change in the sub-region, wherein: The statistical model includes historical data of temperature changes resulting from previous activations of the medical device.

18. The system of claim 1, wherein: The controller is further configured to cause the magnetic resonance imaging unit to acquire a reference library comprising a plurality of baseline phase images of the imaging region, each of the baseline phase images corresponding to a phase background during a different phase of expected motion of the imaging region.

19. The system of claim 18, wherein: The controller is further configured to identify a baseline phase image in the reference library that best matches the treatment phase image based on similarities between the baseline phase image and the treatment phase image, and to generate the thermal map based at least in part on the identified best matching baseline phase image.

20. A method of performing magnetic resonance (MR) thermometry, the method comprising: acquiring at least one baseline phase image of an imaging region and at least one treatment phase image of the imaging region after a temperature change of a subregion within the imaging region; electronically generating a thermal map based at least in part on the acquired baseline phase image and the treatment phase image, the thermal map indicating temperature changes of the sub-region on a pixel-by-pixel basis; computationally predicting a change in temperature of the subregion based at least in part on energy accumulated in the subregion during treatment without reference to the generated thermal map; and Determining whether the thermal map is inaccurate is based at least in part on whether a deviation between a temperature change of the sub-region indicated by the thermal map and a predicted temperature change of the sub-region exceeds a predetermined threshold, wherein the value of the predetermined threshold is based on the amount of acoustic energy delivered to the sub-region, and the predetermined threshold is positively correlated with the energy accumulated in the sub-region during processing.

21. A system for performing magnetic resonance (MR) thermometry, the system comprising: Magnetic Resonance Imaging (MRI) Unit; as well as a controller in communication with the magnetic resonance imaging unit and configured to: (i) causing the magnetic resonance imaging unit to acquire at least one baseline phase image of an imaging region and a plurality of treatment phase images of the imaging region after a temperature change of at least a sub-region within the imaging region; (ii) electronically generating a plurality of thermal maps based at least in part on the acquired baseline phase image and the acquired treatment phase image, each of the thermal maps indicating, on a pixel-by-pixel basis, a temperature change of the sub-region associated with one of the plurality of treatment phase images; and (iii) determining whether a thermal map among the plurality of thermal maps is inaccurate based at least in part on whether a deviation between a temperature change associated with the thermal map and a temperature change associated with at least another thermal map among the plurality of thermal maps exceeds a predetermined threshold, wherein a value of the predetermined threshold is based on an amount of acoustic energy delivered to the sub-region, and the predetermined threshold is positively correlated with energy accumulated within the sub-region during processing.

Images