A method for quantifying dose in tumor treatment fields (TTFIELDS) using power loss density and related measurements
By generating a conductivity 3D model and optimizing the electrode position to maximize or minimize power loss density, the problem of electrode positioning relies on electric field strength in the prior art is solved, and the treatment efficiency of tumor treatment fields and patient survival rate are improved.
Patent Information
- Application Number
- CN201980047938.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-04-15
- Filing Date
- 2019-07-18
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2039-07-18
AI Technical Summary
The existing tumor treatment field (TTFields) electrode positioning methods mainly rely on the simulation of electric field strength, and fail to fully consider the power loss density, resulting in insufficient treatment efficiency and electrode position optimization.
By obtaining images of body parts, a conductivity 3D model is generated, the power loss density in the target volume is determined, and the optimal electrode position is selected according to the power loss density distribution to achieve the maximum or minimum power when an alternating electric field is applied in the target volume Loss density.
Improves the treatment efficiency of TTFields, ensures maximum or minimum average or local power loss density in the target volume, optimizes the position of the electrode array, and improves overall and progress-free survival rates of patients.
Smart Images

Figure CN112601578B_ABST
Abstract
Description
[0001] CROSS-REFERENCE TO RELATED APPLICATIONS
[0002] This application claims the benefit of U.S. Provisional Applications 62 / 700,080 (filed July 18, 2018), 62 / 754,901 (filed November 2, 2018), and 62 / 833,983 (filed April 15, 2019), each of which is incorporated herein by reference in its entirety.
[0003] background
[0004] Tumor Treatment Fields, or TTFields, are low-intensity (e.g., 1-3 V / cm) alternating electric fields in the mid-frequency range (100-300 kHz) that inhibit the growth of cancer cells. This non-invasive treatment targets solid tumors and is described in U.S. Patent 7,565,205, which is incorporated herein by reference in its entirety. TTFields are approved for the treatment of glioblastoma and can be delivered, for example, via the Optune™ system. Optune™ includes a field generator and two pairs of transducer arrays (i.e., electrode arrays) that are placed on the patient's shaved head. One pair of electrodes is located to the left and right (LR) of the tumor, while the other pair of electrodes is located in front and behind (AP) the tumor.
[0005] Preclinical studies have shown that the inhibitory effect of TTFields increases with field strength. Field strength has historically been used to quantify the dose of TTFields, and the positioning of electrodes has historically been adjusted to optimize field strength.
[0006] One example of a prior art method for determining where to position electrodes is described in U.S. Patent 10,188,851, which describes the generation of a 3D map of conductivity directly from MRI-derived measurements without segmenting the anatomical volume into tissue types. Model electrodes are then placed at different locations on this 3D map, and the electric field strength resulting from each of these different electrode layouts is analyzed. The electrode layout that produces the best electric field strength distribution is selected and then used to apply TTFields to the patient. SUMMARY OF THE INVENTION
[0008] One aspect of the invention relates to a first method for planning treatment in a portion of a subject's body using an alternating electric field at a given frequency. The first method includes (a) obtaining at least one image of the portion; (b) generating a 3D model of conductivity or resistivity at a given frequency within the portion based on the at least one image obtained; and (c) identifying a target volume within the 3D model, the target volume comprising a plurality of voxels. The first method also includes (d) adding a first set of model electrodes to the 3D model, wherein the first set of model electrodes are positioned at a first set of positions relative to the 3D model; (e) for each voxel in the target volume, determining a power loss density that will exist when the first set of model electrodes positioned at the first set of positions are used to apply an alternating electric field in the target volume; (f) adding a second set of model electrodes to the 3D model, wherein the second set of model electrodes are positioned at a second set of positions relative to the 3D model; and (g) for each voxel in the target volume, determining a power loss density that will exist when the second set of model electrodes positioned at the second set of positions are used to apply an alternating electric field in the target volume. The first method further includes (h) selecting a set of positions for the electrodes based on the results of steps (e) and (g).
[0009] Some instances of the first method further include outputting a description of the selected set of locations.
[0010] In some cases of the first method, step (e) and step (g) each include determining, for each voxel in the target volume, an electric field strength that will exist when a corresponding set of model electrodes positioned at a corresponding set of locations are used to apply an alternating electric field in the target volume; and determining a corresponding power loss density for each voxel in the target volume based on the conductivity of the 3D model at the voxel and the electric field strength at the voxel. In some of these cases, the power loss density for each voxel in the target volume is determined using the formula L = ½ σ |E|², where σ is the conductivity of the 3D model at the voxel and |E| is the electric field strength at the voxel.
[0011] In some cases of the first method, step (h) includes selecting a set of positions that maximizes an average power loss density in the target volume. In some cases of the first method, step (h) includes selecting a set of positions that maximizes a minimum power loss density in the target volume.
[0012] Some instances of the first method further include (h) adding a third set of model electrodes to the 3D model, wherein the third set of model electrodes are positioned at a third set of locations relative to the 3D model; and (i) for each voxel in the target volume, determining a power loss density that will exist when the third set of model electrodes positioned at the third set of locations are used to apply an alternating electric field in the target volume. In these instances, selecting includes selecting a set of locations for the electrodes based on the results of step (e), step (g), and step (i).
[0013] In some cases of the first method, the at least one image of the portion includes an MRI image of the portion.
[0014] Some aspects of the first method further include attaching a plurality of electrodes to the body of the subject at selected locations; and applying an AC voltage between the attached electrodes to apply an alternating electric field in the target volume.
[0015] Another aspect of the invention relates to a second method for planning treatment in a portion of a subject's body using an alternating electric field at a given frequency. The second method includes (a) obtaining at least one image of the portion; (b) generating a 3D model of conductivity or resistivity at the given frequency within the portion based on the at least one image obtained; and (c) identifying a target volume within the 3D model, the target volume comprising a plurality of voxels. The second method also includes (d) adding a first set of model electrodes to the 3D model, wherein the first set of model electrodes are positioned at a first set of positions relative to the 3D model; (e) for each voxel in the target volume, determining the smaller of a first power loss density and a second power loss density that will exist when the first set of model electrodes positioned at the first set of positions are used to apply an alternating electric field in the target volume in a first orientation and a second orientation, respectively; (f) adding a second set of model electrodes to the 3D model, wherein the second set of model electrodes are positioned at a second set of positions relative to the 3D model; and (g) for each voxel in the target volume, determining the smaller of a first power loss density and a second power loss density that will exist when the second set of model electrodes positioned at the second set of positions are used to apply an alternating electric field in the target volume in a first orientation and a second orientation, respectively. The second method also includes (h) selecting a set of positions for the electrodes based on the results of steps (e) and (g).
[0016] Some instances of the second method further include outputting a description of the selected set of locations.
[0017] In some cases of the second method, step (e) and step (g) each include determining, for each voxel in the target volume, a first orientation electric field strength that will exist when a corresponding set of model electrodes positioned at a corresponding set of positions are used in a first orientation to apply an alternating electric field in the target volume; determining, based on the conductivity of the 3D model at the voxel and the first orientation electric field strength at the voxel, a corresponding first orientation power loss density for each voxel in the target volume; determining, for each voxel in the target volume, a second orientation electric field strength that will exist when a corresponding set of model electrodes positioned at a corresponding set of positions are used in a second orientation to apply an alternating electric field in the target volume; determining, based on the conductivity of the 3D model at the voxel and the second orientation electric field strength at the voxel, a corresponding second orientation power loss density for each voxel in the target volume; and for each voxel in the target volume, selecting the smaller of the corresponding first orientation power loss density and the corresponding second orientation power loss density. In some of these cases, the corresponding first-orientation power loss density and the corresponding second-orientation power loss density are determined for each voxel in the target volume using the formula L = ½ σ |E|², where σ is the conductivity of the 3D model at the voxel and |E| is the corresponding electric field strength at the voxel.
[0018] In some cases of the second method, step (h) includes selecting a set of positions that maximizes an average power loss density in the target volume. In some cases of the second method, step (h) includes selecting a set of positions that maximizes a minimum power loss density in the target volume.
[0019] Some instances of the second method further include (h) adding a third set of model electrodes to the 3D model, wherein the third set of model electrodes are positioned at a third set of positions relative to the 3D model; and (i) for each voxel in the target volume, determining the smaller of a first power loss density and a second power loss density that will exist when the third set of model electrodes positioned at the third set of positions are used to apply an alternating electric field in the target volume in the first orientation and the second orientation, respectively. In these instances, the selecting includes selecting a set of positions for the electrodes based on the results of step (e), step (g), and step (i).
[0020] In some cases of the second method, the at least one image of the portion includes an MRI image of the portion.
[0021] Some aspects of the second method further include attaching a plurality of electrodes to the body of the subject at selected locations; and applying an AC voltage between the attached electrodes to apply an alternating electric field in the target volume.
[0022] In some cases of the second method, the treatment using the alternating electric field is planned such that the average local minimum power density in the target volume is at least 1.0 mW / cm³.
[0023] Another aspect of the invention relates to a third method for planning treatment in a portion of a subject's body using an alternating electric field at a given frequency. The third method includes (a) obtaining at least one image of the portion; (b) generating a 3D model of conductivity or resistivity at the given frequency within the portion based on the at least one image obtained; and (c) identifying a target volume within the 3D model, the target volume comprising a plurality of voxels. The third method also includes (d) adding a first set of model electrodes to the 3D model, wherein the first set of model electrodes are positioned at a first set of positions relative to the 3D model; and (e) for each voxel in the target volume, determining a power loss density that will exist when the first set of model electrodes positioned at the first set of positions are used to apply an alternating electric field in the target volume. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1A An axial slice through the computational model of the head is depicted.
[0026] Figure 1B The corresponding distribution of electric field strength for a given placement of electrodes on the model is depicted.
[0027] Figure 1C The corresponding distribution of power loss density for the same placement of electrodes is depicted.
[0028] Figure 2 is a flow chart of an example of using TTFields for planning treatment in a part of a subject's body that relies on power loss density.
[0029] Figure 3 Describes the implementation Figure 2 A method of step S26 in .
[0030] Figure 4 Depicted is an example of calculating the power loss density by using the lowest power loss density at each voxel.
[0031] Figure 5A and 5B is a Kaplan-Meier curve based on the local minimum power density, which divides the tested patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively.
[0032] Fig. 6A and 6B is a Kaplan-Meier curve based on the local minimum dose density, which divides the tested patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively.
[0033] Fig. 7A and7B is a Kaplan-Meier curve based on the local minimum field strength, which divides the tested patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively.
[0034] Various embodiments are described in detail below with reference to the drawings, wherein like reference numerals refer to like elements.
[0035] DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0036] The electric field strength quantifies the force that the electric field applies to an intracellular object (e.g., an organelle). However, when considering the dose of a physical pattern such as TTFields, it is important to consider not only the force at work, but also the amount of energy transferred from the pattern to the tissue (i.e., the work performed by the pattern). This is because quantifying energy and work can provide a better description of the extent to which a physical pattern changes the state of the object it operates on.
[0037] The present application explains that for any given set of electrode positions, the power loss density of the electric field can be advantageously used to quantify the TTFields dose in the target volume. Furthermore, the power loss density distribution can be analyzed for a plurality of different electrode positions so that the electrode positions that produce the best power loss density distribution can be selected. Then, using the selected set of electrode positions, TTFields are applied to the patient. This new approach is in stark contrast to traditional methods, in which the decision of where to position the electrodes is based on a simulation of the electric field strength.
[0038] Power loss density of the electric field L Defined as
[0039]
[0040] where σ is the electrical conductivity of the tissue, and is the electric field strength. This relationship is referred to in this article as equation (1). The power loss density is measured in milliwatts per cubic centimeter (mW / cm³).
[0041] In order to examine the distribution of TTFields power loss density when TTFields are delivered to the brain, a realistic patient head model is generated based on MRI of a glioblastoma patient using the method described in US 2018 / 0160933, entitled "Treating Patients with TTFields with the Electrode Positions Optimized Using Deformable Templates", which is incorporated herein by reference in its entirety. The realistic head model specifies the conductivity at each voxel in the 3D space within the head (including the brain, skull, and scalp), and any of a variety of alternative methods for generating 3D conductivity maps of the relevant body parts can also be used (for example, as described below in conjunction with step S22). After obtaining the realistic head model, a numerical simulation is run to simulate positioning Optune™ electrodes on those realistic head models, and to simulate applying AC voltages to those electrodes using the Optune™ system in order to simulate the delivery of TTFields to those realistic head models. Of course, in alternative embodiments, different types of electrodes (i.e., different from the Optune™ electrodes) can be used, as long as the simulation is adjusted to account for the differences in the electrodes.
[0042] Calculate and compare the field strength distribution and power loss density distribution within the model, and Figures 1A-1C The results of those comparisons for a representative set of electrode locations are depicted. More specifically, Figure 1A An axial slice through a computational model that specifies the conductivity at each voxel within the body part of interest is depicted. After positioning the electrode model on the computational model (at a representative set of electrode locations) and applying a simulated AC voltage from one side of those model electrodes to the other, the electric field strength (in V / cm) at each voxel was calculated to yield Figure 1B Then, using the above formula (1), Figure 1B The field intensity at each voxel is mapped to the corresponding power loss density, and the resulting power loss density distribution is displayed in Figure 1C The field intensity distribution and the power loss density distribution show a certain degree of correlation, but there are obvious differences between the two distributions. More specifically, the electric field intensity ( Figure 1B ) tends to increase in areas of low conductivity (such as white matter) and is lowest in areas of high conductivity (such as ventricles and resection cavities). On the other hand, the power loss density ( Figure 1C ) tend to increase in areas of higher conductivity. And within the ventricles and resection cavities, the power loss density can present values comparable to those observed in other tissue types.
[0043] Figure 2 Flowchart of an example of planning treatment using TTFields in a part of a subject's body that relies on power loss density. The example begins at step S21, in which relevant images of the subject's body part are obtained. These images can be one or more images of the body part obtained using any imaging modality (including but not limited to MRI, CT, etc.). In some preferred embodiments, at least one MRI image is used.
[0044] Next, in step S22, a conductivity map of the body part is generated based on the one or more images obtained in step S21. Note that because TTFields are alternating electric fields, and conductivity is frequency dependent, the conductivity map should take into account the frequency at which the TTFields are ultimately applied. The conductivity map is preferably a 3D model of the conductivity at the relevant frequencies within the body part. For example, because glioblastoma is treated with 200kHz TTFields, the conductivity map used to plan TTFields treatment of glioblastoma should specify the conductivity of each voxel at 200 kHz AC.
[0045] Various methods can be used to generate conductivity maps. For example, conductivity maps can be generated from MRI images using the method described in U.S. Patent No. 10,188,851, which is incorporated herein by reference in its entirety. Alternatively, conductivity maps can be generated by performing manual segmentation of MRI images (i.e., manually identifying various tissue types in the MRI images, such as white matter, gray matter, necrotic cores, etc.), and assigning appropriate conductivity values to each of the relevant tissue types. Various alternative methods for generating conductivity maps will be apparent to those skilled in the art. In order to produce a good computational model, it is preferred to use high-resolution images (e.g., having a resolution of at least 1 mm × 1 mm × 1 mm). Lower resolution images can also be used, but less accurate results will be produced. Note that although the embodiments described herein discuss mapping conductivity, alternative embodiments can provide similar results by mapping different electrical properties such as resistivity.
[0046] In step S23, a target volume is identified within the 3D model. The target volume includes a large number of voxels. In the context of glioblastoma, the target volume is typically an area within the brain containing the glioblastoma. In some embodiments, the target volume will be a gross tumor volume (GTV) or a clinical target volume (CTV). The GTV is the gross verifiable extent and location of the tumor, while the CTV includes the confirmed tumor (if present) and any other tissue with a presumed tumor. In many cases, the CTV is found by defining a volume that surrounds the GTV and adding a margin with a predefined width around the GTV.
[0047] In order to identify the GTV or CTV, the volume of the tumor must be identified within the MRI image. This can be done manually by the user, automatically, or using a semi-automatic method in which a user-assisted algorithm is used. When this task is done manually, the MRI data can be presented to the user, and the user can be asked to outline the volume of the CTV outline on the data. The data presented to the user can be structural MRI data (e.g., T1, T2 data). Different MRI modes can be displayed (registered) with each other, and the user can be presented with the option of viewing any data set and outlining the CTV outline. The user can be asked to outline the CTV on the 3D volume representation of the MRI, or the user can be given the option of viewing each 2D slice of the data and marking the CTV boundary on each slice. Once the boundary has been marked on each slice, the CTV can be found within the anatomical volume (and therefore within the reality model). In this case, the volume marked by the user will correspond to the GTV. In some embodiments, the CTV can then be found by adding a margin of a predefined width to the GTV. Similarly, in other embodiments, the user can be asked to mark the CTV using a similar procedure.
[0048] Optionally, semi-automatic segmentation methods of MRI data can be implemented. In examples of these methods, the user iteratively provides input to the algorithm (e.g., tumor location on the image, roughly marking the boundaries of the tumor, delineating the region of interest in which the tumor is located), which is then used by the segmentation algorithm. The user can then be given the option to refine the segmentation to improve the estimate of the CTV location and volume within the body part.
[0049] Whether using an automated or semiautomated approach, the identified tumor volume will correspond to the GTV, and the CTV can then be found automatically by extending the GTV volume by a predefined amount (e.g., defining the CTV as the volume of a 20-mm-wide margin surrounding the tumor).
[0050] In other embodiments, the GTV is defined as the volume of enhancing tumor tissue, while the peritumoral border zone (PBZ) is defined as the white matter and gray matter voxels within a 3 mm thick volume surrounding the following tissues: enhancing tumor, necrotic core, and resection cavity. In these embodiments, the target volume can be the combined volume of the GTV and the PBZ.
[0051] In other embodiments, it may be sufficient for the user to define a region of interest in which they want to optimize the power loss density as a target volume. This region of interest may be, for example, a box volume, a spherical volume, or a volume of any shape in the anatomical volume surrounding the tumor. When using this approach, complex algorithms for accurately identifying tumors may not be required.
[0052] In step S24, a set of model electrodes are added to the 3D model at a first set of positions relative to the 3D model so that a simulation of the resulting electric field distribution and power loss density distribution can be calculated for that particular set of positions. In the context of glioblastoma, it is common to position one electrode array (i.e., the anterior electrode array) on the forehead, one electrode array (i.e., the posterior electrode array) on the back of the head, one electrode array on the right side of the head, and one electrode array on the left side of the head. However, there is considerable leeway as to the exact location of each of those electrode arrays.
[0053] In step S26, for each voxel in the target volume, an expected power loss density is determined (ie, the power loss density that will exist when the first set of model electrodes positioned at the first set of locations are used to apply an alternating electric field in the target volume).
[0054] Figure 3 Describes the implementation of Figure 2 A method of step S26 in the method of step S26 in the method. In this method, a power loss density distribution is determined using a two-step process. In the first step (S32), for each voxel in the body part, the electric field strength that will exist when a first set of model electrodes positioned at a first set of positions are used to apply an alternating electric field in the target volume is determined.
[0055] This can be achieved, for example, by setting boundary conditions on the model of the real body part using the positions of the simulated electrodes from step S24, and then calculating (for example using numerical simulation) the electric field that develops within the model of the real body part when the electrodes are placed on the model of the real body part (for example, by applying a simulated AC voltage to the simulated electrodes) at the selected positions and under the applied boundary conditions.
[0056] For example, a model of an electrode array for applying TTFields is placed on a model of a realistic body part (e.g., a head model). A volume mesh suitable for finite element (FE) method analysis can then be generated. Next, boundary conditions can be applied to the model. Examples of boundary conditions that may be used include Dirichlet boundary (constant voltage) conditions on the electrode array, Neumann boundary conditions (constant current) on the electrode array, or a floating potential boundary condition that sets the potential at the boundary so that the integral of the normal component of the current density is equal to a specified magnitude. The model can then be solved using a suitable finite element solver (e.g., a low-frequency quasi-static electromagnetic solver) or alternatively a finite difference (FD) algorithm. Meshing, application of boundary conditions, and solving of the model can be performed using existing software packages such as Sim4Life, Comsol Multiphysics, Ansys, or Matlab. The final solution to the model will be a data set that describes the electric field distribution or related quantities (e.g., potential) within the computational model for a given set of electrode locations.
[0057] Then, in a second step (S33), the power loss density of each voxel in the target volume is calculated based on (a) the conductivity of the 3D model at the voxel and (b) the electric field strength at the voxel. This calculation is preferably implemented using the above formula (1). Note that in some preferred embodiments, it is preferred to determine the electric field strength for each voxel in the entire body part to facilitate the operation of the numerical simulation. However, it is only necessary to calculate the power loss density within the smaller target volume. Of course, if necessary, the power loss density can be optionally calculated within the entire body part.
[0058] Combination of the above Figure 3 The methods described provide power loss density for the case in which a continuous AC voltage is applied to the simulated electrodes. However, for many types of tumors, the orientation of the TTFields is switched periodically. For example, glioblastoma is often treated with TTFields by: (a) applying an AC voltage between the right and left electrodes for one second; (b) applying an AC voltage between the front and back electrodes for one second; and repeating steps (a) and (b) for the duration of the treatment. In these cases, the electric field distribution during step (a) will be different from the electric field distribution during step (b). And because the power loss density is derived from the electric field, this means that the power loss density during step (a) will also be different from the power loss density during step (b). In these cases, appropriate rules can be used to select the power loss density for the numerical simulation (which is ultimately used to determine where to position the electrodes).
[0059] Figure 4An example of a suitable rule is depicted, which determines the power loss density at each voxel by calculating the power loss density separately for each field direction at each voxel and using the one with the lowest power loss density at each voxel. More specifically, in step S42, for each voxel in the body portion, the electric field strength that will exist when an AC voltage is applied between the left electrode and the right electrode is calculated. Then, in step S43, for each voxel in the target volume, the corresponding power loss density is calculated, for example using formula (1). Similarly, in step S45, for each voxel in the body portion, the electric field strength that will exist when an AC voltage is applied between the front electrode and the back electrode is calculated. Then, in step S46, for each voxel in the target volume, the corresponding power loss density is calculated, for example using formula (1). Finally, in step S48, the lowest of the two power loss densities calculated for each voxel (i.e., in steps S43 and S46) is selected. The value of the power loss density selected in step S48 is then used in subsequent steps of the process.
[0060] This method of selecting the lowest power loss density at each voxel can also be extended to embodiments in which the electric field is applied in three or more different directions at different times. In these embodiments, when the electric field is applied in each of the three or more directions, the power loss density distribution is calculated separately, and the lowest of those power loss density values at each voxel is used as the value for subsequent calculations of that voxel. In alternative embodiments, when different power loss densities exist at any given voxel at different times, a different rule can be used (e.g., using an average value instead of a minimum value).
[0061] return Figure 2 , steps S24-S26 provide a method for calculating the power loss density (PLD) that will exist at each voxel in the target volume for a given set of electrode positions. And in step S28, those two steps (i.e., steps S24 and S26) are repeated for one or more other sets of electrode positions. Repeating these two steps for each set of electrode positions generates a separate PLD distribution. After all of these separate PLD distributions have been generated, processing proceeds to step S29, where a set of electrode positions is selected based on all of the generated PLD distributions. This selection can be done, for example, by selecting the best PLD (e.g., as explained below), and then outputting a description of the electrode positions that produced that particular PLD.
[0062] A variety of methods can be used to determine which PLD distribution is optimal for a particular patient. In one example, the PLD distribution that maximizes the mean PLD (averaged over all voxels in the target volume) can be considered optimal. Alternatively, the PLD distribution that maximizes the minimum PLD (measured over all voxels in the target volume) can be considered optimal. In other embodiments, a PLD distribution can be considered appropriate if the PLD averaged over all voxels in the target volume is above a given threshold (e.g., 2 mW / cm³).
[0063] After outputting a description of the electrode positions that produce the optimal PLD, a healthcare professional can use this description to place the electrodes on the patient's head (or other body part). Therapy using TTFields can then be performed by applying an AC voltage to the electrodes (e.g., as described in U.S. Patent 7,565,205, which is incorporated herein by reference).
[0064] Optionally, steps S24-S28 can be automated by running an optimization algorithm to find a layout that produces an optimal PLD distribution within the target volume. Such optimization can be implemented by: identifying a volume targeted for treatment within the body part model (as described above in conjunction with step S23); automatically placing an electrode array and setting boundary conditions on the model of the real body part (as described above in conjunction with step S24); once the electrodes have been placed on the model of the real body part and the boundary conditions have been applied, calculating the PLD distribution developed within the model of the real body part; and running an optimization algorithm to find a layout that produces the desired PLD distribution within the target volume.
[0065] In the context of the optimization, steps S24 and S26 are repeated. In each iteration, step S24 involves automatically positioning the electrode array on the model of the realistic body part.
[0066] Various methods can be used to find the best array layout for any given patient. One example of an optimization method is an exhaustive search. In this method, the optimizer will include a library with a limited number of array layouts that should be tested. The optimizer simulates all array layouts in the library (e.g., by repeating steps S24 and S26 for each layout) and picks the array layout that produces the best PLD distribution in the target volume.
[0067] Another optimization method is iterative search. This method uses algorithms such as minimum descent optimization method and simplex search optimization. Using this method, the algorithm iteratively tests different array layouts on the body part (e.g., head) and calculates the PLD distribution in the target volume for each layout. Therefore, this method also involves repeating steps S24 and S26 for each layout. At each iteration, the algorithm automatically selects the configuration tested based on the results of the previous iteration. The algorithm is designed to converge so that it maximizes (or minimizes) the defined objective function of PLD in the target volume.
[0068] The benefit of using power loss density to plan cancer treatment with TTFields has been demonstrated through a retrospective analysis of clinical trial results. This retrospective analysis showed that overall survival increased with the TTFields power (mW / cm 3 This analysis identified the benefit of optimizing the electrode array location on the patient's body for TTFields therapy by looking at the power loss density in order to maximize power delivery to the tumor.
[0069] In this study, a dose metric was defined that allowed the measurement of the cumulative dose over time. The association between field modeling dose and overall survival and progression-free survival in these patients was studied. The patient pool started with 466 patients. After a two-month treatment duration, 379 patients remained. Only in 317 patients was the MRI quality sufficient to generate a model. A total of 317 patients were generated in this study. The following patient data were used for modeling: (1) the patient's MRI at baseline; (2) the extent of resection (used to delineate the tumor and resection cavity); (3) the recommended transducer array layout as recorded in the patient's record; and (4) the average compliance and current derived from the log files of the devices used by the patient.
[0070] Using the Novocure Optune™ system, TTFields are delivered in two directions during treatment. Therefore, a complete analysis needs to consider the electric field effects in both directions (i.e., when the left / right electrodes are used to apply the field; and both when the front / back electrodes are used to apply the field). The power loss density at each voxel is determined by: determining the power loss density due to the left / right field, determining the power loss density due to the front / back field, and selecting the lower of those two power loss densities at each voxel. (This selection is based on the assumption that at each point, the channel delivering the lower dose (i.e., LR or AP) determines the efficacy). This parameter is referred to herein as the local minimum power density (LMiPD), and it represents the lower of the two power loss densities delivered to each point.
[0071] Then, the Kaplan-Meier curve was used to find the threshold LMiPD that divided the patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively, as Figure 5A and 5B . In each of these graphs, the lower trace represents an average LMiPD in the target volume of <2.4 mW / cm³, while the upper trace represents an average LMiPD in the target volume of ≥2.4 mW / cm³. This data suggests that it may be advantageous to plan treatment with TTFields such that the average LMiPD in the target volume is at least 2.4 mW / cm³. In alternative embodiments, treatment with TTFields may be planned such that the average LMiPD in the target volume is at least 1.15 mW / cm³. In other alternative embodiments, treatment with TTFields may be planned such that the average LMiPD in the target volume is at least 1.0 mW / cm³.
[0072] It is important to note that TTFields were not applied 100% of the time to the patients in the study (for example, if the TTFields were turned off in the shower or for any other reason). To account for this factor, another parameter was developed. This parameter is referred to in this article as the local minimum dose density (LMiDD), and it represents the LMiPD multiplied by the percentage of time that TTFields were applied.
[0073] Once again, Kaplan-Meier curves were used to find the threshold LMiDD that separated patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively, as Fig. 6A and 6B . In each of these graphs, the lower trace represents an average LMiDD in the target volume of <1.6 mW / cm³, while the upper trace represents an average LMiDD in the target volume of ≥1.6 mW / cm³. This data suggests that it may be advantageous to plan treatment with TTFields such that the average LMiDD in the target volume is at least 1.6 mW / cm³. In alternative embodiments, treatment with TTFields may be planned such that the average LMiDD in the target volume is at least 0.77 mW / cm³. In other alternative embodiments, treatment with TTFields may be planned such that the average LMiDD in the target volume is at least 0.7 mW / cm³.
[0074] This analysis shows that power loss density is a feasible physical measurement that can be used to accurately quantify TTFields dose in treatment planning.
[0075] Additional parameters were also analyzed. More specifically, the local minimum field intensity (LMiFI), which is the lower of the two field intensities delivered to each point in the target volume (i.e., the lower of the LR field and the AP field). Again, the Kaplan-Meier curve was used to find the threshold LMiFI that separated the patients into two groups with the most statistically significant differences in overall survival and progression-free survival, respectively, as Fig. 7A and 7B . In each of these figures, the lower trace represents the average LMiFI in the target volume <1.05 V / cm, while the upper trace represents the average LMiFI in the target volume ≥ 1.05 V / cm. This data suggests that it may be advantageous to plan treatment with TTFields such that the average LMiFI in the target volume is at least 1.05 V / cm. In alternative embodiments, treatment with TTFields may be planned such that the average LMiFI in the target volume is at least 1.0 V / cm. In other alternative embodiments, treatment with TTFields may be planned such that the average LMiFI in the target volume is at least 0.7 V / cm.
[0076] In addition, the total power loss (power delivered to the model by the TTFields) was calculated. The power loss density integrated over the head in the simulation revealed that the total power loss of the TTFields in the head during treatment was between 20-40 Watts, which is equivalent to 412-825 Kcal / day. Therefore, the power delivered by the TTFields is comparable to or greater than the resting metabolic rate of the brain (i.e., approximately 20% of the resting metabolic rate of the body, which is typically between 1,400 and 1,800 Kcal).
[0077] Finally, although the present application relies on examples for optimizing array placement on the head, the methods described herein can also be used to optimize array placement for treating other body regions including, but not limited to, the chest and abdomen.
[0078] Although the present invention has been disclosed with reference to certain embodiments, numerous modifications, variations and changes to the embodiments are possible without departing from the scope and range of the invention as defined in the appended claims. Therefore, it is contemplated that the present invention is not limited to the embodiments, but rather has the full scope defined by the language of the appended claims and their equivalents.
Claims
1. A method of operating a device for planning treatment in a part of a subject's body using an alternating electric field at a given frequency, the method comprising: the following: (a) the device obtains at least one image of the portion; (b) the device generates a 3D model of conductivity or resistivity at a given frequency within the portion based on the at least one image obtained; (c) the device identifies a target volume within the 3D model, the target volume comprising a plurality of voxels; and selecting, by the device, an optimal distribution by selecting a series of positions of the model electrodes for each of a plurality of distributions, each distribution having a different arrangement of the model electrodes, wherein selecting the series of positions of the model electrodes for each distribution comprises: (d) the apparatus adds a first set of model electrodes to the 3D model, wherein the first set of model electrodes are positioned at a first set of positions relative to the 3D model; (e) for each voxel in the target volume, the apparatus determines a power loss density that will exist when a first set of model electrodes positioned at a first set of locations is used to apply an alternating electric field in the target volume; (f) the apparatus adds a second set of model electrodes to the 3D model, wherein the second set of model electrodes are positioned at a second set of positions relative to the 3D model; (g) for each voxel in the target volume, the apparatus determines a power loss density that will exist when a second set of model electrodes positioned at a second set of locations is used to apply an alternating electric field in the target volume; and (h) The device selects a set of positions for the electrodes based on the results of steps (e) and (g).
2. The method of claim 1, further comprising the steps of: (i) The device outputs a description of the selected set of locations.
3. The method of claim 1, wherein step (e) and step (g) each comprise: The apparatus determines, for each voxel in the target volume, an electric field strength that will exist when a corresponding set of model electrodes positioned at a corresponding set of locations are used to apply an alternating electric field in the target volume; and The device determines a corresponding power loss density for each voxel in the target volume based on the conductivity of the 3D model at the voxel and the electric field strength at the voxel.
4. The method of claim 3, wherein the formula L = 1 / 2σ|E| 2 The power loss density for each voxel in the target volume is determined, where σ is the conductivity of the 3D model at the voxel and |E| is the electric field strength at the voxel.
5. The method of claim 1, wherein step (h) comprises the device selecting a set of locations that maximizes an average power loss density in the target volume.
6. The method of claim 1, wherein step (h) comprises the device selecting a set of locations that maximizes the lowest power loss density in the target volume.
7. The method of claim 1, further comprising the steps of: (i) the device adds a third set of model electrodes to the 3D model, each distribution comprising the device adding a third set of model electrodes to the model, wherein the third set of model electrodes are positioned at a third set of positions relative to the 3D model; and (j) for each voxel in the target volume, the apparatus determines a power loss density that will exist when a third set of model electrodes positioned at a third set of locations is used to apply an alternating electric field in the target volume; Wherein step (h) comprises the device selecting a set of positions for the electrodes based on the results of steps (e), (g) and (j).
8. The method of claim 1, wherein the at least one image of the portion comprises an MRI image of the portion.
9. A method of operating a device for planning treatment in a part of a subject's body using an alternating electric field at a given frequency, the method comprising: the following: (a) the device obtains at least one image of the portion; (b) the device generates a 3D model of conductivity or resistivity at a given frequency within the portion based on the at least one image obtained; (c) the device identifies a target volume within the 3D model, the target volume comprising a plurality of voxels; and selecting, by the device, an optimal distribution by selecting a series of positions of the model electrodes for each of a plurality of distributions, each distribution having a different arrangement of the model electrodes, wherein selecting the series of positions of the model electrodes for each distribution comprises: (d) the apparatus adds a first set of model electrodes to the 3D model, wherein the first set of model electrodes are positioned at a first set of positions relative to the 3D model; (e) for each voxel in the target volume, the apparatus determines the smaller of a first power loss density and a second power loss density that will exist when a first set of model electrodes positioned at a first set of locations are used to apply an alternating electric field in the target volume in a first orientation and a second orientation, respectively; (f) the apparatus adds a second set of model electrodes to the 3D model, wherein the second set of model electrodes are positioned at a second set of positions relative to the 3D model; (g) for each voxel in the target volume, the device determines the smaller of a first power loss density and a second power loss density that will exist when a second set of model electrodes positioned at a second set of locations are used to apply an alternating electric field in the target volume in a first orientation and a second orientation, respectively; and (h) The device selects a set of positions for the electrodes based on the results of steps (e) and (g).
10. The method of claim 9, further comprising the steps of: (i) The device outputs a description of the selected set of locations.
11. The method of claim 9, wherein step (e) and step (g) each comprise: The apparatus determines, for each voxel in the target volume, a first orientation electric field strength that will exist when a corresponding set of model electrodes positioned at a corresponding set of locations are used in a first orientation to apply an alternating electric field in the target volume; The device determines a respective first-orientation power loss density for each voxel in the target volume based on the conductivity of the 3D model at the voxel and the first-orientation electric field strength at the voxel; for each voxel in the target volume, the device determines a second-orientation electric field strength that will exist when a respective set of model electrodes positioned at a respective set of locations are used in a second orientation to apply an alternating electric field in the target volume; The device determines a corresponding second-oriented power loss density for each voxel in the target volume based on the conductivity of the 3D model at the voxel and the second-oriented electric field strength at the voxel; and For each voxel in the target volume, the device selects the smaller of a corresponding first-orientation power loss density and a corresponding second-orientation power loss density.
12. The method of claim 11, wherein the formula L = 1 / 2σ|E| 2 A corresponding first orientation power loss density and a corresponding second orientation power loss density are determined for each voxel in the target volume, where σ is the conductivity of the 3D model at the voxel and |E| is the corresponding electric field strength at the voxel.
13. The method of claim 9, wherein step (h) comprises the device selecting a set of locations that maximizes an average power loss density in the target volume.
14. The method of claim 9, wherein step (h) comprises the device selecting a set of locations that maximizes a minimum power loss density in the target volume.
15. The method of claim 9, further comprising the steps of: (i) the device adds a third set of model electrodes to the 3D model, each distribution comprising the device adding a third set of model electrodes to the model, wherein the third set of model electrodes are positioned at a third set of positions relative to the 3D model; and (j) for each voxel in the target volume, the apparatus determines the smaller of a first power loss density and a second power loss density that will exist when a third set of model electrodes positioned at a third set of locations are used to apply an alternating electric field in the target volume in a first orientation and a second orientation, respectively; Wherein step (h) comprises the device selecting a set of positions for the electrodes based on the results of steps (e), (g) and (j).
16. The method of claim 9, wherein the at least one image of the portion comprises an MRI image of the portion.
17. The method of claim 9, wherein the treatment using the alternating electric field is planned such that the average local minimum power density in the target volume is at least 1.0 mW / cm 3 .
Citation Information
Patent Citations
TTField treatment with optimization of electrode positions on the head based on MRI-based conductivity measurements
US10188851B2
Treating Patients with TTFields with the Electrode Positions Optimized Using Deformable Templates
US20180160933A1
Treating a tumor or the like with electric fields at different orientations
US7565205B2
Targeted electrical stimulation
US20160055304A1
TTField Treatment with Optimization of Electrode Positions on the Head Based on MRI-Based Conductivity Measurements
US20170120041A1
