METHOD AND SYSTEM FOR SETTING UP A HIGH-INTENSITY FOCUSED ULTRASOUND TREATMENT DEVICE

Abstract

The present invention relates to a parameterization method and system (100) integrated within a high-intensity focused ultrasound (HIFU) treatment device (150). The parameterization system comprises a real-time simulation unit (140) for predicting, based on geometric and physiological parameters of tissue regions in the area to be treated, on the one hand, and treatment parameters, on the other, the distribution of the ultrasound field within that area. The calculation is performed in real time using a meta-model: the ultrasound field is estimated from an interpolation of pre-calculated ultrasound field maps stored in a database, said maps being associated with different values ​​of the geometric and physiological parameters of the tissue regions in question. The thermal dose deposited at each point during treatment is then calculated, and the tissue response is estimated (Fig. 1).

Description

Title of the invention: METHOD AND SYSTEM FOR SETTING A HIGH INTENSITY FOCUSED ULTRASOUND TREATMENT DEVICE Technical field

[0001] The present invention relates generally to high intensity focused ultrasound (HIFU) treatment devices. It is particularly applicable to tumor destruction by remote thermal ablation with preservation of intermediate tissues. State of the prior art

[0002] High-intensity focused ultrasound (HIFU) therapy involves using the energy of a focused ultrasound beam to modify or destroy biological tissues. Tissue destruction occurs through protein coagulation, inducing irreversible cell damage and apoptosis.

[0003] This technique is widely used for thermal ablation of benign or malignant tumors (prostate cancer, liver metastases, brain tumors). It has the advantage of being percutaneous (therefore non-invasive) and of allowing the preservation of surrounding healthy tissue.

[0004] HIFU treatment generally uses a phased array multi-transducer probe. With each ultrasound shot, a delay law and an amplitude law are applied to the different elements of the transducer, so as to form a beam focused at a predetermined point. The operation is repeated at different points of the target area to be destroyed, by choosing the repetition rate (recurrence frequency) and the energy of the pulses.

[0005] Before applying the treatment protocol, the practitioner must ensure that the thermal dose deposited is sufficient to destroy the target area while preserving the healthy areas. To do this, the target area is meshed by a grid of points and a simulation of focused ultrasound shots at each point or a group of neighboring points of the grid is carried out. The temperature rise at each point is then deduced from the resolution of the heat transfer equation in the tissues. A method for validating a HIFU treatment protocol has been described in application US-A-2021 / 0000541.

[0006] This simulation method works relatively well when the medium in which the ultrasonic waves propagate is homogeneous but becomes significantly more complex when several tissue areas are to be modeled, especially when a high spatial resolution is necessary. The simulation of the ultrasonic field- trasonic analysis based on the characteristics of the patient's different tissues can then require several hours on a conventional personal computer, which is incompatible with outpatient treatment.

[0007] It has also been proposed to significantly accelerate the simulation of a HIFU treatment by using a metamodel generated from a simulated database. This is obtained by defining a set of N uncertain parameters, here the physiological and anatomical characteristics of the patient influential for the simulation, as well as their assumed variation domain. The database is obtained by performing once and for all (off-line) the simulation of the ultrasound field for all possible configurations in this N-dimensional space (basic configurations). This first simulation gives the distribution of the ultrasound pressure field for each basic configuration.

[0008] A second simulation is then carried out online by interpolating the simulation results obtained for the different previous configurations, i.e. by interpolating the ultrasonic pressure fields contained in the database. This is called a meta-model since it involves replacing the direct simulation model, assumed to be exact, with a much faster, or even real-time, interpolator, whose error is controlled.

[0009] Such a method of simulating a HIFU ultrasound field was described in the article by S. Chatillon et al., entitled “Applicatons of intensive HIFU simulation based on surrogate models using the CIVA Healthcare platform” published in J. of Phys., 1761, 2021.

[0010] If the physiological and anatomical parameters of the patient are perfectly known, this simulation method makes it possible to estimate the lesion produced for an established protocol, and thus to verify that it respects the constraints imposed in terms of target area to be destroyed and healthy tissue to be preserved. When the treatment protocol envisaged by the practitioner does not respect the aforementioned constraints, the latter must modify it empirically and carry out a new online simulation. The convergence of the process can be relatively slow and lead to a sub-optimal solution.

[0011] The aim of the present invention is therefore to propose a method and a parameterization system embedded within a HIFU treatment device which does not have the aforementioned drawbacks, in particular which makes it possible to obtain very quickly a quasi-optimal parameterization from the physiological parameters of the patient, the target area to be necrotic and the healthy areas to be spared. Presentation of the invention

[0012] The present invention is defined by a method of parameterizing a device of HIFU treatment equipped with an ultrasound probe for necrosing at least one tissue region of an area to be treated of a patient, said treatment being original in that it comprises a plurality of ultrasound shots and being original in that the treatment parameters defining in particular the position and orientation of the probe at each shot relative to the area to be treated, the area to be treated being modeled by different tissue regions, including at least one tissue region to be necrosed and one tissue region to be preserved, each tissue region being characterized by geometric parameters and physiological parameters,said parameterization method comprising: - a step of estimating the ultrasound field in the area to be treated by means of a meta-model using a database in which pre-calculated ultrasound field maps are stored for different values ​​of the geometric and physiological parameters of each tissue region of said area, the estimation of the ultrasound field being obtained by interpolation of pre-calculated maps; - a step of estimating the thermal dose deposited at each point of a grid within the area to be treated from the ultrasound field in this area; - a step of estimating the tissue response at each point of said grid to determine whether or not the deposited thermal dose leads to necrosis at this point, thus defining an estimated necrotic region and an estimated preserved region; - a step of adapting the treatment parameters aimed at minimizing a cost function depending on the difference between the estimated necrotic region and the tissue region to be necrotized, subject to the constraint that the region to be preserved is included in the estimated preserved region.

[0013] Said treatment parameters may also include the duration and repetition rate of the ultrasonic shots.

[0014] When the probe comprises a plurality of transducers, said processing parameters may comprise a phase and / or frequency and / or power law to be applied to all of the transducer elements.

[0015] The physiological parameters of a tissue region comprise, for example, at least one parameter among the density, the specific heat, the thermal conductivity of the tissue, the specific heat of the blood and the blood perfusion rate of the tissue.

[0016] The estimation of the ultrasonic field can be obtained in the local reference frame of the probe by interpolation of said precalculated maps, the distribution of the ultrasonic field in the area to be calculated being deduced by changing the reference frame.

[0017] The estimation of the thermal dose is advantageously obtained by solving a heat transfer equation in the different tissue regions, taking into account the ultrasonic energy absorbed at each point of the grid at each ultrasonic shot.

[0018] The result of the processing can be simulated by calculating, at each point of the grid, the thermal dose absorbed at this point in the form of an exposure time equivalent to a reference temperature, the tissue being considered necrotic at this point if this time is greater than a predetermined threshold value and preserved otherwise.

[0019] According to a variant, the cost function depends on the difference between the extent of the tissue region to be necrotized and that of the estimated necrotic tissue region.

[0020] This cost function can be defined as a sum of elementary cost functions associated with each of the shots, the processing parameters being adapted iteratively by means of gradient descent, stochastic gradient descent or a genetic algorithm.

[0021] For at least one iteration of adaptation of the treatment parameters, the estimated necrotic tissue region and / or the estimated preserved tissue region are / are advantageously displayed on a display with the tissue region to be necrotic and / or the tissue region to be preserved superimposed on the image of the area to be treated.

[0022] The present invention also relates to a system for parameterizing a HIFU treatment device equipped with an ultrasound probe for necrosing at least one tissue region of an area to be treated of a patient, said treatment comprising a plurality of ultrasound shots and being characterized by treatment parameters defining in particular the position and orientation of the probe at each shot relative to the area to be treated, the area to be treated being modeled by different tissue regions, including at least one tissue region to be necrosed and one tissue region to be preserved, each tissue region being characterized by geometric and physiological parameters, said parameterization system comprising: - a module for estimating the ultrasound field in the area to be treated by means of a meta-model, a database in which pre-calculated ultrasound field maps are stored for different values ​​of the geometric and physiological parameters of each tissue region of said area, said estimation module estimating the ultrasound field by interpolation of pre-calculated maps extracted from the database; - a module for estimating the thermal dose deposited at each point of a grid within the area to be treated from the ultrasound field in this area; - a module for estimating the tissue response at each point of said grid to determine whether or not the thermal dose deposited leads to necrosis at this point, thus defining an estimated necrotic region and an estimated preserved region; - a module for adapting the treatment parameters to minimize a cost function depending on the difference between the estimated necrotic region and the tissue region to be necrotized, subject to the constraint that the region to be preserved is included in the estimated preserved region.

[0023] Said treatment parameters also include the duration and repetition rate of the ultrasonic shots.

[0024] When the probe comprises a plurality of transducers, said processing parameters may comprise a phase and / or frequency and / or power law to be applied to all of the transducer elements. Brief description of the figures

[0025] Other characteristics and advantages of the invention will appear on reading a preferred embodiment of the invention, made with reference to the attached figures among which:

[0026] [Fig.l] schematically represents a system for setting the parameters of a HIFU treatment device according to an embodiment of the invention;

[0027] [Fig.2] schematically represents the simulation module used in the parameterization system of [Fig.l];

[0028] [Fig.3] schematically represents a method for predicting the distribution of the intensity of the ultrasonic field, used in [Fig.2];

[0029] [Fig.4] schematically represents a method for estimating the thermal dose deposited in the area to be treated, used in [Fig.2]. Description of the embodiments

[0030] We will consider in the following a device for treatment by high intensity focused ultrasound as described in the introductory part. Such a device can be equipped with a probe with a single piezoelectric transducer whose shape gives the focusing law, or more generally a multi-transducer probe, the focusing law being given by the distribution of the delays and the acoustic powers relative to the different piezoelectric elements. In the following, we will assume for the purpose of simple illustration and without prejudice to generalization that the probe is multi-transducer. The treatment is carried out by successive shots, each shot corresponding to the application of a given distribution of phases, amplitudes, or even frequencies. The presence of heterogeneous media between the transducer and the target zone may give rise to the application, on each transducer element, of a specific value of phase, amplitude, or frequency.For example, absorption by a medium with high acoustic attenuation (organ or tumor for example) can be compensated by increasing the power on certain transducer elements or reducing the emission frequency, so as to balance the energy contributions at the focal point.

[0031] The idea behind the present invention is to provide a module (plug-in) which can be added to or integrated into a HIFU treatment device so as to provide a parameter setting of the ultrasound probe for a given area to be treated.

[0032] More precisely, the practitioner first defines an area to be treated in the patient's body, this area comprising parts to be necrotized (cancerous tissues, metastases, etc.) by thermal effect (raising the local temperature to a predetermined threshold value for a given duration) and parts to be preserved (healthy tissues).

[0033] A HIFU treatment device associated with a parameterization system according to the present invention has been shown schematically in [Fig.l].

[0034] The HIFU device equipped with its ultrasound probe 160 is shown at 150.

[0035] The parameterization system 100 can receive from the HIFU device data representing the area to be treated as well as the characteristics of the probe and provide it in return with the parameters for insonifying the area in question.

[0036] The parameterization system essentially comprises a simulation module 140 making it possible to estimate for each shot the intensity of the ultrasonic field in the area to be treated and to deduce therefrom the thermal dose deposited at each point in this area.

[0037] The configuration system may advantageously be equipped with an imaging device, 110, a user interface, 130, and a graphical interface, 120.

[0038] The practitioner will be able to view the area to be treated on the display of the imaging device and plan the treatment in the form of a sequence of shots, each shot being defined by a disposition of the probe relative to the area to be treated (position and orientation of the probe relative to the latter), a focal point, in other words a law of delay of the pulses applied to the different elements of the probe, a pulse energy (power and duration, if applicable), or even an ultrasonic frequency.

[0039] Alternatively, the firing sequence may have been previously defined by the practitioner using the HIFU device and the corresponding data will have been transmitted to the parameterization system. This may be the case in particular if a previous treatment of the patient has been recorded in the HIFU device.

[0040] In any event, the area to be treated as well as the successive shots can be viewed on the display of the imaging device. This device could be, for example, that of a conventional ultrasound scanner.

[0041] The practitioner can, using the user interface, define additional shots or delete programmed shots, modify the position and / or orientation of the probe, adjust the power and / or duration or even the frequency of repetition of the shots, etc. The sequence of shots can be represented superimposed on the image of the area to be treated.

[0042] Once the nominal treatment parameters have been defined (positions and orientations of the probe, power and duration of the shots, ultrasound frequency, frequency of repetition of the shots, etc.), the simulation module allows, as described below, to calculate in quasi real time the spatial distribution of the ultrasound field at each shot, and to deduce the distribution of the thermal dose in the tissues. This possibility of simulation in Near real-time monitoring of the effects of a shooting sequence is essential as it allows for outpatient treatment.

[0043] It is then possible to view a map of the thermal dose on the display screen and to deduce, by comparison with a threshold, the necrotic parts and the spared parts. The practitioner can thus check whether the HIFU treatment plan is in accordance with the desired objective and, if so, validate the transfer of the treatment parameters to the HIFU device.

[0044] The module 140 can itself compare the result of the simulation with the desired objective and iteratively modify the parameters of the HIFU treatment device to comply with the treatment plan. For example, the simulation module can adjust the position and / or orientation of the probe relative to the area to be treated, the power and / or duration of the shots, the rate of the shots, etc.

[0045] In all cases, the conformity of the parameter setting of the treatment device is validated by the practitioner by means of the user interface before the parameters are transferred to the HIFU device.

[0046] [Fig.2] schematically represents the simulation module used in the setting system of [Fig.l].

[0047] This simulation module uses a meta-model to predict, using a database 210, the intensity of the ultrasound field in the area to be treated. More precisely, the configuration of the area to be treated is modeled by breaking it down into different tissue regions, each tissue region being characterized by geometric and physiological parameters.

[0048] The configuration of the area to be treated may differ from one patient to another depending on the geometric characteristics and physical parameters of the tissue regions that compose it. For example, the tissue regions may be fat, blood, skin, the parenchyma of an organ.

[0049] The geometric parameters of the different tissue regions may be, for example, the thickness, the radius of curvature and any other parameter making it possible to describe the shape or volume of such a region.

[0050] Among the physiological parameters of a tissue region, it will be possible to take into account in particular the acoustic attenuation coefficient of the tissue, its acoustic impedance, its density, its specific heat, its thermal conductivity, its temperature, etc.

[0051] The database contains a map of the intensity of the acoustic field in the area to be treated for a plurality of possible configurations of this area and a plurality of treatment parameters. More precisely, for each configuration of said plurality, the database contains a map of the pressure field in the area to be treated.

[0052] For example, the database will contain a mapping of the pressure field for different thicknesses of skin, fat layer, size or values ​​of geometric and / or physiological parameters of an organ. This mapping may be stored in the database for different treatment parameters and in particular different relative positions and orientations of the probe and the area to be treated.

[0053] Each mapping is associated with an N-tuple of samples of these N geometric and physical parameters. Thus, for example, for a given inclination of the probe, a given skin thickness, a given fat thickness, an acoustic attenuation value in an organ, the database contains the precalculated distribution of the intensity of the acoustic field in said area to be treated. For each of these configurations, this distribution will itself have been obtained by an offline simulation (CIVA simulation platform for example). It is important to note that such a simulation requires significant computing resources and is therefore incompatible with real-time constraints.

[0054] The prediction module 220 extracts from the database the ultrasound field map(s) corresponding to the geometric and physical parameters closest to those defining the envisaged treatment protocol. The treatment protocol includes in particular the successive positions and orientations of the probe relative to the area to be treated.

[0055] These treatment parameters may be nominal parameters initially provided by the HIFU device and / or those corrected by the practitioner using the imaging device 110 and the user interface 130.

[0056] The prediction module performs an interpolation of the ultrasound field maps thus obtained to obtain a map corresponding to the geometric and physical parameters of the treatment.

[0057] This ultrasound field mapping is obtained for each shot in the local reference frame of the probe and then transformed by changing the reference frame into a mapping in the reference frame of the area to be treated.

[0058] The mapping of the area to be treated is provided to the deposited thermal dose estimation module, 230. This calculates the temperature rise induced by the input of ultrasonic energy at each point of a grid of points in the area to be treated.

[0059] The tissue response estimation module, 240, then determines at each point of the grid whether the deposited thermal dose, corresponding to the accumulation of thermal heating over time, leads to tissue necrosis or not. To do this, this module will be able to access the physiological characteristics of the different tissues stored in the database 210. For example, for a given tissue, crossing a deposited thermal dose threshold will be a criterion allowing us to conclude that it has been destroyed. On the other hand, if this threshold is not crossed over the period in question, the tissue may be considered preserved. Where appropriate, the criterion will be expressed in the form of a probability of destruction based on statistics already obtained on the tissues in question.

[0060] A map of the necrotic regions or of the probability of tissue necrosis can then be superimposed on the image of the area to be treated on the display of the imaging device 110.

[0061] Advantageously, a module for adapting the processing parameters will use a cost function depending on the deviation observed at each point of the regions to be destroyed. This cost function is minimized under the constraint of not fulfilling the destruction criterion in the regions to be preserved. Where appropriate, the different regions to be destroyed will be assigned weighting coefficients in the cost function depending on the importance of their consideration in the processing. The cost function can be expressed in the form of a sum of elementary functions associated with each of the shots, each elementary function being able to depend in particular on the position and orientation of the probe during the shot, the power and duration of the pulses as well as the focusing law applied (giving the position of the focal point) during this shot.

[0062] The minimization of the cost function under constraint can be carried out in a conventional manner with Lagrange multipliers, for example by associating a Lagrange multiplier with each region to be preserved.

[0063] The minimization of the cost function can be achieved by means of gradient descent, or even stochastic gradient descent when the number of shots is large, or even genetic algorithms, such as differential evolution for example.

[0064] The new treatment parameters obtained are then injected into the prediction module 220 for a new prediction of the ultrasound field.

[0065] The practitioner can visualize at each iteration the mapping of the necrotic regions and the preserved areas and validate the transfer of the treatment parameters to the HIFU device.

[0066] [Fig.3] schematically represents a method for predicting the distribution of the intensity of the ultrasonic field, used in [Fig.2].

[0067] The prediction method assumes that the area to be treated has been previously modeled in 310 in different tissue regions, each tissue region being characterized by a geometric and / or physiological parameter as indicated above.

[0068] It also assumes having modeled in 320 the treatment protocol by means of treatment parameters, such as the position and orientation of the probe relative to the area to be treated, as well as the power, the duration, the delay / phase law, the frequency of the pulses applied to the different transducer elements of the probe, this at each shot. The sequence of shots can be described by a probe trajectory relative to the area to be treated, the latter being able to be divided into slices if necessary.

[0069] From the geometric and physiological parameters of the tissue regions, as well as the treatment parameters, the relevant configurations of these parameters, for example the configurations closest to or relative to values ​​belonging to ranges around these parameters are searched in the database. The precalculated ultrasound field maps associated with these different configurations are then extracted from the database.

[0070] At 340, the distribution of the intensity of the ultrasound field in the area to be treated is estimated in the local reference frame of the probe by interpolating the field maps extracted from the database. This estimation is repeated for each shot of the treatment protocol.

[0071] Finally, a change of reference is carried out at 350 with each shot so as to obtain the distribution of the ultrasonic field in the reference of the area to be treated.

[0072] [Fig.4] schematically represents a method for estimating the thermal dose deposited in the area to be treated, used in [Fig.2].

[0073] The area to be treated is spatially sampled using a grid of points.

[0074] At each shot, the acoustic energy delivered at each point is estimated at 410, for each point of the grid, taking into account the power and duration of the shot and the attenuation along the propagation paths in the different tissue regions.

[0075] In step 420, the heat transfer equation in biological tissues or BHTE (Bio Heat Transfer Equation) is solved using an explicit finite difference resolution method. A presentation of this resolution method can be found in the work by J. Chato entitled "Fundamentals of Bioheat Transfer; Springer Berlin, Heidelberg, 1990.

[0076] The BHTE equation can be expressed in the following form:

[0077] [Math.l] pC^ = VÀVTP + Mbch(TA-TF)+Qp

[0078] where Tp is the temperature at point P and at time 'J\ is the blood temperature arterial, P, C and k are respectively the density, the specific heat of the tissue and the thermal conductivity of the tissue, Cb and are respectively the specific heat of the blood and the blood perfusion rate, and finally Qp is the density of ultrasonic energy absorbed at point P.

[0079] The initial conditions are given by the body temperature and that of the arterial blood. Boundary conditions can also be set, for example the temperature of the gel or liquid between the probe and the patient's body to facilitate acoustic impedance matching.

[0080] Solving the BHTE equation makes it possible to obtain at each point of the grid the evolution of the temperature over time.

[0081] The actual estimation of the thermal dose deposited at each point of the grid is then carried out at 430. The thermal dose can be conventionally expressed as an equivalent duration of exposure to a reference temperature, i.e. 43°C, as described in the article by P. Lele entitled “Thresholds and mechanisms of ultrasonic damage to “organized” animal tissues” published in Symposium of Biological Effects and Characterization of Ultrasound Sources, Rockville, MD: DHEW (Pub) FDA; 78-8048, pp. 224-239, 1977.

[0082] The equivalent duration of exposure to the reference temperature at point P is given by:

[0083] [Math.2]

[0084] where P = 0.25 if TP < 43 ° C and P — 0.5 otherwise, and tmd are respectively the start and end time of the HIFU treatment.

[0085] If the equivalent exposure time is greater than a predetermined threshold value, the tissue is considered necrotic. Otherwise, it is considered spared.

[0086] Other criteria for the destruction of the holder, for example duration of exposure beyond a threshold temperature, may be envisaged by those skilled in the art without departing from the scope of the present invention.

[0087] The parameterization system of the HIFU treatment device makes it possible to simulate a treatment protocol in near real time and, if necessary, adapt the nominal parameters so as to satisfy an objective defined by the contours of the regions to be necrotic and those of the regions to be spared in the area to be treated. The practitioner can intervene at any time in the iterative process and add constraints or release others. Once the treatment parameters have been validated, they are transferred to the HIFU device to start the treatment.

Claims

Claims

1. Method for configuring a HIFU treatment device (150) equipped with an ultrasound probe for necrosing at least one tissue region of an area to be treated of a patient, said treatment comprising a plurality of ultrasound shots and being characterized by treatment parameters defining in particular the position and orientation of the probe at each shot relative to the area to be treated, the area to be treated being modeled by different tissue regions, including at least one tissue region to be necrosed and one tissue region to be preserved, each tissue region being characterized by geometric parameters and physiological parameters,said parameterization method comprising: - a step of estimating the ultrasound field (220) in the area to be treated by means of a meta-model using a database in which are stored pre-calculated ultrasound field maps for different values ​​of the geometric and physiological parameters of each tissue region of said area, the estimation of the ultrasound field being obtained by interpolation of pre-calculated maps; - a step of estimating the thermal dose (230) deposited at each point of a grid within the area to be treated from the ultrasound field in this area; - a step of estimating the tissue response (240) at each point of said grid to determine whether or not the thermal dose deposited leads to necrosis at this point,thus defining an estimated necrotic region and an estimated preserved region; - a step of adapting the treatment parameters (250) aimed at minimizing a cost function depending on the difference between the estimated necrotic region and the tissue region to be necrotized, under the constraint that the region to be preserved is included in the estimated preserved region, said cost function being a sum of elementary cost functions associated with each of the shots.,

2. Method for setting a HIFU treatment device according to claim 1, characterized in that said treatment parameters also include the duration and repetition rate of the ultrasonic shots.

3. Method for setting a HIFU treatment device according to the re- claim 1 or 2, characterized in that the probe comprises a plurality of transducers and that said processing parameters comprise a phase and / or frequency and / or power law to be applied to all of the transducer elements.

4. Method for setting a HIFU treatment device according to one of the preceding claims, characterized in that the physiological parameters of a tissue region comprise at least one parameter among the density, the specific heat, the thermal conductivity of the tissue, the specific heat of the blood and the blood perfusion rate of the tissue.

5. Method for parameterizing a HIFU treatment device according to one of the preceding claims, characterized in that the estimation of the ultrasound field is obtained in the local reference frame of the probe by interpolation of said precalculated maps, the distribution of the ultrasound field in the area to be calculated being deduced by changing the reference frame.

6. Method for setting a HIFU treatment device according to one of the preceding claims, characterized in that the estimation of the thermal dose is obtained by solving a heat transfer equation in the different tissue regions, taking into account the ultrasound energy absorbed at each point of the grid at each ultrasound shot.

7. Method for setting a HIFU treatment device according to one of the preceding claims, characterized in that the result of the treatment is simulated by calculating, at each point of the grid, the thermal dose absorbed at this point in the form of an exposure duration equivalent to a reference temperature, the tissue being considered necrotic at this point if this duration is greater than a predetermined threshold value and preserved otherwise.

8. Method for parameterizing a HIFU treatment device according to claim 7, characterized in that the cost function depends on the difference between the extent of the tissue region to be necrotized and that of the estimated necrotic tissue region.

9. Method for parameterizing a HIFU treatment device according to claim 8, characterized in that the treatment parameters are adapted iteratively by means of gradient descent, stochastic gradient descent or a genetic algorithm.

10. Method of setting a HIFU treatment device according to one of of the preceding claims, characterized in that for at least one iteration of adaptation of the treatment parameters, the estimated necrotic tissue region and / or the estimated preserved tissue region are / are visualized on a display with the tissue region to be necrotic and / or the tissue region to be preserved superimposed with the image of the area to be treated.

Images