Deep frequency conversion phased array tumor focusing hyperthermia method and system

Abstract

The present application belongs to the technical field of hyperthermia equipment control, and particularly relates to a deep frequency conversion phased array focusing tumor hyperthermia method and system. The backscattering signal containing tissue dielectric mutation characteristics is collected during the radio frequency emission pause period, a coherent interference evaluation matrix is constructed, and a residual error compensation phase sequence is extracted to reshape the coherent superposition of the beam in the target area; the target area plan and actual temperature rise rate deviation are converted into a global power pull-up compensation coefficient through nonlinear logarithmic mapping to smooth the compensation energy drop; the real-time monitoring of the standing wave ratio of each channel is carried out, the segmented exponential penalty mechanism is used to calculate the safety tolerance degree factor, the power of the high-risk channel is distributed to the safe channel according to the weight, and the total energy flux is maintained constant. The present application improves the focal point accuracy of the inhomogeneous target area, effectively suppresses the thermodynamic shock, and safely avoids the risk of hardware power amplifier damage.

Description

Technical Field

[0001] This invention relates to the field of thermotherapy equipment control technology. More specifically, this invention relates to a deep frequency conversion phased array focused tumor thermotherapy method and system. Background Technology

[0002] In the field of variable frequency phased array focused hyperthermia for deep tumors, working under an idealized electromagnetic transmission model, living tissue is often simplified as an isotropic homogeneous medium.

[0003] However, actual human tissue is a complex heterogeneous structure composed of skin, fat, muscle and bone, and its dielectric constant exhibits drastic abrupt changes in space.

[0004] When multiple radio frequency electromagnetic waves radiate toward the target area according to a preset phase, this dielectric abrupt change will trigger significant multipath scattering and secondary refraction effects, causing non-orthogonal coherent crosstalk between the beams of each channel in the target area, forming a destructive interferometric standing wave field.

[0005] This can lead to unpredictable distortion and shift of the energy focal point, defocusing, reduced heat deposition efficiency of the target lesion, and potential unintended overheating damage to superficial normal tissues.

[0006] Currently, backscattered signals carrying spatial differences in organizational electrical parameters are treated as harmful interference and discarded directly without extraction and analysis, thus losing the opportunity to perceive and compensate for non-ideal characteristics of the transmission path. Summary of the Invention

[0007] To address the technical problems of beam interference and focus distortion caused by abrupt changes in the heterogeneous dielectric properties of living tissue, resulting in low heat deposition efficiency in the target area and susceptibility of superficial normal tissue to overheating damage, the present invention provides solutions in the following aspects.

[0008] In a first aspect, the present invention provides a method and system for deep frequency conversion phased array focused tumor hyperthermia, the method comprising: During the radio frequency transmission pause, each channel is switched to listening mode to collect backscattered signals containing tissue dielectric mutation characteristics; Based on the backscattered signal, a cross-channel coherent crosstalk evaluation matrix is ​​constructed, and a residual compensation phase sequence for canceling destructive interference is obtained according to the matrix to reshape the coherent superposition of the beam in the target area. The deviation between the planned target area temperature rise rate and the actual temperature rise rate is obtained, and the deviation is converted into a global RF power boost compensation coefficient through nonlinear logarithmic envelope mapping to compensate for the energy drop caused by phase reshaping. Real-time monitoring of the standing wave ratio (SWR) of each channel; and calculation of the safety tolerance evaluation factor of each channel based on the SWR and a preset warning threshold using an exponential penalty mechanism. For channels with low security tolerance, power is reduced and the reduced power is allocated to secure channels according to the security tolerance assessment factor to maintain a constant total energy flux.

[0009] Preferably, the acquisition of backscattered signals containing tissue dielectric mutation characteristics includes: during the radio frequency transmission pause period, switching each channel from the transmission state to the high-sensitivity reception state, capturing weak radio frequency detection signals backscattered at the boundary of heterogeneous deep tissue, and obtaining the time-frequency domain backscattered complex impedance characteristic set of each channel, wherein the time-frequency domain backscattered complex impedance characteristic set includes the amplitude attenuation degree and phase shift hysteresis information of the signal.

[0010] Preferably, the construction of the cross-channel coherent crosstalk evaluation matrix based on the backscattered signal includes: using a channel cross-correlation topology algorithm with time-alignment characteristics, constructing the cross-channel coherent crosstalk evaluation matrix using the time-frequency domain backscattered complex impedance feature set, wherein the elements in the cross-channel coherent crosstalk evaluation matrix are complex numbers, the deflection angle of the complex number represents the radio frequency phase lead or lag caused by non-uniform transmission of the organization dielectric between any two independent channels, and the modulus of the complex number quantifies the energy intensity of the coherent crosstalk between channels.

[0011] Preferably, obtaining the residual compensation phase sequence for offsetting destructive interference based on the matrix includes: using the cross-channel coherent crosstalk evaluation matrix as a reverse analytical constraint, extracting the complex deflection information of the matrix elements, and inversely deriving the destructive interference residual compensation phase sequence for offsetting destructive standing wave effects.

[0012] Preferably, the step of converting the deviation into a global RF power boost compensation coefficient through nonlinear logarithmic envelope mapping includes: using natural logarithmic envelope mapping, with the ratio of the planned target area temperature rise rate to the actual temperature rise rate as the independent variable, introducing a response agility factor and an anti-collapse smoothing constant to prevent singularities in the denominator, and calculating the global RF power boost compensation coefficient, so that when the actual temperature rise rate lags behind the planned temperature rise rate, it provides boost momentum, while when the actual temperature rise rate approaches the planned temperature rise rate, the boost gradient decays logarithmically.

[0013] Preferably, the global RF power boost compensation coefficient is obtained by multiplying a constant 1 by the natural logarithm of the ratio of the sum of the planned target area temperature rise rate and the first anti-collapse smoothing constant to the sum of the actual target area temperature rise rate and the second anti-collapse smoothing constant.

[0014] Preferably, the calculation of the security tolerance evaluation factor for each channel using an exponential penalty mechanism includes: triggering an exponential penalty when the VSWR of a single channel is greater than the health baseline value; the security tolerance evaluation factor is defined using piecewise logic: when the VSWR of a single channel is greater than the health baseline value but less than a preset warning threshold, the security tolerance evaluation factor is an exponential function with the natural constant e as its base, and its exponential term is the penalty amplification exponent multiplied by the difference between the VSWR and the health baseline value divided by the difference between the preset warning threshold and the VSWR plus the negative value of the zero tolerance constant; when the VSWR is greater than or equal to the preset warning threshold, the security tolerance evaluation factor is forcibly set to zero.

[0015] Preferably, the step of reducing the power of channels with low security tolerance by allocating the reduced power to secure channels according to the proportion of the security tolerance evaluation factor includes: multiplying the global radio frequency power boost compensation coefficient by the initial power reference value of each channel, then multiplying by the normalized weight of the security tolerance evaluation factor of each channel in the sum of all channel security tolerance evaluation factors, and multiplying by the total number of channels to obtain the final target power transmitted by each channel.

[0016] Preferably, maintaining a constant total energy flux is achieved through normalized weight allocation.

[0017] Secondly, this invention provides a deep-field variable frequency phased array focused tumor hyperthermia system, employing the following technical solution: A deep frequency conversion phased array focused tumor hyperthermia system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a deep frequency conversion phased array focused tumor hyperthermia method as described above is implemented.

[0018] The embodiments of the present invention have at least the following beneficial effects: 1. By switching each channel to a high-sensitivity listening mode during the radio frequency (RF) transmission pause, backscattered complex envelope signals embedded with tissue dielectric mutation characteristics are captured. Based on this, a cross-channel coherent crosstalk assessment matrix is ​​constructed. This matrix, through cross-correlation integration of the time-domain envelope, removes random white noise and purifies the physical interference components caused by reflections from heterogeneous tissue interfaces. From this, a destructive interference residual compensation phase sequence is derived. When this compensation sequence is actively injected into each RF front-end, it forces the multi-beams, which were originally phase-displaced due to scattering, to achieve strict temporal and spatial alignment at the three-dimensional focal point of the deep tumor. This allows the RF peaks to form a constructive coherent superposition in the target area, effectively offsetting wavefront distortion and focal shift caused by tissue inhomogeneity. The diameter of the half-power focusing area is constrained within the expected range, significantly improving the energy deposition accuracy and efficiency in the deep target area, while avoiding large-area overheating damage to superficial normal tissue caused by standing wave fields.

[0019] 2. By comparing the deviation between the planned and actual temperature rise rates of the target area, a nonlinear mapping logic based on the natural logarithmic envelope is used to generate a radio frequency power boost compensation coefficient. This coefficient replaces the traditional linear product amplification with the smoothing characteristics of a logarithmic function. When the actual temperature rise rate lags significantly behind the planned value, it provides strong boost kinetic energy to counteract the energy deficit caused by blood flow heat dissipation; while when the temperature rise rate gradually approaches the target, the boost gradient decays logarithmically, avoiding overshoot and underheating oscillations caused by fixed linear feedback gain. This coefficient is applied to the base drive level of each channel of the phased array, comprehensively compensating for the absolute radiation energy sacrificed to eliminate destructive interference, ensuring that the target area can continuously and stably reach and maintain the effective thermal dose required to induce tumor cell apoptosis, achieving highly safe and precise temperature control.

[0020] 3. Based on real-time VSWR monitoring values, an exponential mapping mechanism with negative penalty characteristics is introduced to calculate the safety tolerance assessment factor for each channel. When the channel VSWR deteriorates and approaches the preset red line, this factor decays exponentially, driving a forced reduction in its RF drive level. Simultaneously, according to normalized weights, the reduced power is proportionally and safely transferred and superimposed onto channels with excellent impedance matching and sufficient redundancy. During this power reconfiguration process, by maintaining a constant total energy flux through normalized distribution, the overall electromagnetic dose integral density at the target area center remains unaffected. Thus, without disrupting the established precise focusing phase field, the risk of severe damage to the solid-state power amplifier module due to a surge in local VSWR reflection power is effectively avoided, improving the reliability of all-day operation under complex conditions. Attached Figure Description

[0021] Figure 1 The schematic diagram illustrates the steps of a deep frequency conversion phased array focused tumor hyperthermia method according to the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. S1: During the radio frequency transmission pause period, each channel is switched to listening mode to collect backscattered signals containing tissue dielectric change characteristics.

[0024] In the actual operation of deep frequency conversion phased array focused tumor hyperthermia, the heterogeneous structure of living tissue is composed of media such as skin, fat, muscle and bone, and the dielectric constant of each interface exhibits drastic spatial abrupt changes.

[0025] When multiple radio frequency electromagnetic waves are emitted toward the target area according to a preset phase, this dielectric abrupt change will inevitably cause significant electromagnetic multipath scattering and secondary refraction effects, causing a large amount of radio frequency energy to carry spatial difference information of tissue electrical parameters back to each radiation unit in the opposite direction of the incident path.

[0026] Traditional equipment treats such reflected signals as harmful interference and discards them directly without extraction and analysis, thus losing the opportunity to perceive the non-ideal characteristics of the transmission path.

[0027] Therefore, during the duty cycle pauses of the continuous radio frequency transmission pulses, each channel is switched from high-power transmission mode to high-sensitivity listening mode to capture the backscattered signal generated by dielectric abrupt changes at deep tissue interfaces.

[0028] During the rest period window Inside, the RF front-end circuits corresponding to each RF feed channel perform transmit / receive switching actions, disconnecting the power amplifier link and connecting the low-noise amplifier and down-conversion link, so that the array antenna unit works in passive receiving mode.

[0029] At this time, the channel exist to The received complex envelope signal within the time range is The signal is a complex sequence in the time domain in which the amplitude and phase change continuously over time.

[0030] Since the absorption, scattering, and reflection characteristics of radio frequency energy by heterogeneous living tissue are directly related to the spatial distribution of its dielectric constant, The system fully encapsulates the amplitude attenuation and phase shift hysteresis information caused by interface reflections at different depths along the transmission path of the channel.

[0031] All data collected synchronously Each channel , forming a dimension Time-frequency domain backscattering complex impedance characteristic set ,in and These represent the start and end times of the effective acquisition segment within the rest period, respectively, in seconds (s).

[0032] The complex envelope of the backscattered signal exhibits differentiated characteristics across different channels, the degree of which depends on the spatial location of the tissue interface faced by each radiating element and the inconsistency of the gradient of the dielectric constant abrupt change.

[0033] For example, channel If the corresponding transport path includes a thicker interface between the fat and muscle layers, its A relatively strong reflection peak often appears in the channel, accompanied by a group delay hysteresis on the order of tens of picoseconds (ps); while the channel If the proportion of bones in the path is high, the attenuation of the echo signal will increase significantly, and the phase rotation angle will also increase due to the difference in propagation constant.

[0034] These differences, presented in the form of complex envelopes, essentially constitute the fingerprint information of the tissue dielectric mutation characteristics in the channel dimension.

[0035] Instead of averaging or thresholding such signals, their dynamic range of amplitude and periodic winding characteristics of phase are fully preserved to ensure that the electrical details of the layered interfaces within the tissue are not lost.

[0036] S2: Construct a cross-channel coherent crosstalk evaluation matrix based on the backscattered signal, and obtain the residual compensation phase sequence for canceling destructive interference based on the matrix to reshape the coherent superposition of the beam in the target area.

[0037] The process involves completing the backscattered signal set containing the characteristics of tissue dielectric mutation during the resting period. After acquisition, the multipath scattering and secondary refraction effects caused by the heterogeneous medium of the living body have caused the phase relationship of the received signals of each channel to deviate from the preset empty carrier beam synthesis conditions.

[0038] Because the tissue interfaces faced by each radiating unit differ in dielectric constant gradient, layer thickness, and spatial location, any two channels... With channel The captured and Between them, there must exist a parasitic propagation delay parameter determined by the non-uniform propagation path of the medium. And phase offset.

[0039] This phase inconsistency between channels is the direct physical cause of non-orthogonal coherent crosstalk between multiple radio frequency beams in the target area, which in turn generates a destructive interferometric standing wave field.

[0040] To obtain the strength and phase relationship of this cross-channel coherent crosstalk, the values ​​of each channel are no longer considered. Instead of treating them as isolated scalars, a channel cross-correlation topology algorithm with time-aligned properties is executed. This algorithm applies to any channel pair. , channel Backscattering complex envelope With channel Backscattered complex envelope via time shift Compensated conjugate signal Effective scoring window during the rest period The relevant operations are performed and multiplied by an attenuation weighting factor used to characterize spatial distance loss. The channel cross-correlation topology algorithm is a well-known technique and will not be described in detail here.

[0041] Thus, a dimension is constructed. The cross-channel coherent crosstalk assessment matrix, its matrix node elements The calculation process is defined by the following integral: in, It is a channel At any moment The amplitude of the received complex envelope signal, in volts per meter (V / m); It is a channel At any moment The conjugate value of the received complex envelope signal, in volts per meter (V / m); For channel With channel The dynamic propagation delay parameter caused by the difference in physical distribution and dielectric environment is calculated based on the distance between the two channels and the equivalent electromagnetic wave velocity within the tissue, and the unit is picosecond (ps). It is an exponential term used to compensate for the additional attenuation caused by spatial distance; As an energy attenuation factor for biological tissue media, it is determined by a preset frequency-varying attenuation model based on the current operating frequency. Dynamically acquired, unit is per centimeter ( ); The equivalent spatial electromagnetic propagation distance between two feed radiation units is expressed in centimeters (cm). to This constitutes an effective integration time window during the rest period. Experience value , Experience value All units are microseconds (μs). Parameters and As a hyperparameter, it can be adjusted by the implementer based on the actual tissue model and array geometry. In addition, the difference between the start and end times of the effective integration time window is preferably selected in the microsecond range, and the specific value can be adapted according to the radio frequency pulse duty cycle and the maximum round-trip time of the tissue echo.

[0042] The frequency-varying attenuation model can employ the classic multi-pole Cole-Cole dielectric model, based on a tissue parameter lookup table from a fundamental biological dielectric database, by acquiring the current operating frequency in real time. Substituting the loss tangent formula into Maxwell's electromagnetic wave propagation equation, the energy attenuation factor of the corresponding tissue at the current frequency is dynamically calculated. This is a conventional technical method in this field and will not be elaborated further.

[0043] This integral operation effectively suppresses uncorrelated random white noise components in the received signals of each channel by multiplying the conjugate of the time-domain signals and integrating over time, and effectively purifies and solidifies the deterministic physical interference components caused by reflections from heterogeneous interfaces.

[0044] Matrix elements As a complex number, its modulus is The channel was calculated With channel The energy intensity of coherent crosstalk between them is dimensionless; its complex deflection angle This indicates the amount of radio frequency phase lead or lag between the channel pairs due to a change in tissue dielectric, expressed in radians (rad).

[0045] The entire matrix This constitutes a complete full-dimensional spatial topology correlation map, which comprehensively reflects the non-orthogonal coherent interference state between all channels in the array.

[0046] In obtaining the cross-channel coherent crosstalk evaluation matrix Then, it is used as a constraint condition for reverse analysis, and reverse derivation is carried out with the goal of forcing the coherent superposition of the target area beams.

[0047] The essence of destructive interference is that the beams of each channel cannot form a consistent wavefront at the target point due to phase clutter. Therefore, the key to counteracting destructive interference is to actively inject a pre-bias amount that is opposite in phase to the original crosstalk into each channel.

[0048] Directly extract the matrix By using the complex deflection information of the diagonal and off-diagonal elements, and solving a constraint equation that maximizes the vector superposition magnitude of the synthesized field in the target region, a set of residual compensation phase sequences designed to actively disrupt anomalous coherent superposition states deviating from the target region is derived. .

[0049] This sequence is a sequence of length [length]. The phase angle vector, its first... element That is, it needs to be injected into the channel. The phase correction amount of the RF front end, measured in radians (rad).

[0050] Will The sequence is written into the direct digital frequency synthesizer of each channel, forcibly realigning the peaks of the multi-channel radio frequency energy in time and space at the expected three-dimensional focal point of the deep tumor.

[0051] S3: Obtain the deviation between the planned target area temperature rise rate and the actual temperature rise rate, and convert the deviation into a global RF power boost compensation coefficient through nonlinear logarithmic envelope mapping to compensate for the energy drop caused by phase reshaping.

[0052] The cross-channel phase reshaping step is performed, and the residual compensated phase sequence is generated by a direct digital frequency synthesizer. After injection into each channel, the coherent superposition of multiple radio frequency beams in the target area is reconstructed, and the focus distortion is corrected.

[0053] However, this forced phase intervention caused the angle between the superposition of radio frequency power vectors between the array radiation units to deviate from the optimal value under no-load conditions, resulting in a significant drop in the total synthesized radio frequency electric field intensity at the center of the target area. The actual electromagnetic energy deposited at the lesion site was far lower than the treatment plan requirements.

[0054] To restore the required thermal radiation level of the target area after phase refocusing, a target scalar of dynamic thermal radiation rate of increase for a specific tumor volume is extracted from 3D images and treatment prescriptions using a medical preprocessing planning module. The unit is degrees Celsius per minute (°C / min), with an empirical value of 3.0. This parameter defines the expected rate of temperature rise in the lesion area during the current treatment phase. Its value is based on the biological heat conduction equation and is preset in combination with tumor size, depth and pathological type. It can be adjusted by the implementer according to the specific implementation scenario.

[0055] At the same time, a network of high-precision fiber optic temperature probes with no electromagnetic interference, pre-embedded in the target area and its periphery, is used to... The Hertz sampling rate continuously captures real-time temperature values ​​at discrete points in space, and after differential calculation, generates a real-time dynamic thermal radiation temperature rise rate monitoring scalar for the target area. The unit is also degrees Celsius per minute (°C / min), where the sampling rate of 10 Hz is an empirical value that can be adjusted by the implementer according to the specific sampling scenario.

[0056] The value of [value] remains consistently lower than [value] due to the dynamic fluctuations in the heat dissipation rate of in vivo blood perfusion and the aforementioned drop in synthetic field strength. This forms a deviation that characterizes the current degree of thermal radiation hysteresis.

[0057] When faced with this deviation, traditional linear feedback control directly uses a power amplification factor proportional to the deviation amplitude to increase the RF drive level.

[0058] This method is prone to overshoot oscillations when the temperature approaches the target due to the nonlinear time-varying characteristics of heat dissipation from blood flow in living tissue, and it cannot provide a sufficiently rapid pull-up force when the deviation is extremely large.

[0059] To avoid this high-frequency oscillation problem, a nonlinear mapping algorithm based on the natural logarithm envelope is used to smoothly transform the hysteresis of the temperature rise rate into a global RF power boost compensation coefficient. This coefficient directly affects the baseline of the original driving amplitude for all channels in the array. The calculation formula for this mapping model is: In the formula, The thermodynamic response sensitivity constant factor, which is pre-calibrated, determines the degree of compensation intervention for deviations in the temperature rise rate; its empirical value is [value missing]. , dimensionless.

[0060] parameter and These are the boundary values ​​of the anti-collapse smoothing constants added to the numerator and denominator respectively, used to prevent the collapse of the smoothing constants when the collapse occurs. or Arithmetic overflow occurs when the instantaneous noise in the measurement reaches zero, causing the logarithmic argument or denominator to return to zero. Experience value The unit is degrees Celsius per minute (°C / min); Experience value The unit is degrees Celsius per minute (°C / min).

[0061] Among them, parameters , and All parameters are over-parameters and can be adjusted by the practitioner according to the blood perfusion characteristics and temperature measurement accuracy of the specific treatment site.

[0062] The core mechanism of this logarithmic envelope mapping is as follows: When much smaller That is, when the target area temperature rise is significantly delayed, the ratio of the two items is... For a much larger The value of , its natural logarithm Output a sufficiently large positive number, adjusted by the agility factor. Scaling and constant Adding them together makes Become significantly greater than The pull-up factor drives a significant surge in total RF power to quickly catch up with the heat deficit.

[0063] And when the actual temperature rise rate Gradually approaching the planned value At that time, the ratio approaches Its logarithm approaches , Smoothly converges to .

[0064] This logarithmic convergence characteristic ensures that the power growth gradient decays exponentially in the compensation tail stage, thereby achieving extremely smooth control of the target area temperature rise when it approaches the target value, eliminating the repeated oscillations of overheating peak and underheating fall caused by linear regulation.

[0065] Calculated global RF power boost compensation coefficient As a numerical scalar, dimensionless, it represents the total dynamic amplification factor that must be applied on the existing power baseline to compensate for the energy drop caused by phase reshaping.

[0066] S4: Monitor the VSWR of each channel in real time, and calculate the safety tolerance assessment factor of each channel based on the VSWR and the preset warning threshold using an exponential penalty mechanism.

[0067] Based on the global RF power boost compensation coefficient output in step three When the RF drive amplitude of each channel is uniformly amplified, the impedance mismatch at the feed end ports varies significantly due to the different gradients of dielectric constant and thickness distribution of the living tissue faced by each radiating element. This results in a significant difference in the forward standing wave ratio at the end of some channels. After the power is boosted, it rises sharply, approaching or even exceeding the physical tolerance limit of the solid-state power amplifier hardware. Therefore, it is necessary to establish a mechanism to calculate and evaluate the hardware load tolerance margin of each channel.

[0068] A directionally coupled bidirectional sampling module is integrated between the end solid-state power amplifier and the antenna radiating element of each RF feed channel to extract the forward incident wave voltage amplitude in real time at a millisecond refresh rate. With the amplitude of the reflected wave voltage Both are measured in volts (V) and are based on the definition of standing wave ratio. Simultaneously calculate the first The instantaneous standing wave ratio (VSWR) of the channel. As a dimensionless scalar, this value directly reflects the severity of impedance mismatch and the intensity of reflected power accumulation in the channel.

[0069] Get all After obtaining the VSWR sequence of each channel at the current moment, a preset VSWR redline warning threshold constant is introduced for each channel. and daily health baseline standing wave ratio parameters Both are dimensionless constants, among which The upper limit of the VSWR is taken from the empirical value corresponding to the rated withstand reflection power of the solid-state power amplifier module. , Take the empirical value of the upper limit of slight fluctuations allowed by the channel under normal matching conditions. .

[0070] When a certain channel Not greater than When the channel is in the healthy impedance range, no penalty measures are required, and its safety tolerance assessment factor... Maintain as a dimensionless unit value ,when Greater than At this point, the reflected power has already posed a threat to the power amplifier, triggering a tolerance calculation based on an exponential penalty mechanism to determine the degree of danger that the channel is approaching the hardware damage boundary.

[0071] The exponential penalty mechanism is defined using piecewise logic: when hour, ;when Force setting .

[0072] In this mechanism, when hour This indicates that the channel is safe; when the VSWR is in the middle range, the numerator of the fractional term represents the extent to which the VSWR deviates from the healthy baseline, and the denominator is the difference between the warning threshold and the current VSWR plus a positive decimal excluding zero. The meaning of "composition" is that the closer one gets to the red line. The smaller the denominator, the more linearly the penalty fraction increases; when the red line is reached or crossed, the factor returns to zero, directly cutting off power allocation to high-risk channels, amplifying the risk penalty exponentially. The sensitivity to this deteriorating trend is determined empirically by a value of [value missing]. Dimensionless, excluding zero-tolerance constants Prevent when Approaching Arithmetic overflow occurs when the denominator is reduced to zero; empirically, the value is taken as... Dimensionless For the natural constant An exponential function with base 0.

[0073] After exponential penalty mapping, the channel's security tolerance assessment factor For the range of values dimensionless weights, when When within the healthy range This indicates that the channel has a full power carrying capacity margin, when Rise and approach hour, Rapidly decaying and approaching This indicates that the channel is extremely close to its hardware breakdown limit, and its power carrying capacity should be significantly reduced. (Parameters) , , and These are all hyperparameters, which can be adjusted by the implementer based on the solid-state power amplifier module model, the nominal value of the rated tolerable reflected power, and the safety boundaries for clinical treatment.

[0074] At this point, the safety tolerance calculation for each channel under the current impedance state has been completed and evaluated, resulting in a dimension of [missing information]. weight vector .

[0075] S5: Reduce the power of channels with low security tolerance and allocate the reduced power to secure channels according to the proportion of the security tolerance evaluation factor to maintain a constant total energy flux.

[0076] Obtain the security tolerance evaluation factor vector for each channel. ,in For values ​​in the interval The dimensionless weights.

[0077] At the current moment, the global RF power boost compensation coefficient output in step two is... To compensate for the energy drop caused by phase reshaping, it is required to uniformly amplify the drive power of all channels.

[0078] However, some channels are affected by the standing wave ratio (SWR). Approaching or even breaching the hardware's protection threshold ,That It has decayed dramatically and is much smaller than This indicates that the power margin that the channel can safely carry is almost exhausted. If these channels with low safety tolerances are directly subjected to... If the power factor is fully increased, the reflected power will exceed the tolerance limit of the solid-state power amplifier, leading to irreversible hardware damage.

[0079] Therefore, it is necessary to study the classics. The boosted total power is then non-uniformly redistributed among the array channels.

[0080] The calculation of power redistribution follows the zero-sum rebalancing principle and requires that the preset initial power reference values ​​of each channel remain equal, that is, the initial power of all channels is equal. .

[0081] First, adjust the global RF power boost compensation coefficient. Compared with this initial power reference value Multiplying these yields the boosted power reference that a single channel should carry under ideal compensation conditions. This multiplication operation inherits the global energy boosting requirement necessary to counteract target energy collapse.

[0082] Subsequently, a normalized weight allocation factor based on the safety tolerance assessment factor is introduced to... Reshape the terminal physical radio frequency drive transmit target power for this channel .

[0083] The calculation process is as follows: In the formula, For the first allocation The final transmit power of the channel, measured in watts (W); A unified initial power reference preset value for all channels, in watts (W); The global RF power boost compensation coefficient obtained in step three is dimensionless; For the first The security tolerance assessment factor for the channel is dimensionless. For all The algebraic sum of the security tolerance assessment factors for each channel, dimensionless; The total number of channels in the array is dimensionless.

[0084] Normalized weight term This achieves a redistribution of energy load, which means that, without changing the constraint of the total energy flux integral, the power share that high-risk channels cannot bear is proportionally transferred to healthy channels with high safety margins.

[0085] When a channel's security tolerance assessment factor When the channel attenuates rapidly due to an excessively high VSWR, the normalized weights of the channel... will be significantly smaller than , making Far below the power benchmark after the boost This means achieving a significant power reduction for the channel.

[0086] Conversely, for those in the healthy resistance range, Maintain as or close to The secure channel has a significantly increased normalized weight. Will surpass it This allows it to carry the transferred energy.

[0087] Because the normalization factor strictly follows The mathematical identity is given, and the initial power references for each channel are equal, so the total radio frequency energy flux transmitted by the entire array is maintained at [value]. This ensures that the total energy integral flux fed into the target area remains absolutely constant before and after dynamic load reconfiguration, preventing energy deficits due to power reduction in high-risk channels and the introduction of excess energy due to power enhancement in safe channels. In actual operation, steps S1 to S5 are executed cyclically during treatment. Backscatter signal acquisition is synchronized with the radiofrequency pulse pause (acquisition window approximately 40 μs), temperature monitoring is updated at a 10 Hz sampling rate, and VSWR monitoring and power redistribution are adjusted in real time at millisecond refresh rates. Thus, an optimal solution is established within the limits of hardware safety constraints and therapeutic energy requirements, ultimately achieving precise global energy control during deep-conversion phased array focused tumor therapy.

[0088] To verify the applicability of this method under different organizational characteristics and equipment conditions, the following specific embodiments are provided. Each embodiment strictly follows the control logic from S1 to S5, only with different initial conditions and monitoring data.

[0089] Example 1: Standard Uniform Tissue Condition. This example uses a 16-channel phased array as an example, with muscle tissue as the target area. During the radio frequency transmission pause, the amplitude fluctuation of the backscattered signal collected by each channel is less than 3%, and the phase hysteresis difference does not exceed 5 degrees. In the constructed cross-channel coherent crosstalk evaluation matrix, the mean magnitude of the off-diagonal elements is less than 0.12, indicating good tissue uniformity and weak inter-channel interference. At this time, the deviation between the planned temperature rise rate and the actual temperature rise rate is 0.4℃ / min, and the global radio frequency power boost compensation coefficient calculated by nonlinear logarithmic mapping is approximately 1.10. The initial VSWR of each channel is between 1.1 and 1.3, and the safety tolerance evaluation factor is maintained at a dimensionless unit value of 1. The channel is in the healthy impedance range and does not require triggering exponential penalties. The system distributes power uniformly according to normalized weights, and the total energy flux remains unchanged. Measurements after treatment show that the focal offset is less than 2 mm, and the superficial skin temperature rise does not exceed the safety limit.

[0090] Example 2: Fat Layer Dominant Scenario. In this example, the RF beam needs to penetrate a thick fat layer. Multiple channels in the backscattered signal exhibited amplitude attenuation exceeding 40% and delays of tens of picoseconds. The constructed evaluation matrix showed that the phase deviation of some channel pairs exceeded 25 degrees, the mode length increased to 0.35, and the target area energy dropped by approximately 18% after phase reshaping. At this time, the actual temperature rise rate lagged behind the planned value by 1.1℃ / min, and the global RF power boost compensation coefficient calculated by nonlinear logarithmic envelope mapping was approximately 1.35. Channels 6 and 9 experienced a standing wave ratio increase to 1.75 due to reflection from the fat-muscle interface, and the safety tolerance evaluation factor decreased to 0.68 after calculation using the exponential penalty formula. The system automatically redistributed the reduced power to channels with good impedance matching according to their weights. The total energy flux remained constant, and the target area temperature rise steadily reached the target value within the preset time without overheating fluctuations.

[0091] Example 3: Skeletal Dominance Scenario. This example simulates a high-density skeletal region within the RF beam path. The overall amplitude of the backscattered signal is weak, but the phase shift is significant. The evaluation matrix reflects a large fixed phase shift difference between channels, with destructive interference mainly concentrated at the target edge. The planned and actual temperature rise rate deviation is 0.9℃ / min, and the calculated global RF power boost compensation coefficient is approximately 1.27. Multiple channels have VSWRs ranging from 1.8 to 2.5. After incorporating an exponential penalty mechanism, the safety tolerance evaluation factor exponentially decreases from 0.62 to 0.03 as the VSWR increases. Based on this weight distribution, the system significantly reduces the drive level of high-risk channels and concentrates power on healthy channels with higher safety tolerance. After phase reshaping and power reconfiguration, the focus energy concentration returns to the expected level, and the temperature rise of normal tissues surrounding the bone is controlled within a safe range.

[0092] Example 4: Single-channel hardware failure scenario. This example simulates the aging of the power amplifier module in channel 3, where the backscattered signal amplitude is less than 15% of the normal value and the VSWR remains above 3.0. Based on the piecewise exponential penalty logic, the channel's safety tolerance assessment factor is forcibly reset to zero. The system immediately cuts off the transmit power of channel 3 and redistributes the power carrying capacity to the remaining 15 healthy channels according to the safety tolerance normalized weight. The total energy flux remains constant throughout the treatment, the three-dimensional distribution of the focal spot shows no significant distortion, and the equipment does not trigger hardware protection shutdown, ensuring the continuity of treatment.

[0093] This invention also discloses a deep frequency conversion phased array focused tumor hyperthermia system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the deep frequency conversion phased array focused tumor hyperthermia method of this invention.

[0094] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0095] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as resistive random access memory (DRAM), dynamic random access memory (DRAM), static random access memory (SRAM), etc., or any other medium that can be used to store desired information and can be accessed by an application program, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device.

Claims

1. A deep-field frequency conversion phased array focused tumor hyperthermia method, characterized in that, include: During the radio frequency transmission pause, each channel is switched to listening mode to collect backscattered signals containing tissue dielectric mutation characteristics; Based on the backscattered signal, a cross-channel coherent crosstalk evaluation matrix is ​​constructed, and a residual compensation phase sequence for canceling destructive interference is obtained according to the matrix to reshape the coherent superposition of the beam in the target area. The deviation between the planned target area temperature rise rate and the actual temperature rise rate is obtained, and the deviation is converted into a global RF power boost compensation coefficient through nonlinear logarithmic envelope mapping to compensate for the energy drop caused by phase reshaping. Under the premise that the initial power reference values ​​of each channel are preset and equal, the standing wave ratio of each channel is monitored in real time. Based on the standing wave ratio and the preset warning threshold, the safety tolerance evaluation factor of each channel is calculated using an exponential penalty mechanism. For channels with low security tolerance, power is reduced and the reduced power is allocated to secure channels according to the security tolerance assessment factor to maintain a constant total energy flux.

2. The method for deep frequency conversion phased array focused tumor hyperthermia according to claim 1, characterized in that, The acquisition of backscattered signals containing tissue dielectric mutation characteristics includes: During the radio frequency transmission pause, each channel is switched from the transmission state to the high-sensitivity reception state to capture the weak radio frequency detection signals backscattered at the boundary of heterogeneous deep tissues, and to obtain the time-frequency domain backscattering complex impedance feature set of each channel. The time-frequency domain backscattering complex impedance feature set includes the amplitude attenuation and phase shift hysteresis information of the signal.

3. The deep frequency conversion phased array focused tumor hyperthermia method according to claim 2, characterized in that, The construction of the cross-channel coherent crosstalk evaluation matrix based on the backscattered signal includes: A channel cross-correlation topology algorithm with time-alignment characteristics is adopted. The cross-channel coherent crosstalk evaluation matrix is ​​constructed using the time-frequency domain backscattering complex impedance feature set. The elements in the cross-channel coherent crosstalk evaluation matrix are complex numbers. The deflection angle of the complex number represents the radio frequency phase lead or lag caused by non-uniform transmission of the organization dielectric between any two independent channels. The modulus of the complex number quantifies the energy intensity of the coherent crosstalk between channels.

4. The method for deep frequency conversion phased array focused tumor hyperthermia according to claim 1, characterized in that, The step of obtaining the residual compensation phase sequence for canceling destructive interference based on the matrix includes: Using the cross-channel coherent crosstalk evaluation matrix as a reverse analytical constraint, the complex deflection information of the matrix elements is extracted, and the destructive interferometric residual compensation phase sequence for offsetting the destructive standing wave effect is derived in reverse.

5. A deep-field variable frequency phased array focused tumor hyperthermia method according to claim 1, characterized in that, The process of converting the deviation into a global RF power boost compensation coefficient through nonlinear logarithmic envelope mapping includes: Using a natural logarithmic envelope mapping, with the ratio of the planned target area temperature rise rate to the actual temperature rise rate as the independent variable, a response agility factor and a collapse-prevention smoothing constant to prevent denominator singularities are introduced to calculate the global RF power boost compensation coefficient. This provides boost momentum when the actual temperature rise rate lags behind the planned temperature rise rate, and the boost gradient decays logarithmically when the actual temperature rise rate approaches the planned temperature rise rate.

6. The method for deep frequency conversion phased array focused tumor hyperthermia according to claim 1, characterized in that, The global RF power boost compensation coefficient is derived by multiplying a constant 1 by the natural logarithm of the ratio of the sum of the planned target area temperature rise rate and the first anti-collapse smoothing constant to the sum of the actual target area temperature rise rate and the second anti-collapse smoothing constant.

7. The method for deep frequency conversion phased array focused tumor hyperthermia according to claim 1, characterized in that, The calculation of the security tolerance evaluation factor for each channel using an exponential penalty mechanism includes: When the VSWR of a single channel exceeds the healthy baseline value, an exponential penalty is triggered. The safety tolerance assessment factor is defined using piecewise logic: when the VSWR of a single channel is greater than the health baseline value and less than the preset warning threshold, the safety tolerance assessment factor is an exponential function with the natural constant e as the base, and its exponential term is the penalty amplification exponent multiplied by the difference between the VSWR and the health baseline value divided by the difference between the preset warning threshold and the VSWR plus the negative value of the zero tolerance constant; when the VSWR is greater than or equal to the preset warning threshold, the safety tolerance assessment factor is forcibly set to zero.

8. The method for deep frequency conversion phased array focused tumor hyperthermia according to claim 1, characterized in that, The power reduction for channels with low security tolerance, which involves allocating the reduced power to secure channels according to the security tolerance assessment factor, includes: The global RF power boost compensation coefficient is multiplied by the initial power reference value of each channel, then multiplied by the normalized weight of the security tolerance assessment factor of each channel in the sum of all channel security tolerance assessment factors, and multiplied by the total number of channels to obtain the final target power of each channel.

9. A deep-field frequency conversion phased array focused tumor hyperthermia method according to claim 1, characterized in that, Maintaining a constant total energy flux is achieved through normalized weight allocation.

10. A deep-field variable frequency phased array focused tumor hyperthermia system, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a deep frequency conversion phased array focused tumor hyperthermia method according to any one of claims 1-9.

Images