Focused sound field generation system and method based on three-dimensional transcranial ultrasonic imaging

Abstract

The invention discloses a focused sound field generation system and method based on three-dimensional transcranial ultrasonic imaging, relates to the technical field of ultrasonic imaging and sound focusing, and solves the technical problem that significant spatial deviation exists between a focusing focus and a target spot. A three-dimensional full-waveform inversion and / or tomographic inversion algorithm is adopted to jointly invert parameters such as sound velocity, attenuation and density of the skull and the brain tissue, and an individualized three-dimensional acoustic atlas is constructed. And on the premise of meeting the safety threshold parameter, performing multi-target optimization on the array element phase, amplitude, frequency band and emission time sequence of the multilayer ring array according to the three-dimensional acoustic atlas to obtain an individualized focusing parameter and an acoustic dose parameter for the predetermined brain region. Individualized modeling and imaging-regulation closed loop integration based on ultrasonic self-imaging are achieved, and low-cost, repeatable and real-time three-dimensional transcranial imaging and accurate nerve regulation guidance can be provided in a bedside environment.

Description

Technical Field

[0001] This application relates to the fields of ultrasound imaging and acoustic focusing technology, and in particular to a focused acoustic field generation system and method based on three-dimensional transcranial ultrasound imaging. Background Technology

[0002] Neuromodulation techniques are a class of methods that use physical or chemical means to intervene in the central nervous system and regulate the activity of neurons to improve patients' motor, sensory, and cognitive functions. Currently, commonly used invasive methods of neuromodulation include deep brain stimulation (DBS), which, while having clear therapeutic effects, require electrode implantation and carry surgical risks, infection risks, and long-term maintenance costs. As a non-invasive alternative, transcranial focused ultrasound (tFUS) has received widespread attention in recent years. It can focus ultrasound energy on specific areas deep in the brain without opening the skull, achieving reversible and controllable neuromodulation, while also possessing high spatial resolution and depth of modulation.

[0003] In existing technologies, to improve the accuracy of transcranial focusing, MRI or CT imaging is typically used to provide anatomical information and skull structural parameters. This information is used to guide transducer placement and target selection, as well as for sound field simulation and phase correction. However, MRI equipment is expensive and has long scan times, making it less suitable for emergency and bedside settings. CT imaging involves ionizing radiation, posing a radiation exposure risk to patients, and has limited soft tissue contrast resolution, increasing the overall burden on both patients and healthcare professionals.

[0004] Furthermore, while some existing solutions attempt to combine skull information provided by MRI or CT scans for time reversal or ray tracing correction, they often simplify the focusing process to linear propagation in sound field simulations, failing to fully consider complex fluctuations such as multiple scattering, strong refraction, and mode conversion under transcranial conditions. This can still lead to significant spatial deviations between the focal point and the target. Moreover, these solutions lack real-time visualization and closed-loop assessment of in vivo sound field distribution and physiological responses, making it difficult to detect risks such as focus drift and abnormal energy accumulation in a timely manner.

[0005] In terms of imaging and navigation, current transcranial ultrasound applications mainly rely on two-dimensional temporal window exploration, which has a limited field of view and is heavily dependent on the window. It can only provide two-dimensional cross-sectional information of local brain regions, making it difficult to achieve a unified assessment of the three-dimensional structure and function of the whole brain. Introducing full waveform inversion (FWI) or ultrasound tomography (USCT) based on acoustic wave theory can directly invert the acoustic parameter fields such as sound velocity and attenuation of the skull and brain tissue, which can help achieve integrated imaging and focusing. However, there are still shortcomings in three-dimensional array engineering, CT-free modeling in vivo verification, and closed-loop standardized evaluation.

[0006] On the other hand, the development of ultrafast ultrasound and functional ultrasound (fUS) technologies has made it possible to observe cerebral hemodynamics at sub-second temporal resolution, which can be used to assess neural activity, functional connectivity, and regulatory responses. However, most existing transcranial ultrasound systems separate "imaging" and "modulation": the imaging and modulation devices are independent, spatial registration is complex, and it is difficult to use functional imaging results in a timely manner for target screening and dose adjustment; at the same time, there is a lack of a systematic evaluation system and a unified set of indicators covering "phantom-ex vivo-small animal / preclinical", which is not conducive to the comparability between different systems and multi-center promotion.

[0007] In summary, the existing technology has at least the following problems: (1) Over-reliance on expensive or radiation-emitting imaging methods such as MRI / CT is not conducive to bedside deployment and repeated monitoring; (2) Transcranial acoustic field modeling often uses approximate rays or simplified models, which makes it difficult to accurately handle multiple scattering and mode conversion caused by complex skulls, and there is still a deviation between the focal point and the predetermined target point; (3) The imaging and control links are disconnected, and there is a lack of three-dimensional navigation and real-time closed-loop evaluation based on ultrasound self-imaging; (4) A multi-level verification system and standardized evaluation indicators for three-dimensional transcranial imaging-neuromodulation system have not yet been formed. Summary of the Invention

[0008] This application provides a focused sound field generation system and method based on three-dimensional transcranial ultrasound imaging. The technical purpose is to improve the accuracy of the sound field focusing point, so as to achieve precise control of the intracranial target location, reduce the dependence on MRI / CT, ​​and improve the bedside applicability, safety and scalability of the system.

[0009] The above-mentioned technical objective of this application is achieved through the following technical solution: A focused sound field generation system based on three-dimensional transcranial ultrasound imaging, comprising: A head-mounted three-dimensional ultrasound transducer module (10) is used to fit the subject's head for ultrasound transmission and echo reception. The front-end drive and acquisition module (20) is used to drive each element of the head-mounted three-dimensional ultrasonic transducer module (10) under high voltage, switch between transmission and reception, and acquire multi-channel three-dimensional full-aperture ultrasonic signals. The imaging and acoustic modeling module (30) inverts the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion of the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic map covering the whole brain. Under the constraints of the three-dimensional acoustic map, it performs three-dimensional structural imaging and three-dimensional functional ultrasound imaging. The focusing control module and beam optimization module (40) construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module (10) are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

[0010] Preferably, the focused sound field generation system further includes: The closed-loop monitoring and safety control module (50) identifies focal point position deviation, energy deposition anomaly and potential cavitation events, and performs online evaluation and adaptive adjustment of focusing parameters based on focal point position deviation, energy deposition anomaly and potential cavitation events, and triggers automatic power reduction or shutdown protection when safety parameters exceed limits. The human-computer interaction and navigation module (60) is used to display the three-dimensional acoustic atlas, the three-dimensional structural imaging and the three-dimensional functional ultrasound imaging in three dimensions, support users to preset or interactively select transcranial nerve modulation target points, and realize the alignment and registration with external images.

[0011] Preferably, the three-dimensional wave equation is expressed as: ; in, Indicates spatial location; , represents angular frequency; Represents frequency domain sound pressure level; Indicates the speed of sound; Indicates density; and Both represent power-law decay parameters; Indicates the sound source; Represents the imaginary unit; Represents the Laplace operator; The three-dimensional acoustic map is represented as follows: ; ; in, Indicates data mismatch items, Represents the regularization term. Indicates the measured sound signal. Represents analog sound signals. Indicates the launch index. Indicates the receiving index. Indicates normalized cross-correlation; Indicates the parameters to be inverted; Priors are represented, including image priors and velocity priors; , , , All represent weighting coefficients; Represents the sound field Spatial gradient; The constraints in the inversion and joint inversion processes are expressed as follows: ; in, Indicates the first Wheel parameters, Indicates the first One frequency band, This represents the overall objective function.

[0012] Preferably, the multi-objective optimization model is expressed as: , , ; , ; ; , , , ; in, Represents the complex sound pressure vector at each sampling point; This represents the acoustic propagation matrix from the array element to the sampling point; Represents the complex weights of array elements; This represents the propagation matrix corresponding to the target point; Represents the target propagation vector The conjugate transpose of; This indicates the density at the target point. Indicates the speed of sound at the target point; Indicates the mechanical index; The value representing the ultrasonic operating frequency or center frequency in MHz; Indicates the reduction of radiation amplitude; This represents the spatial peak-time average sound intensity after attenuation. This indicates the reduction of sound intensity; Represents the bone window energy proxy matrix; Represents the sidelobe sampling matrix. Represents the sampling matrix of the constraint region. This indicates the sidelobe suppression penalty weight; Indicates the energy load penalty weight for bone windows; Indicates the average time window length; Indicates the maximum power limit; This represents the upper limit threshold for the allowed de-attenuation spatial peak time average acoustic intensity. This indicates the upper limit threshold of sound pressure or energy in the constrained area.

[0013] Preferably, the head-mounted three-dimensional ultrasonic transducer module (10) is an imaging-control integrated three-layer coaxial ring array ultrasonic transducer. The three-layer coaxial ring array ultrasonic transducer includes three layers of coaxial ring arrays. The three-layer coaxial ring arrays include a first ring array layer, a second ring array layer and a third ring array layer distributed axially from top to bottom. The first ring array layer, the second ring array layer and the third ring array layer are all concentric rings with a diameter of 22cm. Each ring array layer contains at least 512 independent array elements.

[0014] Preferably, the circumferential spacing of the three-layer coaxial ring array is 0.5mm to 1mm, and the axial spacing between each ring array layer is 2mm; the array element material is piezoelectric ceramic material or piezoelectric composite material, the effective operating frequency band of the array element covers 0.1MHz to 1.2MHz, and the center frequency is 0.8MHz.

[0015] Preferably, the front-end driving and acquisition module (20) includes a multi-channel transmit / receive and phased array control submodule and a data acquisition and host computer processing submodule electrically connected thereto, wherein: The multi-channel transmit / receive and phased array control submodule is used to output multi-band pulse signals and coded transmit signals, and to perform high-voltage driving, transmit or receive switching on each array element through the multi-band pulse signals and the coded transmit signals; wherein, the multi-band pulse signals include narrowband pulse signals and linear chirped signals, and the coded transmit signals include complementary coded signals; The data acquisition and host computer processing submodule is used to acquire multi-channel three-dimensional full-aperture ultrasonic signals and perform gain control, anti-aliasing filtering, and A / D conversion on the multi-channel three-dimensional full-aperture ultrasonic signals; wherein the sampling rate is not less than 10MHz.

[0016] Preferably, the closed-loop monitoring and safety control module (50) includes: The passive acoustic monitoring and functional imaging submodule is used to acquire the echo phase, passive acoustic monitoring signal and functional ultrasound hemodynamic response of the target area and adjacent areas in real time during transcranial ultrasound neuromodulation. The drift of the echo phase is used to monitor the thermal expansion of the bone window and the change of tissue sound velocity. The passive acoustic monitoring signal is used to monitor potential cavitation events and abnormal scattering. The functional ultrasound hemodynamic response is used to evaluate the neuromodulation effect, so as to identify focal point position shift, abnormal energy deposition and potential cavitation events. The safety monitoring and closed-loop control submodule is used to perform online evaluation and adaptive adjustment of focusing parameters based on focus point position offset, energy deposition anomalies and potential cavitation events, and to trigger automatic power reduction or shutdown protection when safety parameters exceed limits; wherein, the adaptive adjustment includes real-time adjustment of transmit power, pulse repetition frequency, duty cycle and array element phase.

[0017] Preferably, the human-computer interaction and navigation module (60) is integrated into the imaging and acoustic modeling module (30) or the host computer processing terminal or set up separately.

[0018] A method for generating a focused sound field based on three-dimensional transcranial ultrasound imaging, wherein the method is implemented through the aforementioned focused sound field generation system, comprising: The head-mounted three-dimensional ultrasonic transducer module (10) performs multi-band, wide solid angle ultrasonic transmission and echo reception on the subject's head. During the ultrasonic transmission and echo reception process, the front-end drive and acquisition module (20) performs high-voltage drive, transmission or reception switching and multi-channel signal acquisition on each array element of the head-mounted three-dimensional ultrasonic transducer module (10) to obtain multi-channel three-dimensional full-aperture ultrasonic signals covering the subject's head. The imaging and acoustic modeling module (30) inverts the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion of the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic spectrum covering the whole brain. Under the constraints of the three-dimensional acoustic spectrum, it performs three-dimensional structural imaging and three-dimensional functional ultrasound imaging. The focusing control module and the beam optimization module (40) construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module (10) are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

[0019] The above technical solution can achieve at least some of the following technical effects: (1) It realizes the integration of individualized three-dimensional acoustic modeling and imaging-control, reducing the dependence on MRI / CT; (2) It provides three-dimensional structural and functional navigation capabilities for the whole brain, which is beneficial for accurate target screening and efficacy evaluation; (3) Achieving individualized transcranial focusing and dosimetric control based on multi-objective optimization; (4) It has the ability to perform in vivo closed-loop monitoring and adaptive safety control, which can improve the safety and robustness of transcranial nerve modulation. Attached Figure Description

[0020] Figure 1 This is a framework diagram of the focused sound field generation system based on three-dimensional transcranial ultrasound imaging as described in the embodiments of this application; Figure 2 This is a schematic diagram of the structure of the three-layer coaxial ring array ultrasonic transducer described in the embodiments of this application; Figure 3 This is a schematic diagram of the closed-loop process of "imaging-modeling-focusing-monitoring" in the embodiments of this application; In the diagram: 101 - First ring array layer; 102 - Second ring array layer; 103 - Third ring array layer. Detailed Implementation

[0021] To make the objectives, technical solutions, and beneficial effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of this invention. Without departing from the spirit and substance of this invention, those skilled in the art can make various improvements or equivalent substitutions to the following embodiments, all of which should fall within the scope of protection of this invention. For ease of explanation, this application establishes a three-dimensional rectangular coordinate system. Using the approximate center revealed by the subjects as the origin ; Use the position vector for any point in space Unless otherwise specified, "array element number" is indicated by "layer number". Circular serial number It is expressed in the way of "".

[0022] like Figure 1 As shown, the focused sound field generation system based on three-dimensional transcranial ultrasound imaging described in this application is characterized by comprising a head-mounted three-dimensional ultrasound transducer module 10, a front-end drive and acquisition module 20, an imaging and acoustic modeling module 30, and a focusing control module and beam optimization module 40. The front-end drive and acquisition module 20 is electrically connected to the head-mounted three-dimensional ultrasound transducer module 10, the imaging and acoustic modeling module 30 is signal-connected to the front-end drive and acquisition module 20, and the focusing control module and beam optimization module 40 is signal-connected to the imaging and acoustic modeling module 30.

[0023] The head-mounted three-dimensional ultrasound transducer module 10 is used to fit the subject's head for ultrasound transmission and echo reception.

[0024] Preferably, the head-mounted three-dimensional ultrasonic transducer module 10 is an imaging-control integrated three-layer coaxial ring array ultrasonic transducer, such as... Figure 2 As shown, the three-layer coaxial ring array ultrasound transducer comprises three layers of coaxial ring arrays. The array elements of these three layers are approximately centered on the center of the subject's head and include a first ring array layer, a second ring array layer, and a third ring array layer distributed axially from top to bottom. Each of these three ring array layers is a concentric ring with a diameter of 22 cm. Each ring array layer contains at least 512 independent elements, evenly arranged at equal angles along the circumferential direction.

[0025] Preferably, the circumferential element spacing of the three-layer coaxial ring array is 0.5mm to 1mm to suppress grating lobes and improve sidelobe suppression capability. The axial layer spacing between each ring array layer is 2mm. By reasonably selecting the layer spacing, the solid angle coverage can be improved, the condition number of the forward operator can be improved, and the observability of the three-dimensional parameter inversion can be enhanced. The azimuth angle of the array elements in each ring array layer is expressed as: ; in, This refers to the index of the circumpolar array element in this layer. Indicates the first Number of elements in the layered array , For the first The first in the layer The azimuth angle of each array element.

[0026] The array element material is made of piezoelectric ceramic or piezoelectric composite material. The effective operating frequency band of the array element covers 0.1MHz to 1.2MHz, with 0.8MHz as the center frequency. Multi-scale imaging and full waveform inversion are achieved through multi-band excitation. A flexible coupling layer and a waterproof layer can be attached to the front end of the array element to improve acoustic energy transmission efficiency and enhance long-term stability in the coupling medium.

[0027] The electrical interface component is used to reliably connect all array elements to the front-end driver and acquisition module 20. It can use high-density connectors and flexible cables, and work with the channel calibration matrix to achieve channel mapping and management.

[0028] The front-end drive and acquisition module 20 is used to drive each array element of the head-mounted three-dimensional ultrasonic transducer module 10 under high voltage, switch between transmission and reception, and acquire multi-channel three-dimensional full-aperture ultrasonic signals.

[0029] Preferably, the front-end driving and acquisition module 20 includes a multi-channel transmit / receive and phased array control submodule and a data acquisition and host computer processing submodule electrically connected thereto, wherein: The multi-channel transmit / receive and phased array control submodule is used to output multi-band pulse signals and coded transmit signals. These signals are used to drive, transmit, or receive each array element under high voltage, thereby improving the signal-to-noise ratio and bone-penetrating imaging capability while meeting acoustic safety thresholds. The multi-band pulse signals include narrowband pulse signals and linearly chirped signals, and the coded transmit signals include complementary coded signals.

[0030] The data acquisition and host computer processing submodule is used to acquire multi-channel three-dimensional full-aperture ultrasonic signals and perform gain control, anti-aliasing filtering, and A / D conversion on the multi-channel three-dimensional full-aperture ultrasonic signals; wherein, the sampling rate is not less than 10MHz to fully retain the effective information in the 0.1 MHz to 1.2 MHz frequency band.

[0031] The observed echo is represented as: ; ; in, The first The number of launch elements to the first The observed echo, transmitted signal, equivalent impulse response, and noise term of each receiving array element. The echo spectrum after calibration and compensation. This is used to estimate the channel frequency response obtained from calibration.

[0032] The front-end driver and acquisition module 20 is connected to the host computer via a high-speed data interface, such as fiber optic or high-speed Ethernet, to achieve continuous acquisition and buffering of raw data across the entire aperture, and supports trigger control, transmission sequence configuration and safety parameter monitoring.

[0033] The imaging and acoustic modeling module 30 inverts the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion on the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic map covering the whole brain. Under the constraints of the three-dimensional acoustic map, three-dimensional structural imaging and three-dimensional functional ultrasound imaging are performed to provide integrated structural-functional indicators for target screening and efficacy evaluation.

[0034] Specifically, the three-dimensional transcranial acoustic modeling in this application employs a framework based on full-waveform inversion (FWI) and / or ultrasound computed tomography (USCT) to determine the sound velocity in the skull and brain tissue. ,attenuation and density The parameters are jointly inverted to form an individualized three-dimensional acoustic map. To this end, a three-dimensional forward propagation model is first established and simulation data is generated. Then, a multi-objective function is constructed and the parameters are iteratively updated using the gradient method, such as... Figure 3 As shown, the typical steps are as follows: S1: Three-dimensional full-aperture ultrasonic signal acquisition: According to the preset transmission scheme, through the multi-layer ring array grouping and element-by-element excitation of the head-mounted three-dimensional ultrasonic transducer module 10, full-aperture transmission / scattering data covering multiple incident angles and multiple frequency bands are acquired. The measured echo recorded in the r-th receiving channel after the s-th transmission is taken as... , ; ; That is, the first Receive channel during secondary transmission The measured time-domain signal.

[0035] S2: Data Preprocessing and Quality Control The imaging and acoustic modeling module 30 performs time window truncation, DC removal, bandpass filtering, and code demodulation (for chirped or complementary coded emissions) on the raw data, while removing obviously abnormal channels and saturated segments. The preprocessing flow is as follows: ;in, This is the preprocessed measured signal; This represents a preprocessing operator, which includes combinations of time windows, DC removal, bandpass filtering, demodulation, etc.

[0036] S3: 3D Forward Modeling A three-dimensional wave equation is used as the forward model, and absorption and viscosity terms are introduced to characterize the attenuation effect. Finite-difference time-domain (FDTD) or frequency-domain multi-frequency coupling methods are employed at the current speed of sound. ,attenuation and density Under the model, the wave field propagation and data recording of each transmit / receive pair are simulated to obtain a simulated dataset.

[0037] Preferably, the three-dimensional wave equation is expressed as: ; in, Indicates spatial location; , represents angular frequency; Represents frequency domain sound pressure level; Indicates the speed of sound; Indicates density; and Both represent power-law decay parameters; Indicates the sound source; Represents the imaginary unit; This represents the Laplace operator.

[0038] S4: Construction of Multi-Objective Functions and Gradient Calculation: The imaging and acoustic modeling module 30 compares measured data with simulated data, constructs an objective function primarily based on data error, and can enhance sensitivity to waveform deformation by incorporating metrics such as optimal transmission distance, cross-correlation, and time alignment; it calculates the relevant information using either the adjoint state method or an automatic differentiation framework. , , The gradient components provide direction for parameter updates.

[0039] Preferably, the three-dimensional acoustic spectrum is represented as follows: ; ; in, This indicates a data mismatch term (data fitting target / data term). This represents the regularization term / prior constraint term. Indicates the measured sound signal. Represents analog sound signals. Indicates the launch index. Indicates the receiving index. Indicates normalized cross-correlation; Indicates the parameters to be inverted; Priors are represented, including image priors and velocity priors; , , , All represent weighting coefficients; This represents the spatial gradient of the sound field.

[0040] S5: Regularization and Prior Introduction: In the gradient update process, prior and regularization terms consistent with brain anatomy are introduced, including piecewise smoothing or total variational (TV) regularization, tomographic priors, low-rank / sparse constraints, and frequency recursion constraints, to alleviate strong nonlinearity and multiple solutions, and improve inversion stability under strong skull scattering conditions. The constraints in the inversion and joint inversion processes are then expressed as follows: ; in, Indicates the first Wheel parameters, Indicates the first One frequency band, This represents the overall objective function.

[0041] S6: Multi-scale frequency recursive inversion: The inversion is performed sequentially from low frequency to high frequency. The inversion result of each frequency band is used as the initial model for the next frequency band. The large-scale structure is recovered first through a multi-scale strategy, and then high-frequency details are gradually introduced. This achieves a spatial resolution of no more than 1.0 mm to 1.5 mm for the deep region of interest, and the mean square relative error of the sound velocity / attenuation spectrum is preferably controlled within 2% to 5%.

[0042] S7: Inversion Convergence and Spectrum Generation: When the objective function's downward trend flattens out or the parameter update amount is lower than the set threshold, the inversion convergence is automatically determined and the iteration stops, resulting in a three-dimensional sound velocity, attenuation, and density map covering the entire brain. Furthermore, secondary indicators such as acoustic impedance and reflection coefficient can be derived, serving as the basis for subsequent beam optimization and security assessment.

[0043] Preferably, the anatomical structure of deep brain regions can be reconstructed based on the three-dimensional structural imaging, i.e., brain ultrasound; and the power Doppler intensity spectrum, i.e., cerebral blood flow, can be obtained based on the three-dimensional functional ultrasound imaging.

[0044] Preferably, obtaining the power Doppler intensity spectrum based on the three-dimensional functional ultrasound imaging includes: (1) The three-dimensional functional ultrasound image is processed by ultrafast three-dimensional imaging and Doppler estimation technology to obtain a three-dimensional hemodynamic map at the whole brain scale; (2) The clutter components in the three-dimensional hemodynamic spectrum are removed by singular value decomposition filtering to obtain the three-dimensional blood flow spatiotemporal signal. Then, the three-dimensional blood flow spatiotemporal signal is subjected to frequency domain Doppler estimation to obtain the power Doppler intensity spectrum.

[0045] The above process can be represented as follows: , ; ; in, Represents a spatiotemporal data matrix; This represents blood flow-related data after removing clutter components; This represents the matrix obtained from SVD decomposition; Indicates the removal of the previous Individual singular components Indicates Fourier transform, Indicates the Doppler frequency. Represents the power Doppler intensity spectrum; This represents the three-dimensional spatiotemporal signal of blood flow after removing clutter components; Indicates the first A singular value, Indicates the first A left singular vector Indicates the first A right singular vector.

[0046] The focusing control module and beam optimization module 40 are used to construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module 10 are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

[0047] Specifically, after obtaining the individualized three-dimensional acoustic map, this application utilizes the focus control and beam optimization module 40 to accurately predict the energy deposition distribution at the target point under safety constraints such as MI / TI / Ispta.3. The process is as follows: T1: Definition of target point and constraint region: Based on the structural reconstruction and functional imaging results of the imaging and acoustic modeling module 30, the human-computer interaction and navigation module 60 provides a three-dimensional visualization interface for the operator to select the target point to be controlled (such as a specific cortical area or deep nucleus) and specify the constraint area to be protected (such as the temporal bone window, important blood vessels and tissues near the optic nerve).

[0048] T2: Construction of the transport operator: The focusing control and beam optimization module 40, combined with the three-dimensional acoustic spectrum, calculates the complex transfer function or Green's function from each element in the head-mounted three-dimensional ultrasonic transducer module 10 to the target point and the key constraint region, forming a multi-channel-multi-point propagation matrix, expressed as: , , ; in, Represents the complex sound pressure vector at each sampling point; The acoustic propagation matrix from the array element to the sampling point is obtained by discretizing the complex transfer function under acoustic spectrum constraints; Represents the complex weights of array elements; Indicates the complex sound pressure at the target point; This represents the propagation matrix corresponding to the target point; Represents the target propagation vector Hermitian transpose; Indicates the speed of sound at the target point; This indicates the density at the target point.

[0049] T3: Multi-objective optimization solution: Under constraints such as total power, mechanical index MI, thermal index TI, focal volume, sidelobe suppression, and bone window energy load, a multi-objective optimization model is constructed with the objectives of maximizing the target sound pressure and / or sound intensity and minimizing the energy of the sidelobes and constrained regions, as follows: , ; ; , , , ; in, Indicates the mechanical index; The value representing the ultrasonic operating frequency or center frequency in MHz; Indicates the reduction of radiation amplitude; This represents the spatial peak time average sound intensity after attenuation. Indicates the reduction of sound intensity. Represents the bone window energy proxy matrix; Represents the sidelobe sampling matrix. The sampling matrix of the constrained region is obtained by stacking the transmission vectors of non-target points; This represents the sidelobe suppression penalty weight (weighting coefficient); Bone window energy load penalty weight (weighting coefficient); Indicates the average time window length; Indicates the maximum power limit; This represents the upper limit threshold for the allowed de-attenuation spatial peak time average acoustic intensity. This indicates the upper limit threshold of sound pressure (or energy) in the constrained area.

[0050] Then, the complex weights (amplitude and phase) of each array element are solved by methods such as convex optimization, quadratic programming or iterative beamforming to generate single-target or multi-target focusing parameters and corresponding acoustic dose parameters.

[0051] T4: Multi-target and pathway planning: For situations requiring simultaneous or sequential modulation of multiple targets, weight vectors are designed for different targets using time or frequency segmentation, and then optimized and combined on the time or frequency axis to achieve energy allocation and path planning for multiple targets, ensuring the desired neuromodulation effect is achieved while meeting safety constraints.

[0052] After the path planning is completed, the transmission power signal is used to test the target focusing effect and monitor whether the acoustic focusing signal meets the requirements. If not, the process returns to T2 to rebuild the path; if so, subsequent adjustments are made.

[0053] Preferably, the focused sound field generation system further includes a closed-loop monitoring and safety control module 50 and a human-computer interaction and navigation module 60. The closed-loop monitoring and safety control module 50 is signal-connected to the front-end drive and acquisition module 20 and the imaging and acoustic modeling module 30, and is used to monitor acoustic and physiological surrogate quantities in vivo and realize closed-loop safety control; the human-computer interaction and navigation module 60 is signal-connected to the imaging and acoustic modeling module 30, the focusing control and beam optimization module 40, and the closed-loop monitoring and safety control module (50), and is used to display three-dimensional imaging results, acoustic and functional spectra, and support target selection and parameter setting.

[0054] The closed-loop monitoring and safety control module 50 is used to identify focus point position deviation, energy deposition anomalies and potential cavitation events, and to perform online evaluation and adaptive adjustment of focus parameters based on focus point position deviation, energy deposition anomalies and potential cavitation events, and to trigger automatic power reduction or shutdown protection when safety parameters exceed limits.

[0055] Preferably, the closed-loop monitoring and safety control module 50 includes a passive acoustic monitoring and functional imaging submodule and a safety monitoring and closed-loop control submodule.

[0056] The passive acoustic monitoring and functional imaging submodule is used to acquire echo phase, passive acoustic monitoring signal and functional ultrasound hemodynamic response in real time during transcranial ultrasound neuromodulation. The echo phase drift is used to monitor changes in bone window thermal expansion and tissue sound velocity. The passive acoustic monitoring signal is used to monitor potential cavitation events and abnormal scattering. The functional ultrasound hemodynamic response is used to evaluate the neuromodulation effect, so as to identify focal point position shift, abnormal energy deposition and potential cavitation events.

[0057] Specifically, in order to obtain information on cerebral blood flow and neural function based on structural imaging, this application utilizes the imaging and acoustic modeling module 30 and the closed-loop monitoring and safety control module 50 to realize functional ultrasound (fUS) and ultrasound localization microscopy (ULM) based on three-dimensional bone correction.

[0058] In fUS mode, the system employs an ultrafast three-dimensional imaging strategy, utilizing multi-angle plane wave composite and three-dimensional aperture synthesis to obtain high frame rate volumetric data; through Doppler estimation and singular value decomposition (SVD) or robust filtering to suppress static tissue clutter, it obtains cerebral blood flow perfusion and dynamic change maps, i.e., power Doppler intensity maps.

[0059] To compensate for head movement and long-range drift, volumetric motion correction and cross-time registration steps are introduced during the reconstruction process.

[0060] In ULM or super-resolution mode, the system can intravenously inject ultrasound contrast microbubbles at safe doses and reconstruct sub-diffraction-level cerebral vascular networks using long-term acquisition and single-bubble localization algorithms. After bone-penetrating correction, the effective vascular line density of ULM can be significantly improved, and it can be three-dimensionally co-registered with structural / acoustic maps to obtain integrated indicators of "structure-function-microcirculation".

[0061] The safety monitoring and closed-loop control submodule 50 is used to perform online evaluation and adaptive adjustment of focusing parameters based on focus point position offset, energy deposition anomalies and potential cavitation events, and to trigger automatic power reduction or shutdown protection when safety parameters exceed limits; wherein, the adaptive adjustment includes real-time adjustment of transmit power, pulse repetition frequency, duty cycle and array element phase.

[0062] Specifically, this application constructs a closed-loop sound field focusing framework based on in vivo measurable acoustic and functional proxies, which include echo phase / deformation proxies, passive acoustic monitoring signals, and fUS blood flow response, etc.

[0063] In the actual focusing process, baseline echoes and functional signals are first acquired in low-power pre-detection mode to construct a baseline acoustic field and blood flow reference. Then, focusing is performed according to the focusing scheme output by the focusing control and beam optimization module 40. The safety monitoring and closed-loop control submodule 50 performs anomaly identification and indicator evaluation based on these surrogate quantities. For example, it monitors the slow drift of the echo phase to reflect changes in bone window thermal expansion and tissue sound velocity; monitors cavitation or high-frequency scattering spectra in passive acoustic signals to identify potential unsafe conditions; and monitors fUS blood flow response to assess neuromodulation effects and hemodynamic changes.

[0064] Based on this, the safety monitoring and closed-loop control submodule 50 can employ model predictive control (MPC) or adaptive gain scheduling strategies to adjust the transmit power, pulse repetition frequency, duty cycle, and phase delay in real time according to recently observed trends, thereby maintaining them within safety thresholds while ensuring the effectiveness of neural modulation. When critical safety parameters are detected to be close to or exceed the set thresholds, the system can automatically execute power reduction or shutdown strategies, as shown below:

[0065] in, Indicates the number of functional proxies. Indicates the expected value. This indicates the control increment (power / PRF / duty cycle / phase). and Indicates the weight.

[0066] The human-computer interaction and navigation module 60 is used to overlay and display the three-dimensional acoustic atlas, the three-dimensional structural imaging, and the three-dimensional functional ultrasound imaging in three dimensions. It supports user-preset or interactive selection of transcranial nerve modulation target points and enables alignment and registration with external images. These external images include MRI, CT, fMRI, DSA, etc.

[0067] Preferably, the human-computer interaction and navigation module 60 is integrated into the imaging and acoustic modeling module 30 or the host computer processing terminal, or is set up separately.

[0068] The focused sound field generation method based on three-dimensional transcranial ultrasound imaging described in this application includes: S1: The head-mounted three-dimensional ultrasound transducer module 10 performs multi-band, wide solid angle ultrasound transmission and echo reception on the subject's head. During the ultrasound transmission and echo reception process, the front-end drive and acquisition module 20 performs high-voltage drive, transmission or reception switching and multi-channel signal acquisition on each array element of the head-mounted three-dimensional ultrasound transducer module 10 to obtain multi-channel three-dimensional full-aperture ultrasound signals covering the subject's head. S2: The imaging and acoustic modeling module 30 performs inversion on the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion on the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic map covering the whole brain. Under the constraints of the three-dimensional acoustic map, three-dimensional structural imaging and three-dimensional functional ultrasound imaging are performed. S3: The focusing control module and beam optimization module 40 construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module 10 are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

[0069] It should be noted that the parameters in the above embodiments (such as the number of array elements, frequency band range, interlayer distance, power and dose threshold, etc.) are all preferred examples. Those skilled in the art can adjust them according to different species, skull conditions, target locations and specific clinical / research needs. As long as the head-mounted three-layer coaxial ring array geometry, three-dimensional acoustic map inversion and imaging-control closed-loop strategy are still used, they are all considered to fall within the protection scope of this invention.

Claims

1. A focused sound field generation system based on three-dimensional transcranial ultrasound imaging, characterized in that, include: A head-mounted three-dimensional ultrasound transducer module (10) is used to fit the subject's head for ultrasound transmission and echo reception. The front-end drive and acquisition module (20) is used to drive each element of the head-mounted three-dimensional ultrasonic transducer module (10) under high voltage, switch between transmission and reception, and acquire multi-channel three-dimensional full-aperture ultrasonic signals. The imaging and acoustic modeling module (30) inverts the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion of the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic map covering the whole brain. Under the constraints of the three-dimensional acoustic map, it performs three-dimensional structural imaging and three-dimensional functional ultrasound imaging. The focusing control module and beam optimization module (40) construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module (10) are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

2. The focused sound field generation system as described in claim 1, characterized in that, The focused sound field generation system also includes: The closed-loop monitoring and safety control module (50) identifies focal point position deviation, energy deposition anomaly and potential cavitation events, and performs online evaluation and adaptive adjustment of focusing parameters based on focal point position deviation, energy deposition anomaly and potential cavitation events, and triggers automatic power reduction or shutdown protection when safety parameters exceed limits. The human-computer interaction and navigation module (60) is used to display the three-dimensional acoustic atlas, the three-dimensional structural imaging and the three-dimensional functional ultrasound imaging in three dimensions, support users to preset or interactively select transcranial nerve modulation target points, and realize the alignment and registration with external images.

3. The focused sound field generation system as described in claim 2, characterized in that, The three-dimensional wave equation is expressed as: ； in, Indicates spatial location; , represents angular frequency; Represents frequency domain sound pressure level; Indicates the speed of sound; Indicates density; and Both represent power-law decay parameters; Indicates the sound source; Represents the imaginary unit; Represents the Laplace operator; The three-dimensional acoustic map is represented as follows: ； ； in, Indicates data mismatch items, Represents the regularization term, Represents the measured sound signal. Represents analog sound signals. Indicates the launch index. Indicates the receiving index. Indicates normalized cross-correlation; Indicates the parameters to be inverted; Priors are represented, including image priors and velocity priors; , , , All represent weighting coefficients; Represents the sound field Spatial gradient; The constraints in the inversion and joint inversion processes are expressed as follows: ； in, Indicates the first Wheel parameters, Indicates the first One frequency band, This represents the overall objective function.

4. The focused sound field generation system as described in claim 3, characterized in that, The multi-objective optimization model is expressed as follows: ， ， ； ， ； ； ， ， ， ； in, Represents the complex sound pressure vector at each sampling point; This represents the acoustic propagation matrix from the array element to the sampling point; Represents the complex weights of array elements; This represents the propagation matrix corresponding to the target point; Represents the target propagation vector The conjugate transpose of; This indicates the density at the target point. Indicates the speed of sound at the target point; Indicates the mechanical index; The value representing the ultrasonic operating frequency or center frequency in MHz; Indicates the reduction of radiation amplitude; This represents the spatial peak-time average sound intensity after attenuation. This indicates the removal of attenuation of sound intensity; Represents the bone window energy proxy matrix; Represents the sidelobe sampling matrix. Represents the sampling matrix of the constraint region. This indicates the sidelobe suppression penalty weight; Indicates the energy load penalty weight for bone windows; Indicates the average time window length; Indicates the maximum power limit; This represents the upper limit threshold for the allowed de-attenuation spatial peak time average acoustic intensity. This indicates the upper limit threshold of sound pressure or energy in the constrained area.

5. The focused sound field generation system as described in claim 4, characterized in that, The head-mounted three-dimensional ultrasonic transducer module (10) is an imaging-control integrated three-layer coaxial ring array ultrasonic transducer. The three-layer coaxial ring array ultrasonic transducer includes three layers of coaxial ring array. The three-layer coaxial ring array includes a first ring array layer, a second ring array layer and a third ring array layer distributed axially from top to bottom. The first ring array layer, the second ring array layer and the third ring array layer are all concentric rings with a diameter of 22cm. Each ring array layer contains at least 512 independent array elements.

6. The focused sound field generation system as described in claim 5, characterized in that, The circumferential spacing of the three-layer coaxial ring array is 0.5mm to 1mm, and the axial spacing between each ring array layer is 2mm. The array element material is piezoelectric ceramic material or piezoelectric composite material. The effective operating frequency band of the array element covers 0.1MHz to 1.2MHz, with 0.8MHz as the center frequency.

7. The focused sound field generation system as described in claim 6, characterized in that, The front-end driver and acquisition module (20) includes a multi-channel transmit / receive and phased array control submodule and a data acquisition and host computer processing submodule electrically connected thereto, wherein: The multi-channel transmit / receive and phased array control submodule is used to output multi-band pulse signals and coded transmit signals, and to perform high-voltage driving, transmit or receive switching on each array element through the multi-band pulse signals and the coded transmit signals; wherein, the multi-band pulse signals include narrowband pulse signals and linear chirped signals, and the coded transmit signals include complementary coded signals; The data acquisition and host computer processing submodule is used to acquire multi-channel three-dimensional full-aperture ultrasonic signals and perform gain control, anti-aliasing filtering, and A / D conversion on the multi-channel three-dimensional full-aperture ultrasonic signals; wherein the sampling rate is not less than 10MHz.

8. The focused sound field generation system as described in claim 7, characterized in that, The closed-loop monitoring and safety control module (50) includes: The passive acoustic monitoring and functional imaging submodule is used to acquire the echo phase, passive acoustic monitoring signal and functional ultrasound hemodynamic response of the target area and adjacent areas in real time during transcranial ultrasound neuromodulation. The drift of the echo phase is used to monitor the thermal expansion of the bone window and the change of tissue sound velocity. The passive acoustic monitoring signal is used to monitor potential cavitation events and abnormal scattering. The functional ultrasound hemodynamic response is used to evaluate the neuromodulation effect, so as to identify focal point position shift, abnormal energy deposition and potential cavitation events. The safety monitoring and closed-loop control submodule is used to perform online evaluation and adaptive adjustment of focusing parameters based on focus point position offset, energy deposition anomalies and potential cavitation events, and to trigger automatic power reduction or shutdown protection when safety parameters exceed limits; wherein, the adaptive adjustment includes real-time adjustment of transmit power, pulse repetition frequency, duty cycle and array element phase.

9. The focused sound field generation system as described in claim 8, characterized in that, The human-computer interaction and navigation module (60) is integrated into the imaging and acoustic modeling module (30) or the host computer processing terminal or set up separately.

10. A method for generating a focused sound field based on three-dimensional transcranial ultrasound imaging, wherein the method is implemented using the focused sound field generation system described in any one of claims 1-9, characterized in that... include: The head-mounted three-dimensional ultrasonic transducer module (10) performs multi-band, wide solid angle ultrasonic transmission and echo reception on the subject's head. During the ultrasonic transmission and echo reception process, the front-end drive and acquisition module (20) performs high-voltage drive, transmission or reception switching and multi-channel signal acquisition on each array element of the head-mounted three-dimensional ultrasonic transducer module (10) to obtain multi-channel three-dimensional full-aperture ultrasonic signals covering the subject's head. The imaging and acoustic modeling module (30) inverts the acquired three-dimensional full-aperture ultrasound signal based on the three-dimensional wave equation, and performs joint inversion of the sound velocity, attenuation and density parameters of the skull and brain tissue to obtain a three-dimensional acoustic map covering the whole brain. Under the constraints of the three-dimensional acoustic map, it performs three-dimensional structural imaging and three-dimensional functional ultrasound imaging. The focusing control module and the beam optimization module (40) construct a multi-objective optimization model under the constraint of a three-dimensional acoustic spectrum as a safety condition. Based on the objective optimization model, the phase, amplitude, frequency band and emission timing of each element of the head-mounted three-dimensional ultrasound transducer module (10) are calculated to obtain individualized focusing parameters and acoustic dose parameters for one or more predetermined brain regions. The sound field is focused based on the focusing parameters and the acoustic dose parameters to determine the focal point. The safety conditions include mechanical index MI, thermal index TI, spatial average time and average sound intensity.

Images