A transcranial magnetic acoustic stimulation regulation system and method based on the encoding and consolidation of memory of the medial septal-hippocampal-thalamic-cortical loop

Abstract

The application provides a transcranial magnetic acoustic stimulation regulation system and method based on memory encoding and consolidation of a medial septum-hippocampus-thalamus-cortex loop, which is realized based on a medial septum-hippocampus-thalamus-cortex (SHTC) loop architecture design and comprises an SHTC model construction module: firstly, a local network of the medial septum containing cholinergic, glutamatergic and GABAergic neurons is constructed, then based on the regulation of the medial septum on the hippocampus, a hippocampus-thalamus-cortex coupling network is integrated to form a closed loop model, thereby providing a simulation basis for memory consolidation regulation. The application optimizes the memory consolidation regulation process through double-target point stimulation, thereby providing strong support for precise regulation strategies for memory disorders.

Description

Technical Field

[0001] This invention relates to the field of neuromodulation technology, and in particular to a transcranial magnetic acoustic stimulation modulation system and method based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit. Background Technology

[0002] Memory impairment is a chronic, non-communicable brain disease. According to data released by the World Health Organization, more than 55 million people worldwide currently suffer from dementia, and the number of patients continues to grow at a rate of nearly 10 million new cases per year. Memory impairment is a common feature of neurological diseases such as Alzheimer's disease, vascular dementia, and traumatic brain injury; its essence lies in abnormalities in the neural circuits involved in the processes of memory encoding, consolidation, storage, or retrieval. Current treatments (such as cholinesterase inhibitors) can only alleviate symptoms and cannot reverse neurodegenerative changes.

[0003] Memory encoding and memory consolidation are two core and continuous processes by which the brain transforms short-term experiences into long-term memories: memory encoding relies on the precise regulation of the hippocampal theta rhythm to transform new information into neural signals that can be recognized by the brain; while memory consolidation relies on non-theta states (mainly composed of delta slow waves, hippocampal sharp wave ripples, and thalamic-cortical sleep spindles) to transfer these signals from the hippocampus to the cortex for long-term storage.

[0004] Recent studies have found that techniques targeting neural oscillations (such as transcranial magnetoacoustic stimulation) can partially restore memory function by enhancing cross-frequency synchronicity. However, existing technologies face two major bottlenecks: first, insufficient analysis of the dynamic correlation mechanism between memory encoding and consolidation, with most strategies focusing only on a single stage and lacking systematic regulation of the entire process; second, individual heterogeneity limits their efficacy, with most clinical neuromodulation strategies relying on the "gold standard," making it difficult to establish a clear correlation between stimulation protocol parameters and individual efficacy. The individual sensitivity of magnetoacoustic stimulation parameters and the impact of long-term intervention on neural plasticity remain unclear. Therefore, designing personalized, optimized stimulation protocols that cover both memory encoding and the entire consolidation cycle has become a common challenge for neuroscience and clinical medicine.

[0005] Therefore, it is necessary to study a transcranial magnetic acoustic stimulation modulation system and method based on the memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit to address the shortcomings of existing technologies and solve or alleviate one or more of the above-mentioned problems. Summary of the Invention

[0006] In view of this, the present invention provides a transcranial magnetoacoustic stimulation (TMS) modulation system and method based on the medial septal nucleus-hippocampus-thalamus-cortex circuit for memory encoding and consolidation. By constructing a complete circuit model that includes the memory encoding and memory consolidation processes, integrating a dual-target magnetoacoustic stimulation scheme, and designing a module for quantifying the memory consolidation effect to visualize and evaluate the effect of the stimulation scheme, the present invention solves the problem of poor efficacy of current clinical stimulation schemes for the entire process of memory encoding and consolidation. It can be widely applied to memory enhancement management in healthy individuals and clinical diagnosis and treatment of patients with memory disorders.

[0007] On the one hand, the present invention provides a transcranial magnetic acoustic stimulation modulation system for memory encoding and consolidation based on the medial septal nucleus-hippocampus-thalamus-cortex circuit, including an SHTC model construction module, a transcranial magnetic acoustic stimulation modulation module, and a memory consolidation effect quantification module; The SHTC model construction module is used to construct a heterogeneous medial septum nucleus network containing cholinergic neurons, glutamatergic neurons and GABAergic neurons, and to integrate the heterogeneous medial septum nucleus network into the hippocampus-thalamus-cortex circuit to form the SHTC circuit network model. The SHTC model construction module adjusts the key parameters of the inner diaphragm loop to enable the hippocampus network to switch between theta state and non-theta state. The theta state corresponds to the memory encoding state, and the non-theta state corresponds to the memory consolidation state. The transcranial magnetoacoustic stimulation modulation module includes a TMS unit and a TUS unit. The TMS unit targets the cortex to apply a voltage offset term, and the TUS unit targets the medial septal nucleus to introduce a mechanosensitive current term. The transcranial magnetoacoustic stimulation modulation module uses the membrane potential signal of pyramidal neurons in the CA1 region of the hippocampus as the core modulation basis to achieve dual-target synergistic stimulation. The memory consolidation effect quantification module is used to quantify and evaluate the memory consolidation effect. The quantification indicators include the number and duration of ripples, the number, amplitude and duration of spindles, the number of SOs, theta score, co-occurrence percentage, synchronization index SI value and phase lock value PLV value.

[0008] In addition to the aspects described above and any possible implementation, a further implementation is provided in which the inhibitory-excitation synaptic strength of the CA1 and CA3 regions in the SHTC model construction module is dynamically adjusted according to the average firing rate of GABAergic neurons in the medial septal nucleus; and the regulatory parameters of the medial septal nucleus on the hippocampus and cortex are dynamically adjusted according to the ratio of the theta band power to the delta band power of the membrane potential in the CA1 region.

[0009] As described above and in any possible implementation, a further implementation is provided in which the voltage offset term S(t) applied by the TMS unit satisfies S(t) = A stim·sin(2πf stim ·t), where A stim To stimulate amplitude, f stim The stimulation frequency; the A stim The value range is 0 to 100mV, and the f stim The coverage range is from 1 to 100 Hz.

[0010] As described above and in any possible implementation, a further implementation is provided in which the mechanically sensitive current term I introduced by the TUS unit is... TUS (t) satisfies I TUS (t)=A·G·sin(2πf mod ·t)·sin(2πf US ·t)·P(t), where A is the stimulus amplitude, G is the focusing gain coefficient, and f mod f is the modulation frequency. US Where A is the ultrasonic carrier frequency, and P(t) is the pulse window function; the value of A ranges from 0.1 to 1.0 mV, and f... mod A theta rhythm interval of 4 to 12 Hz is selected, and the value of G ranges from 1.2 to 2.0.

[0011] In addition to the aspects described above and any possible implementation, an implementation is further provided in which the pulse window function P(t) is a binary function, the pulse width is 100 to 300 μs, the pulse repetition period is 5 to 15 ms, and the duty cycle is less than 50%.

[0012] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the memory consolidation effect quantification module detects spindle by performing a 12-16Hz bandpass filter on the membrane potential signal of pyramidal neurons in the CA1 region, extracting the instantaneous amplitude through Hilbert transform, smoothing it through a 300ms Gaussian window, and determining the amplitude based on an amplitude threshold; detects ripple by performing a 100-250Hz bandpass filter on the membrane potential signal, calculating the root mean square, smoothing it through a 50ms Gaussian window, and determining the ripple based on a power threshold; and detects SO by performing a 0.5-2Hz bandpass filter on the membrane potential signal and determining the SO based on the time interval between positive and negative zero crossings.

[0013] In addition to the aspects described above and any possible implementation, a further implementation is provided in which the theta score is based on the membrane potential signal of the pyramidal neurons in the CA1 region, and the theta band component and the delta band component are extracted respectively through a third-order zero-phase-delay Butterworth bandpass filter, and the ratio of their average power is calculated.

[0014] In addition to the aspects and any possible implementations described above, an implementation is further provided in which the co-occurrence percentage includes SO-ripple co-occurrence percentage, SO-spindle co-occurrence percentage, and spindle-ripple co-occurrence percentage; The condition for determining SO-ripple co-occurrence is that there is an SO peak within ±0.5s before and after the peak of the ripple; The condition for determining SO-spindle co-occurrence is that there is an SO peak within ±1.5s before and after the spindle peak moment; The spindle-ripple co-occurrence is determined by a two-way decision logic, that is, if a ripple peak exists within a specified time window before or after the spindle peak, or if a spindle peak exists within a specified time window before or after the ripple peak, co-occurrence is determined if either condition is met.

[0015] In addition to the aspects described above and any possible implementation, a further implementation is provided in which the synchronization index SI value is obtained by extracting the instantaneous phase of the low-frequency oscillation and the instantaneous phase of the high-frequency oscillation power fluctuation through Hilbert transform, and calculating the modulus of the complex exponential mean of the phase difference between the two.

[0016] In accordance with the aspects and any possible implementations described above, a transcranial magnetic acoustic stimulation (TMS) modulation method based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit is further provided. The method, using the aforementioned TMS modulation system, includes the following steps: Step 1: Construct the SHTC loop network model. By adjusting the key parameters of the inner diaphragm loop, the hippocampal network is transformed between theta state and non-theta state, which correspond to memory encoding and memory consolidation, respectively. Step 2: Based on the SHTC loop network model, taking the membrane potential signal of the pyramidal neurons in the CA1 region as the core regulation basis, the TMS unit targets the cortex to apply a voltage offset term and the TUS unit targets the medial septal nucleus to introduce a mechanosensitive current term, thereby implementing dual-target synergistic stimulation. Step 3: Quantitatively analyze the output signal of the SHTC loop network model after dual-target synergistic stimulation, and evaluate the memory consolidation effect based on the number and duration of ripples, the number and amplitude and duration of spindles, the number of SOs, theta score, co-occurrence percentage, synchronization index SI value and phase lock value PLV value.

[0017] Compared with the prior art, the present invention can achieve the following technical effects: 1. By constructing a heterogeneous medial septal nucleus network containing cholinergic, glutamatergic, and GABAergic neurons, and integrating it into the hippocampus-thalamus-cortex circuit to form a complete SHTC closed-loop network model, the dynamic transformation between the hippocampal theta and non-theta states was simulated, covering the complete physiological process of memory encoding and memory consolidation. This design solves the technical problem of existing technologies that only focus on a single stage and lack regulation of the entire process system, providing a simulation basis for optimizing dual-target stimulation schemes; 2. A dual-target synergistic stimulation strategy, employing the TMS unit to apply a voltage offset to the cortex and the TUS unit to introduce a mechanosensitive current term targeting the medial septal nucleus, was developed. Based on the membrane potential signal of pyramidal neurons in the CA1 region of the hippocampus, precise regulation of the cortex and medial septal nucleus was achieved. TMS alters the membrane potential of cortical neurons by simulating an induced electric field through voltage offset, while TUS activates ion channels by simulating focused ultrasound through mechanosensitive current. The synergistic effect of both achieves systematic regulation of the entire process of memory encoding and consolidation, overcoming the limitations of existing single-target stimulation techniques. 3. By embedding a multi-dimensional quantitative system covering the quality of theta-state encoding and the efficiency of non-theta-state consolidation, including seven categories of indicators such as the number / duration / amplitude of ripples, spindles, SOs, theta score, co-occurrence percentage, SI value, and PLV value, the dynamic synergistic efficiency of the SHTC network is comprehensively quantified from the perspectives of state spectrum transition, phase coupling, co-occurrence probability, and single-rhythm characteristics, establishing a clear correlation between stimulation parameters and efficacy. This design solves the technical problem of unclear correlation between stimulation parameters and individual efficacy in existing technologies, providing a visual evaluation tool for the quantitative assessment and precise control of memory consolidation efficacy. 4. By introducing the dynamic adjustment function of the inhibition-excitation synapse intensity in the CA1 and CA3 regions and the regulation parameter function of the medial septal nucleus on the hippocampus / cortex, the synapse intensity and connectivity coefficient are dynamically adjusted according to the average firing rate of GABAergic neurons in the medial septal nucleus and the theta / delta power ratio of the CA1 membrane potential. This accurately simulates the physiological mechanism of GABAergic neurons driving hippocampal theta oscillations, realizes the dynamic regulation of the medial septal nucleus on the transformation between the theta and non-theta states in the hippocampus, and improves the simulation accuracy of the model for long-term neural activities.

[0018] Of course, any product implementing this invention does not necessarily need to achieve all of the technical effects described above at the same time. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the transcranial magnetic acoustic stimulation modulation strategy and system based on the memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit provided in the embodiments of the present invention. Figure 2 This is the main flowchart of the TMS unit control module provided in the embodiment of the present invention; Figure 3 This is the main flowchart of the TUS unit control module provided in the embodiment of the present invention. Detailed Implementation

[0021] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0022] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0023] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0024] Embodiments of the present invention provide a transcranial magnetic acoustic stimulation modulation strategy and system implementation based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit, such as... Figure 1 As shown, the system includes: The SHTC model building module is used to build SHTC loop network models to simulate the generation and coupling of various rhythms between different brain regions, including spindle, ripple, and SO. The transcranial magnetoacoustic stimulation modulation module is configured as follows: the TMS unit targets the cortex, and the TUS unit targets the medial septal nucleus, respectively simulating the neuromodulation effects of magnetic field and focused ultrasound; The memory consolidation effect quantification module embeds a multimodal quantification system to measure the memory consolidation enhancement effect under different parameters.

[0025] The system described in this invention can be used to simulate the rhythms of multiple brain regions, including the hippocampus, thalamus, cortex, and medial septal nucleus, in patients with memory disorders. It also establishes a magnetic and acoustic neuromodulation experimental platform, enabling the exploration of memory consolidation mechanisms based on the SHTC circuit and the design of precise neuromodulation schemes.

[0026] In the SHTC model building module, based on the anatomical structures of the medial septal nucleus, hippocampus, thalamus, and cortex, and the characteristics of excitatory / inhibitory propagation, an SHTC circuit network model is established. This network comprises four main subnetworks: the medial septal nucleus, hippocampus, cortex, and thalamus. The medial septal nucleus network contains a group of cholinergic neurons, a group of glutamatergic neurons, and a group of GABAergic neurons. The hippocampal network consists of CA1 and CA3 regions, each containing a group of excitatory neurons and a group of inhibitory neurons, which connect to form local circuits. The cortical network is divided into two regions, CX1 and CX2, each containing a group of excitatory neurons and a group of inhibitory neurons. The thalamic network (TH) consists of a group of TC neurons and a group of RE neurons in the reticular nucleus. The cortical network interacts with the hippocampus and thalamus through long-range connections, and the medial septal nucleus network interacts with the hippocampus through long-range connections.

[0027] Sleep stages can be divided into rapid eye movement (REM) sleep, non-rapid eye movement (NREM) sleep, and deep sleep (slow-wave sleep). Each stage exhibits two main types of hippocampal firing activity: theta state and non-theta state. Specifically, the REM stage is characterized by a theta state with dominant power at 4-12 Hz, driven by GABAergic neurons in the medial septal nucleus. This state induces synchronized theta oscillations through rhythmic inhibition of the hippocampal network and is used for memory encoding related learning and exploratory behaviors. The NREM stage is dominated by a non-theta state with dominant delta rhythm (0.5-4 Hz). When the theta score is <0.34, it is identified as deep sleep (slow-wave sleep) within NREM. This stage is characterized by hippocampal wavy patterns and forms a cross-brain region synergy with spindle waves generated by the thalamus and slow oscillations originating from the cortex. The precise coupling of these three factors can promote the conversion of newly acquired information in the hippocampus into long-term cortical storage, which is the core physiological process of offline memory processing. In addition, slow waves and slow oscillations are typical features of the deep sleep stage and belong to the global synchronous activity of the brain, which further provides a stable neural environment for memory consolidation.

[0028] The SHTC model can simulate the membrane potentials of various brain regions through parameters, thereby generating key intracranial rhythms such as spindles, ripples, and SOs. The specific method is as follows: Spindle: After obtaining the local field potential (LFP) signal from the membrane potential of excitatory neurons in the cortical CX2 network, it is first filtered by a bandpass filter (12~16 Hz). The filter order corresponds to three cycles of the low-frequency cutoff frequency. Then, the instantaneous amplitude of the signal is calculated by Hilbert transform and smoothed by a 300 ms Gaussian window. Candidate spindles must meet three conditions to be determined as valid events: the smoothed amplitude exceeds the set threshold (1 standard deviation for model data), the duration is in the range of 0.4~3.5 s, and the interval between two adjacent candidate events is < 0.5 s, which should be merged into a single spindle event.

[0029] Ripple: The hippocampal LFP signal is obtained from the CA1 region (the membrane potential of pyramidal neurons is directly used in the model data). First, it is filtered by a 100~250 Hz fourth-order zero-phase-delay Butterworth bandpass filter. Then, the root mean square (RMS) of the filtered signal is calculated and smoothed by a 50 ms Gaussian window. When the smoothed RMS signal exceeds the set threshold (median + 2 standard deviation in the model data) and the duration is ≥30 ms, it is determined to be a valid ripple event.

[0030] SO: After obtaining the LFP signal from the membrane potential of excitatory neurons in the cortical CX2 network, the signal is filtered by an FIR bandpass filter at a frequency of 0.5~2 Hz. The filter order corresponds to three cycles of the low-frequency cutoff frequency. The LFP signal is determined to be a valid SO event when it meets three conditions simultaneously: First, the time interval between consecutive positive and negative zero crossings is in the range of 0.5~2 s (corresponding to a frequency of 0.5~2 Hz); Second, the negative peak value exceeds the set threshold (1.4 times the average peak value of the model data); Third, the difference between the positive and negative peak values ​​exceeds the set threshold (1.4 times the overall average amplitude of the model data signal).

[0031] In this embodiment of the invention, the simulation results of SHTC on several rhythms can be compared with existing experimental data to ensure the authenticity and validity of the system simulation results, and to analyze the potential pathological mechanisms inside the loop that are difficult to obtain with general instruments.

[0032] In this study, the transcranial magnetoacoustic stimulation (TMS) modulation module employs two precise targeting neuromodulation strategies: TMS targeting the cerebral cortex and TUS targeting the medial septal nucleus, respectively, to achieve precise reinforcement of memory consolidation.

[0033] Furthermore, as the first specific implementation of the stimulation scheme of the present invention, the TMS unit has the cortex as the stimulation target point. The cortical targeted voltage offset superposition method is used to accurately simulate the neuromodulation effect of TMS, effectively solving the technical defects of traditional whole-brain stimulation models such as target ambiguity and cross-brain region interference.

[0034] The core design logic of this module is as follows: a voltage offset term is applied to the state variables representing the excitatory and inhibitory neuronal activities in the cortex on the right side of the differential equation of the SHTC loop network model. The time derivative term on the left side of the equation remains unchanged, ensuring that the simulated effect is strictly confined to the cortex and does not affect the normal dynamic characteristics of other brain regions such as the medial septal nucleus, hippocampus, and thalamus. From a physical equivalence perspective, the voltage offset term... The superposition is mathematically equivalent to introducing an additional voltage field into the synaptic input of cortical neurons, which is highly consistent with the physical process of TMS changing the membrane potential of cortical neurons through induced electric fields, ensuring the scientific accuracy of the simulation.

[0035] As a core improvement of this invention, the voltage offset term An alternating function is used to accurately simulate the rhythmic driving effect of TMS, and its expression is as follows:

[0036] in, To stimulate amplitude (the range can be flexibly adjusted from 0-100 mV to adapt to different intensities of TMS effect according to simulation requirements). The stimulation frequency (can be set as needed to characteristic frequencies such as 1Hz, 5Hz, 10Hz, 20Hz, 50Hz, 100Hz, etc.) The stimulus time variable.

[0037] To analyze the amplitude With frequency To independently influence the efficacy of neuromodulation and explore the optimal stimulation parameter window, this invention can conduct systematic simulations by expanding the multi-dimensional parameter space: the frequency dimension covers the typical TMS stimulation frequency band of 1-100 Hz, and the amplitude dimension sets a gradient range of 0-100 mV. By combining different parameter combinations, the dynamic changes of core indicators related to memory consolidation are continuously monitored and quantified. This design can uncover potentially better amplitude-frequency combination schemes, providing direct theoretical guidance for parameter design in subsequent in vivo TMS experiments. The model of this invention can drive the hippocampus to achieve dynamic transformation between theta and non-theta states, supporting the simulation of memory encoding and consolidation processes. Transcranial magnetic acoustic stimulation modulation module: divided into transcranial acoustic stimulation (TUS) modulation unit and transcranial magnetic stimulation (TMS) modulation unit. The former targets the medial septal nucleus, using focused ultrasound parameters to optimize and match the impedance characteristics of deep brain tissue to form a mechanical vibration effect; the latter targets the cortex, using parameters to optimize and match the cortical tissue resistance characteristics to form an effective stimulation electric field. Both can independently modulate the neural activity of corresponding brain regions and related circuits. The memory consolidation effect quantification module embeds a quantification system that covers the quality of theta-state encoding and the efficiency of non-theta-state consolidation, measuring the enhancement effect of memory consolidation under different parameters.

[0038] Furthermore, as a second specific implementation of the stimulation scheme of the present invention, the TUS unit uses the inner septum as the precise target point and innovatively adopts the mechanically sensitive current equivalent modeling method.

[0039] The core design logic of this module is as follows: In the SHTC loop network model, for the core state variables representing the activity of excitatory neurons (cholinergic and glutamatergic) and inhibitory neurons (GABAergic) in the medial septal nucleus, a mechanosensitive current term is introduced on the right side of its current balance differential equation. The time derivative term on the left side of the equation remains unchanged, enabling specific regulation of deep targets.

[0040] From a physical equivalence perspective, the mechanically sensitive current term The introduction of this technology is mathematically equivalent to simulating the mechanical vibration effect of focused ultrasound in transcranial acoustic stimulation (TUS) – by activating mechanosensitive ion channels such as Piezo1 / 2 on the neuronal membrane of the medial septal nucleus, it alters ion permeability and generates a controllable ion flow, which aligns with the core mechanism of TUS in regulating neuronal firing through mechanical force. At the same time, this design fully leverages the precise focusing advantage of focused ultrasound on deep brain regions, providing a reliable paradigm for modeling research on the neural modulation of deep nuclei.

[0041] As the core improvement of this invention, the mechanically sensitive current term adopts a composite form of "amplitude-modulated sine wave + focusing gain + pulse window" to reproduce the pulse characteristics, rhythmic driving effect, and deep focusing advantages of focused ultrasound. Its expression is as follows:

[0042] The definitions and design basis of each parameter are as follows: To stimulate amplitude, the value range is 0.1-1.0 mV, which can be flexibly adapted to different sound pressure intensities of the TUS effect according to simulation needs, while avoiding over-excitation of medial septal nucleus neurons; The focusing gain coefficient is used to simulate the energy focusing effect when focused ultrasound targets deep brain regions (medial septal nucleus). Real TUS focuses ultrasound energy onto the target point through a focusing lens, which will produce a certain energy gain. This coefficient can accurately match this physical characteristic and improve the realism of the simulation. The modulation frequency; It uses the ultrasound carrier frequency to balance the penetration depth and focusing accuracy in deep brain regions; This is a pulse window function (binary function), primarily used to simulate the pulsed application characteristics of real TUS, avoiding tissue thermal damage caused by continuous waves. Its expression is:

[0043] in The pulse width. It is the pulse repetition period, and the duty cycle must be satisfied. Safety constraints.

[0044] Analyze the amplitude of the system Modulation frequency Ultrasonic carrier frequency To investigate the independent regulatory effects and synergistic mechanisms of TUS, this invention employs a multi-dimensional parameter space simulation scheme: the modulation frequency is selected within the key theta rhythm range of 4-12 Hz, the amplitude is set within a gradient range of 0.1-1.0 mV, and the focusing gain coefficient is set within the range of 1.2-2.0. By continuously monitoring key indicators such as the synchronicity of medial septal nucleus discharge, hippocampal theta rhythm intensity (e.g., theta score), and cross-brain region circuit coupling efficiency (e.g., PLV value, percentage of co-occurrence), the optimal parameter combination suitable for different regulatory needs is identified, providing direct theoretical support for the parameter design of subsequent in vivo in vivo medial septal nucleus-targeted TUS experiments.

[0045] The quantitative module for memory consolidation effectiveness selects indicators based on the physiological processes of memory processing. Memory processing includes two core stages: the information encoding state dominated by theta oscillations and the offline slow-wave sleep consolidation state dominated by non-theta oscillations. The cross-frequency coupling, spatiotemporal co-occurrence, and feature evolution of multiple neural rhythms in these two states jointly determine memory consolidation efficiency. Therefore, constructing a quantitative module covering the entire memory cycle and formulating hierarchical evaluation indicators are essential. Traditional studies are mostly limited to static spectral analysis of LFP in a single brain region, failing to quantitatively characterize phase synchronization and spatiotemporal co-occurrence between rhythms, and lacking standardized calculations of the rhythm transition laws between encoding and consolidation states. This makes it difficult to systematically analyze the hierarchical dynamic collaborative mechanism of the SHTC network throughout the entire memory cycle.

[0046] Furthermore, this invention constructs a multimodal quantization system comprising basic rhythmic feature parameters and advanced coupled quantization indices. The specific settings and physiological significance of each index are as follows: The basic rhythmic characteristic parameters include ripple, spindle, and SO-related characteristic parameters. Focusing on the duration of slow-wave sleep ripples, this study quantifies their dynamic changes to assess the potential impact of neuromodulation interventions on offline memory consolidation. Slow-wave sleep is characterized by three dimensions: the number, amplitude, and duration of spindle waves. The number of spindle waves reflects the rhythmic activation frequency of the thalamic reticular nucleus, and its cumulative effect can periodically enhance the plasticity of cortical neurons. Duration characterizes the oscillation stability of a single spindle wave event; long-term oscillations help maintain the continuity of phase synchronization across brain regions. Average amplitude indicates the synchronization strength of local neuronal clusters; high-amplitude events can effectively drive information encoding in downstream brain regions. Slow oscillations, with the number of events as the core quantitative parameter, have two physiological significances: first, characterizing the global regulatory strength of cortical slow-wave activity on the thalamus-hippocampus network; second, assessing the cumulative effect of the neural plasticity time window during sleep, providing a key observational dimension for analyzing the regulatory mechanism of exogenous interventions on offline memory replay.

[0047] Building upon the basic rhythm feature detection described above, this invention further utilizes theta score, percentage of rhythm co-occurrence, SI value, and PLV value to achieve quantitative assessment of rhythmic state and cross-regional coupling. The specific logic for the advanced brain rhythm calculation is as follows: Theta score: Theta score is a core spectral quantitative indicator for dual-state classification of memory processing. It is calculated based on the raw membrane potential signal of pyramidal neurons in the CA1 region. The core method involves extracting the 4-12Hz theta band and the 0.5-4Hz delta band components using a third-order zero-phase-delay Butterworth bandpass filter, and then calculating the average power ratio of these two components. Its core value lies in directly reflecting the relative strength of theta oscillations to delta rhythms in both states. It can quantitatively characterize the dynamic changes in the rhythmic spectrum during the encoding-consolidation transition, and provides a standardized basis for determining whether theta oscillations are at a relatively high expression level. Furthermore, it provides a global observation dimension for analyzing the regulatory effects of exogenous interventions on the temporal transitions throughout the entire memory cycle.

[0048] The calculation of the co-occurrence percentage requires first extracting rhythmic events and peak values ​​using a specialized algorithm. This involves detecting valid events for SOs, spindles, and ripples, and accurately recording the peak times of each type of event. Following this, the core time window overlap determination step is performed, defining three core co-occurrence relationships based on standardized windows: SO-ripple co-occurrence is determined by the presence of an SO peak within ±0.5s before and after the peak of the ripple. The co-occurrence of SO-spindle waves is determined by the presence of an SO peak within ±1.5s before and after the peak value of the spindle wave. The co-occurrence of spindle waves and ripples adopts a two-way judgment logic - it is necessary to determine whether there is a spindle wave peak within ±0.5s before and after the ripple peak, and also to determine whether there is a ripple peak within ±1.0s before and after the spindle wave peak. Finally, the number of events that meet the time window overlap condition is counted, divided by the total number of rhythmic events used as the baseline in this type of relationship, and multiplied by 100% to obtain the percentage of common occurrence.

[0049] As a core indicator for quantifying the spatiotemporal synergy of the three, its core significance lies in reflecting the coupling strength across brain regions and the characteristics of associated rhythms, while verifying the regulatory role of long-range connectivity and providing data support for the memory consolidation hypothesis.

[0050] The core of SI value calculation is to quantify the synchronization lock relationship between the power fluctuations of low-frequency oscillations and high-frequency oscillations. First, signal preprocessing is performed. To ensure the accuracy of phase estimation, both the low-frequency oscillation signal and the power signal of the high-frequency oscillation need to be filtered to a specific frequency range of the low-frequency modulation oscillation, for example, SOs is 0.5-2 Hz, spindles is 12-16 Hz, and ripples is 100-250 Hz. A second-order FIR bandpass filter is used, with the order corresponding to three cycles of the low-frequency cutoff frequency. Subsequently, the instantaneous phase of the two time series is extracted using Hilbert transform. The phase value represents the low-frequency oscillation band. The SI value represents the phase value of the high-frequency oscillation power fluctuation. Based on this, the SI value is calculated using the following formula:

[0051] In the formula, n is the number of time points within the selected time interval. The calculation requires limiting the time window for phase estimation—±1 s before and after the SO peak and ±0.25 s before and after the spindle wave peak. The final SI is in complex form, where the angle reflects the phase difference between the two oscillating events, and the absolute value represents the coupling strength between them, thereby achieving a quantitative characterization of the degree of oscillation coordination in different brain regions.

[0052] The core of PLV value calculation is to quantify the lock-in stability of high-frequency rhythmic events on low-frequency phase signals. The specific steps are as follows: First, using the peak time of each event of a high-frequency rhythm (such as ripples and spindles) as an anchor point, extract the instantaneous phase of the low-frequency phase modulation signal (such as SOs and spindles) at that moment after purification and filtering. Define this phase as the lock-in phase of the high-frequency event. ( For the event sequence number, =1,2,...,N, where N is the total number of valid high-frequency events); then all locked phases are converted into complex exponential form. The PLV is calculated by summing all complex exponents and taking the average value. The modulus of this average value is then calculated, which is the PLV. The formula for PLV is as follows:

[0053] The value of this indicator ranges from 0 to 1. The closer it is to 1, the more concentrated all high-frequency events are locked in the same position in the low-frequency phase, and the stronger the phase locking. The closer it is to 0, the more random the high-frequency events are in the low-frequency phase, and the no significant locking relationship.

[0054] The temporal coupling, spatiotemporal co-occurrence, and dual-state transition of cross-band oscillations are the core mechanisms of sleep memory consolidation. The above-mentioned multi-dimensional indicator system can comprehensively quantify the dynamic collaborative efficiency of the SHTC network from the perspectives of state spectrum transition, phase coupling, co-occurrence probability, and single rhythm characteristics, and fully characterize the organizational principles and regulatory laws of rhythmic collaboration throughout the memory cycle.

[0055] Furthermore, as a specific implementation of the present invention, the technical details of the present invention are as follows: 1. SHTC Network (1) Network structure The SHTC network structure comprises four core subnetworks: the medial septal nucleus, hippocampus, cortex, and thalamus. Each subnetwork has distinct structural and connectivity characteristics: the medial septal nucleus network contains one set each of cholinergic, glutamatergic, and GABAergic neurons, which synergistically regulate circuit rhythms; the hippocampal network consists of two functional regions, CA1 and CA3, each containing one set of excitatory neurons and one set of inhibitory neurons, forming a local closed loop; the cortical network is divided into two regions, CX1 and CX2, each equipped with one set of excitatory neurons and one set of inhibitory neurons, enabling cross-brain region interaction with the hippocampus and thalamus through long-range projections; and the thalamic network consists of one set each of thalamic cortical TC neurons and reticular nucleus RE neurons, participating in circuit information integration and regulation.

[0056] (2) Network Model The model can be described by the following 21 differential equations: Medial septal network:

[0057]

[0058]

[0059]

[0060]

[0061] Haima Networks:

[0062] Cortical network: Thalamic network:

[0063] In the above differential equation, the neuron membrane potential is: k represents one of the brain regions or a class of neurons, such as CA1, CA3, CX, 1CX, TH, Ch, Glu, or GABA. m represents neurons in a brain region, such as TC neurons and RE neurons in the thalamus, or excitatory and inhibitory neurons in the cortex, medial septal nucleus, and hippocampus. The firing rate of neurons in the medial septal nucleus, hippocampus, and cerebral cortex is also considered. It is about membrane potential A sigmoid function (represented as the G3 function in the formula).

[0064] in, It is the maximum (minimum) firing rate of neuron m in a completely depolarized state. It is the firing threshold potential of neuron m. It is the dependence of discharge rate on membrane potential.

[0065] neurons To neurons The network connection probability and strength are respectively determined by and Given, among which and It is one of CA1, CA3, CX, 1CX, TH, Ch, Glu, or GABA, where m and n are TC neurons and RE neurons of the thalamus, or excitatory and inhibitory neurons of the cortex, medial septal nucleus, and hippocampus. Similarly, This represents the connection strength from neuron m to neuron n within the group, where It is one of CA1, CA3, CX, 1CX, or TH. Specifically, it refers to all excitability-to-excitability connections. It is a decreasing sigmoid function of the adaptation variable c, where c represents the average fitness level of excitatory neurons (represented as the G4 function in the formula):

[0066]

[0067]

[0068] in, It is the maximum synaptic strength corresponding to the fitness level at the equilibrium value (set to zero). It is the threshold adaptation level. This indicates the steepness of the dependence of excitatory-to-excitatory connections on adaptation variables. Each synaptic activity elicits an adaptation level... Increase .

[0069] Functions of the thalamus module It is an sigmoid function (represented as the G1 function in the formula), which can be expressed as:

[0070] This function can quantify the magnitude of the T-type calcium current, which depends on the burst variable. ,in, It is the amplitude of the maximum burst current. It is the burst threshold. It is the burst sharpness function. It is a specific film capacitance. Among them, It is a burst variable specifically defined for thalamic neurons to simulate T-type calcium currents:

[0071] in, The burst time constant is much larger than the membrane time constant. It is the burst baseline, which depends on the dynamic adjustment of the membrane potential. Or when hour, ,otherwise The firing rate of thalamic neurons is defined as (expressed as a G2 function in the formula):

[0072] in, These are the maximum discharge rate, the threshold discharge potential, and the sharpness dependent on the discharge rate-membrane potential, respectively. It is the mode switching coefficient, which controls... The adjustment strength of the weights for the two modes.

[0073] The dynamic regulation function of the inhibitory-excitation synaptic strength in CA1 and CA3 regions is defined as follows:

[0074]

[0075] in, This represents the average firing rate of GABAergic neurons in the medial septal nucleus. ( )and ( These are the minimum and maximum thresholds for synaptic strength, respectively, which limit the dynamic adjustment range of synaptic strength; parameters This is an adjustment coefficient used to control the rate at which synaptic strength changes with the discharge rate. The characteristic discharge rate, serving as a benchmark reference value for the influence of discharge rate on synaptic strength, together determine the dynamic variation law of synaptic strength.

[0076] The regulatory function of the medial septum on the hippocampus and cortex is:

[0077] in, For adjustment coefficients, It is the theta / delta power ratio of the CA1 membrane potential.

[0078] 2. The simplified model of external stimulus regulation is as follows:

[0079] here, It is the membrane potential. It is a specific film capacitor. and Neuron The adaptive variables and their increments. and It is the number of neurons, in addition, , , , ; ; ; , ; , ; , .in Apply voltage offset terms to state variables Specifically, it is added to the cortical network. For mechanically sensitive current terms, they are specifically added to the inner partition network.

[0080] The specific parameter values ​​are as follows:

[0081]

[0082]

[0083]

[0084]

[0085] Example 1 This embodiment provides an overall implementation scheme for a transcranial magnetic acoustic stimulation modulation system based on the memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit.

[0086] like Figure 1 As shown, the system includes an SHTC model construction module, a transcranial magnetoacoustic stimulation modulation module, and a memory consolidation effect quantification module.

[0087] The SHTC model building module is used to construct a complete closed-loop network model of the medial septum-hippocampus-thalamus-cortex.

[0088] The module first constructs a medial septal nucleus network containing three types of heterogeneous neurons, including 1000 cholinergic neurons, 1000 glutamatergic neurons, and 50 GABAergic neurons.

[0089] The hippocampal network consists of CA1 and CA3 regions. CA1 contains 1000 excitatory neurons and 100 inhibitory neurons, while CA3 contains 1000 excitatory neurons and 100 inhibitory neurons.

[0090] The cortical network is divided into CX1 and CX2 regions. CX1 contains 8,000 excitatory neurons and 2,000 inhibitory neurons, while CX2 contains 8,000 excitatory neurons and 2,000 inhibitory neurons.

[0091] The thalamic network contains 500 TC neurons and 500 RE neurons.

[0092] Based on the anatomical structure of the medial septal nucleus, hippocampus, thalamus, and cortex, a SHTC circuit network model was established, taking into account excitatory and inhibitory connections.

[0093] Short-range connections have high synaptic strength, forming local small-world networks. For example, the excitatory-to-excitatory neuron connection strength in the CA3 region is 0.7 mV, and the connection probability is 0.25.

[0094] Long-range connections have lower synaptic strength but wider coverage, supporting rhythmic coupling across brain regions. For example, the connection strength from CA1 excitatory neurons to cortical excitatory neurons is 2 mV, with a connection probability of 0.002.

[0095] Dendritic spine frequency adaptation is introduced to describe synaptic dynamics, and the strength of excitatory connections is adjusted through adaptation variables.

[0096] The inhibitory-excitatory synaptic strengths in the CA1 and CA3 regions are based on the average firing rate f of GABAergic neurons in the medial septal nucleus. avg Dynamic adjustment.

[0097] The intensity of the inhibitory-excitation synapse in region CA1 is expressed as an exponentially decreasing function, while the intensity of the inhibitory-excitation synapse in region CA3 is expressed as an exponentially increasing function. The adjustment coefficient k = 0.8, and the characteristic discharge rate f* = 1 Hz.

[0098] The medial septum regulates the parameter k of the CA1 area of ​​the hippocampus, the CX area of ​​the cortex, and the 1CX area of ​​the cortex. MS →*Based on the power ratio P of the theta band (4-12Hz) and delta band (0.5-4Hz) of the membrane potential in the CA1 region. θ / P δ Dynamic adjustment, adjustment coefficient k coeff =0.6, with a value range of 0.1-1.0.

[0099] By adjusting key parameters of the inner septal nucleus loop, this model can drive the hippocampal network to transition between theta and non-theta states.

[0100] The theta state corresponds to the REM phase, with theta power dominating at 4-12 Hz. It is driven by GABAergic neurons in the medial septal nucleus and induces synchronized theta oscillations through rhythmic inhibition of the hippocampal network. It is used for memory encoding-related learning and exploratory behaviors.

[0101] The non-theta state corresponds to the NREM stage, dominated by a delta rhythm power of 0.5-4 Hz. When the theta score is <0.34, it is identified as deep sleep (slow-wave sleep). This stage is characterized by hippocampal ripples, which, together with spindles generated by the thalamus and slow oscillations (SO) originating from the cortex, form a cross-brain region synergy to promote the conversion of newly acquired information in the hippocampus into long-term cortical storage.

[0102] The transcranial magnetoacoustic stimulation modulation module includes a TMS unit and a TUS unit, which uses the membrane potential signal of pyramidal neurons in the CA1 region of the hippocampus as the core modulation basis to achieve dual-target targeted modulation.

[0103] The TMS unit targets the cortex, applying a voltage offset term S(t) = A to the state variables of excitatory and inhibitory neurons in the cortex on the right-hand side of the differential equation of the SHTC loop network model. stim ·sin(2πf stim ·t).

[0104] Stimulus A stim The value ranges from 0 to 100 mV, and the stimulation frequency f stim It covers the frequency band from 1 to 100 Hz and can be set to characteristic frequencies such as 1 Hz, 5 Hz, 10 Hz, 20 Hz, 50 Hz, and 100 Hz as needed.

[0105] The superposition of the voltage offset term S(t) is mathematically equivalent to introducing an additional voltage field into the synaptic input of cortical neurons. This is highly consistent with the physical process of TMS changing the membrane potential of cortical neurons through induced electric fields, ensuring that the simulation effect is strictly limited to the cortex and does not affect the normal dynamic characteristics of other brain regions such as the medial septal nucleus, hippocampus, and thalamus.

[0106] The TUS unit targets the medial septal nucleus, introducing a mechanosensitive current term I on the right-hand side of the differential equation for the current balance between excitatory neurons (cholinergic and glutamatergic) and inhibitory neurons (GABAergic) in the medial septal nucleus. TUS (t) = A·G·sin(2πf) mod ·t)·sin(2πf US ·t)·P(t).

[0107] The stimulation amplitude A ranges from 0.1 to 1.0 mV, and the modulation frequency f modThe theta rhythm range of 4 to 12 Hz was selected, the focusing gain coefficient G ranged from 1.2 to 2.0, and the ultrasonic carrier frequency f was... US The value ranges from 0 to 1.5MHz.

[0108] The pulse window function P(t) is a binary function, the pulse width PW = 200 μs, and the pulse repetition period T. PRF =10ms, duty cycle D=(PW / T) PRF ) × 100% = 2% < 50%, which satisfies the safety constraint.

[0109] Mechanically sensitive current term I TUS The introduction of (t) is mathematically equivalent to simulating the mechanical vibration effect of focused ultrasound. By activating mechanically sensitive ion channels such as Piezo1 / 2 on the membrane of medial septal nucleus neurons, it changes ion permeability and generates a controllable ion flow, thereby achieving specific regulation of deep target points.

[0110] The memory consolidation effect quantification module embeds a quantification system covering the quality of theta-state encoding and the efficiency of non-theta-state consolidation, including seven categories of indicators: the number of ripples, the duration of ripples, the number of spindles, the amplitude of spindles, the duration of spindles, the number of SOs, theta score, the percentage of co-occurrence, the synchronization index SI value, and the phase lock value PLV value.

[0111] Working process: First, a complete closed-loop network model of medial septum-hippocampus-thalamus-cortex is built using the SHTC model building module to simulate the dynamic transformation between theta and non-theta states of the hippocampus.

[0112] Then, the TMS and TUS stimulation parameters were set through the transcranial magnetic acoustic stimulation modulation module. The TMS unit applied a voltage offset term S(t) to regulate the excitability of cortical neurons and network rhythm, and the TUS unit introduced a mechanosensitive current term I. TUS (t) Regulates the activity of medial septal nucleus neurons and the intensity of hippocampal theta rhythm.

[0113] Finally, the membrane potential signals of pyramidal neurons in the CA1 region of the hippocampus were monitored using the memory consolidation effect quantification module. Ripple, spindle, and SO rhythmic events were detected, and theta score, co-occurrence percentage, SI value, and PLV value were calculated to evaluate the enhancement effect of dual-target stimulation on memory consolidation and optimize stimulation parameters.

[0114] Technical effect: By constructing a complete SHTC closed-loop circuit model and adopting a TMS / TUS dual-target synergistic stimulation strategy, covering the entire process of memory encoding and consolidation, the technical problem of single-stage regulation in existing technologies has been solved.

[0115] By embedding a seven-category quantitative indicator system, a clear correlation between stimulation parameters and therapeutic effects was established, enabling quantitative assessment of memory consolidation efficacy.

[0116] Example 2 This embodiment, based on Embodiment 1, describes in detail the network structure, differential equations, synaptic dynamics, and dynamic adjustment functions of the SHTC model building module.

[0117] The SHTC network structure comprises four core subnetworks: the medial septal nucleus, hippocampus, cortex, and thalamus.

[0118] The medial septal nucleus network contains one group each of cholinergic neurons, glutamatergic neurons, and GABAergic neurons, which work together to regulate the circuit rhythm.

[0119] The hippocampal network consists of two functional regions, CA1 and CA3. Each region contains a set of excitatory neurons and a set of inhibitory neurons, forming a local closed loop.

[0120] The cortical network is divided into two regions, CX1 and CX2. Each region is equipped with a set of excitatory neurons and a set of inhibitory neurons, which interact across brain regions with the hippocampus and thalamus through long-range projections.

[0121] The thalamic network consists of one group of TC neurons in the thalamic cortex and one group of RE neurons in the reticular nucleus, which participate in the integration and regulation of circuit information.

[0122] The network model can be described by the aforementioned 21 differential equations, including 5 equations for the medial septal nucleus network, 6 equations for the hippocampal network, 6 equations for the cortical network, and 4 equations for the thalamic network.

[0123] Additional technical benefits: By introducing dendritic spine frequency adaptation and dynamic adjustment functions, the model can accurately simulate the regulatory mechanism of the medial septal nucleus on the hippocampal theta / non-theta state transition, improving the simulation accuracy of long-term neural activity and providing a more reliable simulation basis for the optimization of dual-target stimulation parameters.

[0124] Example 3 Based on Example 1, this embodiment describes in detail the mathematical form, physical equivalence, and parameter optimization range of the voltage offset term and mechanically sensitive current term of the TMS unit and TUS unit.

[0125] like Figure 2 As shown, the target of TMS unit stimulation is the cortex, and the cortical targeted voltage offset superposition method is used to accurately simulate the neuromodulation effect of TMS.

[0126] The core design logic of this module is as follows: a voltage offset term S(t) is applied to the state variable representing the excitatory and inhibitory neuronal activity in the cortex on the right side of the differential equation of the SHTC loop network model, while the time derivative term on the left side of the equation remains unchanged, ensuring that the simulation effect is strictly limited to the cortex and does not affect the normal dynamic characteristics of other brain regions such as the medial septal nucleus, hippocampus, and thalamus.

[0127] The voltage offset term S(t) is expressed as an alternating function to accurately simulate the rhythmic driving effect of the TMS, and its expression is S(t) = A. stim ·sin(2πf stim ·t), where A stim To stimulate amplitude, the range can be flexibly adjusted from 0-100mV to adapt to different intensities of TMS effect according to simulation requirements. stim The stimulation frequency can be set as needed to characteristic frequencies such as 1Hz, 5Hz, 10Hz, 20Hz, 50Hz, and 100Hz, while t is the stimulation time variable.

[0128] From the perspective of physical equivalence, the superposition of the voltage offset term S(t) is mathematically equivalent to introducing an additional voltage field into the synaptic input of cortical neurons, which is highly consistent with the physical process of TMS changing the membrane potential of cortical neurons through induced electric fields, thus ensuring the scientific accuracy of the simulation.

[0129] To analyze amplitude A stim With frequency f stim To explore the independent impact on neuromodulation efficacy and the optimal stimulation parameter window, this invention can conduct systematic simulations by expanding the multi-dimensional parameter space: the frequency dimension covers the typical TMS stimulation frequency band of 1-100Hz, and the amplitude dimension sets the gradient range of 0-100mV. By combining different parameter combinations, the dynamic changes of core indicators related to memory consolidation can be continuously monitored and quantified.

[0130] like Figure 3 As shown, the TUS unit uses the inner partition core as the precise target point and innovatively adopts the mechanically sensitive current equivalent modeling method.

[0131] The core design logic of this module is as follows: In the SHTC loop network model, for the core state variables representing the activity of excitatory neurons (cholinergic and glutamatergic) and inhibitory neurons (GABAergic) in the medial septal nucleus, a mechanosensitive current term I is introduced on the right side of its current balance differential equation. TUS (t), the time derivative term on the left side of the equation remains unchanged, achieving specific regulation of deep target points.

[0132] The mechanically sensitive current term adopts a composite form of "amplitude-modulated sine wave + focusing gain + pulse window" to reproduce the pulse characteristics, rhythmic driving effect, and deep focusing advantages of focused ultrasound. Its expression is I. TUS(t) = A·G·sin(2πf) mod ·t)·sin(2πf US ·t)·P(t).

[0133] A represents the stimulation amplitude, ranging from 0.1 to 1.0 mV. It can be flexibly adapted to different sound pressure intensities of the TUS effect according to simulation needs, while avoiding over-excitation of the medial septal nucleus neurons.

[0134] G is the focusing gain coefficient, ranging from 1.2 to 2.0. It is used to simulate the energy focusing effect when focused ultrasound targets deep brain regions (medial septal nucleus). Real TUS focuses ultrasound energy onto the target point through a focusing lens, which will produce a certain energy gain. This coefficient can accurately match this physical characteristic and improve the realism of the simulation.

[0135] f mod The modulation frequency is selected within the 4-12Hz theta rhythm range, which can balance the penetration depth and focusing accuracy of deep brain regions.

[0136] f US The ultrasonic carrier frequency has a value range of 0-1.5MHz.

[0137] P(t) is the pulse window function (binary function), which is used to simulate the pulsed application characteristics of real TUS and avoid tissue thermal damage caused by continuous waves. Its expression is P(t) = {1, when there exists a natural number n such that n·T PRF ≤t} From a physical equivalence perspective, the mechanically sensitive current term I TUS The introduction of (t) is mathematically equivalent to simulating the mechanical vibration effect of focused ultrasound in transcranial acoustic stimulation. By activating mechanically sensitive ion channels such as Piezo1 / 2 on the neuronal membrane of the medial septal nucleus, it changes ion permeability and generates a controllable ion flow, which is consistent with the core mechanism of TUS in regulating neuronal firing through mechanical force.

[0138] At the same time, this design fully leverages the advantages of focused ultrasound in precisely focusing on deep brain regions, providing a reliable paradigm for modeling research on the neural modulation of deep nuclei.

[0139] To analyze the amplitude A and modulation frequency f of the system mod Ultrasonic carrier frequency f US To investigate the independent modulation effect and synergistic mechanism, this invention conducts research through a multi-dimensional parameter space simulation scheme: the modulation frequency is selected in the key interval of theta rhythm of 4-12Hz, the amplitude is set in the gradient range of 0.1-1.0mV, and the focusing gain coefficient is taken in the range of 1.2-2.0.

[0140] By continuously monitoring key indicators such as the synchronicity of medial septal nucleus discharge, hippocampal theta rhythm intensity (e.g., theta score), and cross-brain region circuit coupling efficiency (e.g., PLV value, percentage of co-occurrence), we can identify the optimal parameter combinations to suit different regulatory needs, providing direct theoretical support for the parameter design of subsequent in vivo medial septal nucleus-targeted TUS experiments.

[0141] Dual-target synergistic strategy: TMS uses voltage offset as the key execution parameter, TUS uses mechanically sensitive current as the core regulation carrier, and uses the membrane potential signal of pyramidal neurons in the CA1 region of the hippocampus as the core regulation basis to achieve dual-target targeted regulation.

[0142] Additional technical benefits: Through dual-target synergistic stimulation of the cortex by TMS and the medial septal nucleus by TUS, precise control of the entire process of memory encoding (theta state) and memory consolidation (non-theta state) is achieved.

[0143] TUS's mechanically sensitive current equivalent modeling method solves the technical challenge of non-invasive control of deep internal diaphragm.

[0144] Example 4 Based on Example 1, this embodiment describes in detail the detection algorithm, calculation formula, and physiological significance of the seven quantitative indicators of the memory consolidation effect quantification module.

[0145] The basic rhythmic characteristic parameters include the characteristic parameters related to ripple, spindle, and SO.

[0146] Spindle detection employs a 12-16Hz bandpass filter, Hilbert transform to calculate instantaneous amplitude, and a 300ms Gaussian window for smoothing.

[0147] After obtaining the local field potential (LFP) signal from the membrane potential of excitatory neurons in the cortical CX2 network, it is first filtered by a bandpass filter at 12-16Hz, with the filter order corresponding to three cycles of the low-frequency cutoff frequency. Then, the instantaneous amplitude of the signal is calculated by Hilbert transform and smoothed using a 300ms Gaussian window.

[0148] Candidate spindles must meet three conditions to be considered valid events: the smoothed amplitude exceeds a set threshold (1 standard deviation for model data), the duration is in the range of 0.4-3.5s, and the interval between two adjacent candidate events is less than 0.5s, which should be merged into a single spindle event.

[0149] Ripple detection employs 100-250Hz bandpass filtering, root mean square calculation, and 50ms Gaussian window smoothing.

[0150] Hippocampal LFP signals were obtained from the CA1 region (the membrane potentials of pyramidal neurons were directly used in the model data). First, the signals were filtered by a 100-250Hz fourth-order zero-phase-delay Butterworth bandpass filter. Then, the root mean square (RMS) of the filtered signals was calculated, and the RMS signal was smoothed by a 50ms Gaussian window.

[0151] When the smoothed RMS signal exceeds the set threshold (model data is median + 2 standard deviations) and the duration is ≥30ms, it is determined to be a valid ripple event.

[0152] SO detection uses a 0.5-2Hz bandpass filter and a time interval for determining positive and negative zero crossings.

[0153] After obtaining the LFP signal from the membrane potential of excitatory neurons in the cortical CX2 network, the signal is filtered from 0.5 to 2 Hz using an FIR bandpass filter, with the filter order corresponding to three cycles of the low-frequency cutoff frequency.

[0154] An LFP signal is considered a valid SO event when it meets all three conditions: First, the time interval between consecutive positive and negative zero crossings is within the range of 0.5-2s (corresponding to a frequency of 0.5-2Hz); second, the negative peak value exceeds the set threshold (1.4 times the average peak value of the model data); and third, the difference between the positive and negative peak values ​​exceeds the set threshold (1.4 times the overall average amplitude of the model data signal).

[0155] The theta score is the core spectral quantification index for bistate classification of memory processing. It is calculated based on the original signal of the membrane potential of pyramidal neurons in the CA1 region. The core is obtained by extracting the 4-12Hz theta band and the 0.5-4Hz delta band components through a third-order zero-phase-delay Butterworth bandpass filter, and solving for the average power ratio of the two.

[0156] Its core value lies in directly reflecting the strength of theta oscillation relative to delta rhythm under two states. It can quantitatively characterize the dynamic changes of rhythm spectrum during the transition from encoding to consolidation, and at the same time provide a standardized basis for determining whether theta oscillation is at a relatively high expression level. This provides a global observation dimension for analyzing the regulatory effect of exogenous intervention on the temporal transition of the entire memory cycle.

[0157] The calculation of the co-occurrence percentage requires the extraction of rhythmic events and peaks through a dedicated algorithm, namely, detecting valid events of SOs, spindles, and ripples respectively, and accurately recording the peak times of each type of event.

[0158] The next step is to determine the overlap of the core time windows. Based on the standardized window definition, three types of core co-occurrence relationships are defined: SO-ripple co-occurrence is determined by whether there is an SO peak within ±0.5s before and after the ripple peak; SO-spindle co-occurrence is determined by whether there is an SO peak within ±1.5s before and after the spindle peak; and spindle-ripple co-occurrence adopts a bidirectional determination logic, which requires determining whether there is a spindle peak within ±0.5s before and after the ripple peak, and also whether there is a ripple peak within ±1.0s before and after the spindle peak.

[0159] Finally, the number of events that meet the time window overlap condition is counted, divided by the total number of rhythmic events used as the baseline in this type of relationship, and multiplied by 100% to obtain the percentage of common occurrence.

[0160] As a core indicator for quantifying the spatiotemporal synergy of the three, its core significance lies in reflecting the coupling strength across brain regions and the characteristics of associated rhythms, while verifying the regulatory role of long-range connectivity and providing data support for the memory consolidation hypothesis.

[0161] The core of SI value calculation is to quantify the synchronous locking relationship between low-frequency oscillations and high-frequency oscillation power fluctuations.

[0162] First, signal preprocessing is performed. To ensure the accuracy of phase estimation, both the low-frequency oscillation signal and the high-frequency oscillation power signal need to be filtered to a specific frequency range of low-frequency modulation oscillation, such as SOs at 0.5-2Hz, spindles at 12-16Hz, and ripples at 100-250Hz. The filtering adopts a second-order FIR bandpass filter, with the order corresponding to three cycles of the low-frequency cutoff frequency.

[0163] Subsequently, the instantaneous phases of the two time series were extracted using Hilbert transform, where φ lt φ represents the phase value of the low-frequency oscillation band. ut The phase value represents the power fluctuation of high-frequency oscillations.

[0164] Based on this, according to the formula SI=(1 / n)Σexp[i(φ lt -φ ut ] Calculate the SI value, where n is the number of time points within the selected time interval. When calculating, the time window for phase estimation needs to be limited, ±1s before and after the SO peak and ±0.25s before and after the spindle peak.

[0165] The final SI is in complex form, with its angle reflecting the phase difference between the two oscillating events and its absolute value representing the coupling strength between them, thereby achieving a quantitative characterization of the degree of oscillation coordination in different brain regions.

[0166] The core of PLV value calculation is to quantify the lock-in stability of high-frequency rhythmic events on low-frequency phase signals.

[0167] First, using the peak time of each event in a high-frequency rhythm (such as ripples and spindles) as an anchor point, extract the instantaneous phase of the low-frequency phase modulation signal (such as SOs and spindles) at that moment after purification and filtering. Define this phase as the locked phase φ of the high-frequency event. k (k is the event number, k=1, 2, ..., N, N is the total number of valid high-frequency events).

[0168] Then all locked phases are converted into complex exponential form exp(iφ) k The average value of the summation of all complex exponents is calculated by taking the average value, and the modulus of this average value is given by the formula: PLV = |(1 / N)Σexp(iφ) k )|.

[0169] The value of this indicator ranges from 0 to 1. The closer it is to 1, the more concentrated all high-frequency events are locked in the same position in the low-frequency phase, and the stronger the phase locking. The closer it is to 0, the more random the high-frequency events are in the low-frequency phase, and the no significant locking relationship.

[0170] Additional technical benefits: Through multi-dimensional data fusion of seven quantitative indicators, the dynamic collaborative efficiency of the SHTC network is comprehensively quantified from the perspectives of state spectrum transformation (theta score), phase coupling (SI value, PLV value), co-occurrence probability (percentage of co-occurrence), and single rhythm characteristics (number / duration / amplitude of ripple / spindle / SO), thus fully depicting the organizational principles and regulatory laws of memory full-cycle rhythm collaboration.

[0171] Example 5 This embodiment uses simulation data to compare and verify various indicators between the dual-target synergistic stimulation group and the non-stimulation group, the single-target TMS group, and the single-target TUS group.

[0172] Simulation conditions: Based on the SHTC model, the TMS stimulus amplitude A is set. stim =50mV, stimulation frequency f stim =10Hz; TUS stimulation amplitude A=0.5mV, modulation frequency f mod =8Hz, Focusing gain coefficient G=1.5, Pulse width PW=200μs, Pulse repetition period T PRF =10ms.

[0173] Comparison groups: No stimulation group (SHTC model only), single-target TMS group (TMS stimulation only), single-target TUS group (TUS stimulation only), and dual-target synergistic group (TMS and TUS stimulation applied simultaneously).

[0174] The simulation results are shown in the table below: Table 1 Comparison of memory consolidation-related indicators under different stimulus conditions

[0175] Conclusion: Compared with the non-stimulation group, the dual-target synergistic group showed a 46% increase in theta score, a 40% increase in the number of ripples, a 47% increase in the number of spindles, a 42% increase in the number of SOs, an 87% increase in the percentage of co-occurrences, a 50% increase in the SI value, and a 53% increase in the PLV value.

[0176] Compared to the single-target TMS group, the dual-target synergistic group showed a 28% increase in theta score, a 24% increase in the number of ripples, a 32% increase in the number of spindles, a 25% increase in the number of SOs, a 56% increase in the percentage of co-occurrences, a 31% increase in the SI value, and a 35% increase in the PLV value.

[0177] Compared to the single-target TUS group, the dual-target synergistic group showed a 17% increase in theta score, a 16% increase in the number of ripples, a 23% increase in the number of spindles, an 18% increase in the number of SOs, a 33% increase in the percentage of co-occurrences, a 21% increase in the SI value, and a 23% increase in the PLV value.

[0178] The above data demonstrates that the dual-target synergistic stimulation strategy of TMS targeting the cortex and TUS targeting the medial septal nucleus can significantly enhance memory consolidation, achieving precise control over the entire process of memory encoding and consolidation, and bringing unexpected technical effects.

[0179] Variants Understandably, the network structure of the SHTC model is not limited to the four-region configuration of medial septal nucleus (cholinergic / glutamatergic / GABAergic neurons) + hippocampus (CA1 / CA3 area) + cortex (CX1 / CX2 area) + thalamus (TC / RE neurons), and can also be extended to a multi-region circuit model that includes other brain regions such as the prefrontal cortex, amygdala, and striatum.

[0180] Clearly, the TMS stimulation frequency f stim It is not limited to the 1-100Hz range, but can also be set to specific frequency bands such as 0.1-1Hz ultra-low frequency stimulation and 100-200Hz high frequency stimulation according to different control needs.

[0181] It is understandable that the TUS modulation frequency f modIt is not limited to the 4-12Hz theta rhythm range, but can also be set to other characteristic frequency bands such as the 12-16Hz spindle rhythm range and the 0.5-4Hz delta rhythm range.

[0182] Clearly, quantitative indicators are not limited to the seven categories (number / duration / amplitude of ripples / spindles / SOs, theta score, percentage of co-occurrences, SI value, PLV value), and other quantitative indicators such as power spectral density of gamma rhythms (30-100Hz), coherence across brain regions, and Granger causality can also be added.

[0183] Understandably, the dual-target stimulation strategy is not limited to the combination of TMS targeting the cortex and TUS targeting the medial septal nucleus. Other dual-target or multi-target combinations can also be used, such as TMS targeting the cortex and TUS targeting the hippocampus, TMS targeting the cortex and TUS targeting the thalamus, and TUS targeting the medial septal nucleus and TUS targeting the hippocampus.

[0184] The foregoing has provided a detailed description of a transcranial magnetic acoustic stimulation modulation system and method based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit, as provided in the embodiments of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas; furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

[0185] Certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that hardware manufacturers may use different names to refer to the same component. This specification and claims do not distinguish components based on differences in name, but rather on differences in function. The terms "comprising" and "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising / including but not limited to". "Approximately" means that within an acceptable margin of error, those skilled in the art can solve the technical problem and substantially achieve the technical effect within a certain margin of error. The following descriptions in the specification are preferred embodiments for carrying out this application; however, these descriptions are for the purpose of illustrating the general principles of this application and are not intended to limit the scope of this application. The scope of protection of this application shall be determined by the appended claims.

[0186] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes said element.

[0187] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0188] The foregoing description illustrates and describes several preferred embodiments of this application. However, as previously stated, it should be understood that this application is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the application concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of this application should be within the protection scope of the appended claims.

Claims

1. A transcranial magnetic acoustic stimulation modulation system based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit, characterized in that, It includes an SHTC model construction module, a transcranial magnetoacoustic stimulation modulation module, and a memory consolidation effect quantification module; The SHTC model construction module is used to construct a heterogeneous medial septum nucleus network containing cholinergic neurons, glutamatergic neurons and GABAergic neurons, and to integrate the heterogeneous medial septum nucleus network into the hippocampus-thalamus-cortex circuit to form the SHTC circuit network model. The SHTC model construction module adjusts the key parameters of the inner diaphragm loop to enable the hippocampus network to switch between theta state and non-theta state. The theta state corresponds to the memory encoding state, and the non-theta state corresponds to the memory consolidation state. The transcranial magnetoacoustic stimulation modulation module includes a TMS unit and a TUS unit. The TMS unit targets the cortex to apply a voltage offset term, and the TUS unit targets the medial septal nucleus to introduce a mechanosensitive current term. The transcranial magnetoacoustic stimulation modulation module uses the membrane potential signal of pyramidal neurons in the CA1 region of the hippocampus as the core modulation basis to achieve dual-target synergistic stimulation. The memory consolidation effect quantification module is used to quantify and evaluate the memory consolidation effect. The quantification indicators include the number and duration of ripples, the number, amplitude and duration of spindles, the number of SOs, theta score, co-occurrence percentage, synchronization index SI value and phase lock value PLV value.

2. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, In the SHTC model construction module, the inhibitory-excitation synaptic strength of CA1 and CA3 regions is dynamically adjusted according to the average firing rate of GABAergic neurons in the medial septal nucleus; the regulatory parameters of the medial septal nucleus on the hippocampus and cortex are dynamically adjusted according to the ratio of the theta band power to the delta band power of the membrane potential in CA1 region.

3. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, The voltage offset term S(t) applied by the TMS unit satisfies S(t) = A stim ·sin(2πf stim ·t), where A stim To stimulate amplitude, f stim The stimulation frequency; the A stim The value range is 0 to 100mV, and the f stim The coverage range is from 1 to 100 Hz.

4. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, The mechanically sensitive current term I introduced by the TUS unit TUS (t) satisfies I TUS (t)=A·G·sin(2πf mod ·t)·sin(2πf US ·t)·P(t), where A is the stimulus amplitude, G is the focusing gain coefficient, and f mod f is the modulation frequency. US Where A is the ultrasonic carrier frequency, and P(t) is the pulse window function; the value of A ranges from 0.1 to 1.0 mV, and f... mod A theta rhythm interval of 4 to 12 Hz is selected, and the value of G ranges from 1.2 to 2.

0.

5. The transcranial magnetoacoustic stimulation modulation system according to claim 4, characterized in that, The pulse window function P(t) is a binary function with a pulse width of 100 to 300 μs, a pulse repetition period of 5 to 15 ms, and a duty cycle of less than 50%.

6. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, In the memory consolidation effect quantification module, the detection of spindle is achieved by performing a 12 to 16 Hz bandpass filter on the membrane potential signal of pyramidal neurons in the CA1 region, extracting the instantaneous amplitude through Hilbert transform, smoothing it through a 300 ms Gaussian window, and then determining the amplitude based on an amplitude threshold; the detection of ripple is achieved by performing a 100 to 250 Hz bandpass filter on the membrane potential signal, calculating the root mean square, smoothing it through a 50 ms Gaussian window, and then determining the power based on a power threshold; the detection of SO is achieved by performing a 0.5 to 2 Hz bandpass filter on the membrane potential signal and determining the time interval between positive and negative zero crossings.

7. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, The theta score is based on the membrane potential signal of pyramidal neurons in the CA1 region. The theta and delta frequency band components are extracted by a third-order zero-phase-delay Butterworth bandpass filter, and the ratio of their average power is calculated.

8. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, The co-occurrence percentages include SO-ripple co-occurrence percentage, SO-spindle co-occurrence percentage, and spindle-ripple co-occurrence percentage; The condition for determining SO-ripple co-occurrence is that there is an SO peak within ±0.5s before and after the peak of the ripple; The condition for determining SO-spindle co-occurrence is that there is an SO peak within ±1.5s before and after the spindle peak moment; The spindle-ripple co-occurrence is determined by a two-way decision logic, that is, if a ripple peak exists within a specified time window before or after the spindle peak, or if a spindle peak exists within a specified time window before or after the ripple peak, co-occurrence is determined if either condition is met.

9. The transcranial magnetoacoustic stimulation modulation system according to claim 1, characterized in that, The synchronization index SI value is obtained by extracting the instantaneous phase of the low-frequency oscillation and the instantaneous phase of the high-frequency oscillation power fluctuation through Hilbert transform, and calculating the modulus of the complex exponential mean of the phase difference between the two.

10. A transcranial magnetic acoustic stimulation modulation method based on memory encoding and consolidation of the medial septal nucleus-hippocampus-thalamus-cortex circuit, using the transcranial magnetic acoustic stimulation modulation system according to any one of claims 1-9, characterized in that, The method includes the following steps: Step 1: Construct the SHTC loop network model. By adjusting the key parameters of the inner diaphragm loop, the hippocampal network is transformed between theta state and non-theta state, which correspond to memory encoding and memory consolidation, respectively. Step 2: Based on the SHTC loop network model, taking the membrane potential signal of the pyramidal neurons in the CA1 region as the core regulation basis, the TMS unit targets the cortex to apply a voltage offset term and the TUS unit targets the medial septal nucleus to introduce a mechanosensitive current term, thereby implementing dual-target synergistic stimulation. Step 3: Quantitatively analyze the output signal of the SHTC loop network model after dual-target synergistic stimulation, and evaluate the memory consolidation effect based on the number and duration of ripples, the number and amplitude and duration of spindles, the number of SOs, theta score, co-occurrence percentage, synchronization index SI value and phase lock value PLV value.

Images