Neuromodulation device and system

Abstract

The application discloses a nerve regulation device and system, and belongs to the technical field of nerve regulation, which comprises: a transducer unit for targeting the hepatic portal vein nerve plexus of the abdomen, the transducer element is configured to be directly covered on the abdominal skin of a user, and to deliver ultrasonic pulses to the hepatic portal vein nerve plexus of the abdomen when receiving an activation signal; a flexible fixing member for fixing the transducer element on the abdominal epidermis, and a circuit unit electrically connected with the transducer unit. The application integrates four core structures of wireless flexible integrated hardware, epidermal attachment deformable ultrasonic lens, CGM blood glucose closed loop trigger and AI intelligent prescription, fills the technical gap of non-invasive, wearable, closed loop intelligent and targeted portal vein nerve ultrasonic hypoglycemic equipment, opens up a new chronic disease treatment channel of low-intensity ultrasonic regulation of blood glucose, adapts to the auxiliary intervention of type 2 diabetes and related metabolic diseases, and expands the application boundary of medical nerve regulation and closed loop wearable medical equipment.

Description

Technical Field

[0001] This invention belongs to the field of neuromodulation technology, specifically relating to a neuromodulation device and system. Background Technology

[0002] In recent years, bioelectronic medicine has emerged as a promising new alternative intervention method in the management of metabolic diseases. Among these, the use of ultrasound equipment and systems to modulate peripheral neural circuits can effectively regulate the body's metabolic state, intervene in central glucose regulation pathways, and improve insulin sensitivity, providing an important technological direction for the non-invasive treatment and refined management of metabolic diseases.

[0003] However, current mainstream neuromodulation and ultrasound metabolic intervention technologies still have many significant limitations, making it difficult to meet the clinical needs for long-term, dynamic, portable, and comfortable treatment. The specific technical deficiencies mainly include the following aspects:

[0004] First, the technology has invasive drawbacks. Existing commonly used neurostimulation modulation methods require surgical implantation of stimulating devices, which not only poses safety risks such as postoperative infection, electrode displacement, and off-target stimulation, but also has limited stability and safety. Furthermore, implantable structures are difficult to adapt to chronic intervention scenarios such as long-term home treatment and daily mobility, greatly limiting their applicability.

[0005] Secondly, the treatment instruments are bulky and poorly adaptable to different body shapes. Traditional therapeutic ultrasound neuromodulation systems mostly adopt a large desktop equipment architecture, requiring auxiliary equipment such as function generators, radio frequency amplifiers, oscilloscopes, and large-size rigid transducer probes. The overall equipment has poor portability, a fixed structure, and lacks the ability to conform to the human body. In addition, they generally lack closed-loop feedback control functions, making them unsuitable for dynamic usage scenarios such as daily human activities, which seriously restricts the practical application of the technology.

[0006] Third, traditional coupling media have obvious defects in use. Currently, conventional ultrasound acoustic coupling gels gradually dehydrate over time, which can easily lead to unstable ultrasound impedance matching and a continuous decline in acoustic signal transmission efficiency, directly affecting the stability and consistency of treatment. At the same time, traditional coupling methods rely on rigid fixation structures to help fit the human body, which can easily cause pressure and discomfort after long-term wear, and cannot meet the comfort requirements of long-term continuous treatment.

[0007] Fourth, the fixed ultrasound focusing mode limits the therapeutic effect. Most existing ultrasound transducers adopt a fixed focusing design, which cannot dynamically adjust the acoustic energy targeting position according to anatomical differences and treatment needs, making it difficult to flexibly adapt to different surrounding nerve targets; repeated stimulation of a single anatomical site is prone to causing local tissue fatigue, and the therapeutic effect gradually diminishes under long-term intervention, resulting in poor focusing flexibility and insufficient treatment continuity.

[0008] Currently, some existing technologies have attempted to use ultrasound technology to improve metabolism and regulate blood sugar. However, each existing solution has its own technical shortcomings, and a complete, adaptable, and portable ultrasound neuromodulation metabolic intervention system that can be used for long-term applications has not yet been formed. The specific deficiencies of the existing technologies are analyzed as follows:

[0009] GE Healthcare's published PCT patent US2022 / 046440 (publication number WO2023064386) describes a technique for achieving sustained remission of diabetes by activating the delayed-response gene (FGF1) in the hypothalamus using therapeutic ultrasound. While the technique mentions the portal venous plexus as a potential stimulation target, the overall research and technology disclosure focuses on the biological mechanism of gene activation and does not disclose any hardware devices or systems.

[0010] US10188325B2 and US20160192867A1 disclose a non-invasive wearable blood glucose monitoring system based on the principle of ultrasound time-of-flight measurement. This type of solution belongs to blood glucose diagnostic monitoring equipment. Its core principle is to calculate blood glucose values ​​by detecting human tissue characteristics. It only has the function of blood glucose data collection and monitoring, and is limited to the detection and diagnosis level, without the ability to intervene in metabolic treatment.

[0011] The US patent US12165626B2 published by the Korea Institute of Machinery and Materials describes a general-purpose acoustic focusing device that achieves beam focusing through phase overlap of multiple filters. This technology falls under the category of general acoustic equipment, only realizing conventional acoustic focusing functions. It has a wide range of applications but is not specifically designed for medical neuromodulation or intervention for metabolic disorders, and cannot be directly applied to ultrasound neuromodulation treatment of metabolic diseases.

[0012] Chinese patent CN114225243A, published by the Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, describes an ultrasound neuromodulation blood glucose regulation system targeting the cervical vagus nerve. This system relies on real-time vascular imaging combined with electronic focus adjustment to achieve ultrasound intervention. However, this technical solution targets the cervical nerves, neglecting the core metabolic regulation target of the abdominal portal venous plexus. Furthermore, its focusing method is singular, and its target coverage is limited, making it difficult to meet the needs of long-term, comfortable, and dynamic clinical treatment.

[0013] In summary, the industry urgently needs a non-invasive, portable, human-conforming ultrasound technology and device that possesses dynamic ultrasound focusing capabilities, stable coupling performance, and supports closed-loop intelligent control, in order to overcome existing technological bottlenecks and improve the safety, comfort, and treatment continuity of ultrasound intervention for metabolic diseases. Summary of the Invention

[0014] The purpose of this invention is to propose a neuromodulation device and system that precisely stimulates reactive neurons responsible for glucose metabolism through a non-invasive targeted ultrasound pulse delivery device applied to the user's abdomen, thereby improving metabolism and being applied to blood glucose control in type 2 diabetes.

[0015] To achieve the above objectives, the present invention adopts the following technical solution:

[0016] In a first aspect, the present invention provides a neuromodulation device for promoting metabolism and controlling blood sugar in type 2 diabetes, comprising: a transducer unit including one or more transducer elements for targeting the portal venous plexus in the abdomen, the transducer elements being configured to be directly applied to the user's abdominal skin and delivering ultrasound pulses to the portal venous plexus in the abdomen upon receiving an activation signal; and a circuit unit electrically connected to the transducer unit, comprising a microcontroller, a pulse generator, a power supply and power circuitry, and a connector for connecting the transducer elements.

[0017] Furthermore, the neuromodulation device also includes a flexible fixation member for fixing the transducer element to the abdominal skin; the flexible fixation member has a first through hole adapted to the size of the transducer unit so that the transducer unit is in direct contact with the abdominal skin.

[0018] The circuit unit is a flexible printed circuit board integrated in a conformal patch, which is equipped with an external flexible package; the flexible printed circuit board is connected to the flexible fixing member, and a second through hole corresponding to the position of the first through hole is opened on the flexible printed circuit board.

[0019] In one embodiment, the transducer unit includes a single transducer element, the emitting surface of which is attached to a supergel lens configured to change its focal length upon compression or decompression, thereby achieving longitudinal focal length depth adjustment of the single transducer element.

[0020] Furthermore, the longitudinal focal length depth adjustment range provided by the individual transducer element and the supergel lens is 1~100 mm.

[0021] Furthermore, the individual transducer element is inserted into the second through hole; the sidewall of the transducer element is provided with a first connecting part, and the edge area of ​​the second through hole is provided with a second connecting part that cooperates with the first connecting part, so as to realize the relative fixation of the transducer element and the flexible printed circuit board.

[0022] In another embodiment, the transducer unit is a transducer array comprising multiple transducer elements; the multiple transducer elements are sequentially mounted on a flexible substrate conforming to the contour of the abdomen; when the multiple transducer elements receive an activation signal, they collectively form a spherical focusing emission surface, and multiple ultrasound beams are focused to deliver ultrasound pulses to the portal vein nerve plexus in the abdomen; the pulse generator is configured with multiple pulse generator channels corresponding one-to-one with the multiple transducer elements; the connector is a multi-channel connector.

[0023] Furthermore, the plurality of transducer elements are mounted on the flexible substrate via flexible hinges, so that each transducer element does not lose electrical continuity when the flexible substrate is in a bent state.

[0024] Furthermore, when there is an overlap between the flexible substrate and the edge region of the second through hole, a direct contact connection is used to fix the transducer array and the flexible printed circuit board relatively; when there is no overlap between the flexible substrate and the edge region of the fifth through hole, a connector is used to fix the transducer array and the flexible printed circuit board relatively.

[0025] Furthermore, the arrangement of the transducer array can be selected from rectangular grids, hexagonal grids, or linear arrays.

[0026] Furthermore, the circuit unit also includes beamforming logic, a phase delay circuit for electronic beam steering, and individual impedance matching for each transducer element to achieve electronic beam steering and electronic focusing of the transducer array.

[0027] Furthermore, the microcontroller independently manages the timing and amplitude of each pulse generator channel through a multiplexer, and controls the lateral focus adjustment of each transducer element by adjusting the phase and amplitude of the signal sent to each transducer element, so that the ultrasound beams delivered by the multiple transducer elements converge at a common focus in the portal vein plexus.

[0028] Furthermore, each of the transducer elements has a supergel lens attached to its emitting surface, the supergel lens being configured to change its focal length upon compression or decompression, thereby achieving longitudinal focal length depth adjustment for each of the transducer elements.

[0029] Furthermore, the maximum thickness of the device is between 6 mm and 10 mm.

[0030] Secondly, the present invention discloses a neuromodulation system, comprising: the neuromodulation device described in the first aspect, an external mobile device located away from the neuromodulation device, and a data communication module coupled to the microcontroller, for realizing data and signal interaction between the neuromodulation device and the external mobile device. The external mobile device is configured to store the operating data and treatment parameters of the neuromodulation device in real time.

[0031] The data includes: operational data and treatment parameters of the neuromodulation device.

[0032] Furthermore, the external mobile device includes a user interface configured to allow the user to set treatment parameters within FDA safety limits, including one or more of pulse repetition frequency, pulse duration, duty cycle, and sound pressure level.

[0033] Furthermore, the external mobile device includes a display interface for presenting the operating data of the neuromodulation device; the operating data includes one or more of the following: battery level, connection status, thermal index and mechanical index, and the user's cumulative treatment time.

[0034] Furthermore, the microcontroller is configured to deactivate after a predetermined time.

[0035] Furthermore, the data communication module includes a wired communication module and / or a wireless communication module.

[0036] Furthermore, the external mobile device is configured to connect to an external continuous glucose monitoring system to continuously acquire the user's blood glucose value; when the acquired blood glucose value exceeds the user's preset blood glucose threshold, the neural modulation device is activated so that the neural modulation device delivers ultrasound pulses to the portal vein plexus according to a pre-programmed periodic stimulation scheme, thereby regulating the user's blood glucose.

[0037] Furthermore, the blood glucose threshold is configured by the user between 120 mg / dL and 250 mg / dL.

[0038] Furthermore, the activation condition of the neural modulation device is configured such that the blood glucose values ​​of the user are obtained twice consecutively and exceed the user's preset blood glucose threshold.

[0039] Furthermore, the external mobile device also includes a safety interlock procedure: when the thermal index exceeds 0.3, the mechanical index exceeds 0.7, or the battery power is insufficient to complete the periodic stimulation program, activation of the neuromodulation device is prohibited.

[0040] Furthermore, the external mobile device includes an emergency stop procedure that, when triggered, shuts down the neuromodulation device.

[0041] Furthermore, the external mobile device also includes a machine learning model composed of a recurrent neural network, a long short-term memory network, and a gradient boosting decision tree, providing at least one of the following functions:

[0042] The learning function learns the relationship between the user's treatment parameters and blood glucose levels.

[0043] The predictive function predicts future blood glucose levels based on the relationship between the user's treatment parameters and blood glucose levels, and the user's current treatment parameters; and

[0044] The guidance function generates the optimal stimulation plan based on the user's current blood glucose level to prevent or correct hyperglycemia.

[0045] Furthermore, the user interface is configured to display the predicted blood glucose trajectory and the recommended optimal stimulation protocol, and to accept the user's confirmation, modification, or rejection of the optimal stimulation protocol.

[0046] Furthermore, the machine learning model is configured to continuously update its predicted blood glucose levels and recommended optimal stimulation protocols based on updated blood glucose values ​​and treatment parameters.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] (1) The non-invasive neuromodulation device provided by the present invention adopts an integrated patch structure with skin conformal attachment. It can achieve targeted modulation of the portal vein nerve plexus by attaching to the T8-T12 area of ​​the abdomen on the body surface. No surgical implantation is required, and the internal surgical implantation structure is completely eliminated. It avoids complications such as wound infection, body rejection, implant wear, secondary surgery for removal, and long-term implantation tissue damage caused by surgical implantation from the root. It is suitable for long-term chronic type 2 diabetes intervention, and is suitable for patients of all ages and without contraindications for internal implantation, with a wider range of applicable populations.

[0049] (2) The neuromodulation device provided by the present invention is equipped with a special acoustic supergel lens and adopts a unique mechanical compression focusing mechanism to achieve dynamic stepless modulation of the focal length and depth of the sound field. It can accurately match the differences in abdominal wall fat thickness and portal vein nerve plexus burial depth among different patients, accurately target the peripheral nerve circuit of the portal vein, improve nerve activation efficiency, and the sound field focusing stability is better than that of the electrically controlled steering ultrasound structure. It is suitable for the dynamic scenarios of daily walking and body position changes. Even after the body position shifts, it can still lock the target nerve point. In addition, the lens material is optimized to match the acoustic impedance of human soft tissue. Long-term wear can maintain stable acoustic coupling of the epidermis, greatly reduce the problems of local tissue strain, soreness and heat stress of the abdominal wall caused by ultrasound irradiation, alleviate human tissue fatigue under long-term chronic disease intervention, and improve the comfort of long-term wear.

[0050] (3) The neural regulation system provided by this invention is the first to build an integrated closed-loop architecture of continuous glucose monitoring (CGM) and ultrasound therapy. Unlike traditional timed and quantitative ultrasound intervention devices that require manual activation, this system can collect dynamic blood glucose data of patients in real time and use the blood glucose threshold as the only feedback signal for starting, stopping and adjusting parameters of ultrasound therapy. After the high blood glucose reaches the target, it will automatically trigger the delivery of low-intensity ultrasound pulses. After the blood glucose drops back to the safe range, the intervention will automatically terminate. No manual operation by medical staff or manual start and stop by the patient is required. It can follow the dynamic blood glucose fluctuations of the patient's diet, rest, medication and exercise, and adjust the timing of ultrasound intervention in real time to achieve dynamic adaptation of metabolic state to treatment, breaking the defects of traditional fixed intervention plan intervention that is delayed and excessive.

[0051] (4) The neuromodulation device provided by the present invention adopts a fully flexible wireless hardware architecture and is in the form of a flexible patch. It can conform to the curvature of the abdomen, is portable and lightweight, and is unrestricted when worn. At the same time, it is equipped with a dedicated wireless communication and wireless control module, which eliminates the limitations of wired power supply and wired signal transmission. It supports all-weather wearable use in dynamic scenarios such as home, office and outings, and opens up all usage scenarios of clinical intervention in hospitals and home self-management of chronic diseases.

[0052] (5) The neural regulation system provided by the present invention can link artificial intelligence and machine learning models to build an intelligent regulation system. It can complete multi-dimensional feature learning based on the patient's pathological and biochemical indicators, demographic data, previous diagnosis and treatment data, wearable physiological data, and ultrasound intervention history data. It has the ability to predict blood glucose trajectory, warn of hyperglycemia risk, and assess the trend of complications. At the same time, it can automatically generate individualized ultrasound frequency, sound pressure, duration, and frequency prescriptions by combining individual neural response sensitivity and blood glucose fluctuation rhythm, and complete the stratification of patient intervention subgroups. It takes into account the effectiveness and differences of intervention, solves the pain points of homogeneous treatment and poor adaptability of traditional equipment, and ensures the stability of regulation in real and complex use scenarios by relying on model sensitivity and robustness verification. Attached Figure Description

[0053] Figure 1 This is a perspective view of a user wearing a wearable, conformal, non-invasive ultrasonic pulse delivery device provided in Embodiment 1 of the present invention;

[0054] Figure 2 This is a close-up strabismus perspective view of the wearable neuromodulation device employing a single transducer system provided in Embodiment 1 of the present invention;

[0055] Figure 3 Three main orthogonal views of the neuromodulation device employing a single transducer system provided in Embodiment 1 of the present invention: top view, side view (outline), and side view (northward).

[0056] Figure 4 This is a schematic diagram of the flexible printed circuit board in the neuromodulation device provided in Embodiment 1 of the present invention;

[0057] Figure 5 A cross-sectional schematic diagram of a single transducer system neuromodulation device deployed on the epidermis according to Embodiment 1 of the present invention;

[0058] Figure 6 Three main orthogonal views of the neural modulation device employing a multi-transducer system provided in Embodiment 2 of the present invention: top view, side view (outline), and side view (northward).

[0059] Figure 7 This is a cross-sectional schematic diagram of the neuromodulation device of the multi-transducer system deployed on the epidermis provided in Embodiment 2 of the present invention;

[0060] Figure 8 This is a schematic diagram illustrating the mechanism by which the neuromodulation device provided in Embodiment 6 of the present invention generates therapeutic functions.

[0061] In the figure, 1. Piezoelectric transducer probe; 2. External flexible encapsulation; 3. Perforated skin interface pad; 4. Flexible printed circuit board; 5. Flexible substrate; 6. Annular fixation component; 7. Supergel lens; 8. Portal vein plexus; 9. Target tissue; 10. Epidermis and percutaneous region. Detailed Implementation

[0062] To better understand this invention, the following description, in conjunction with the accompanying drawings of the embodiments of this invention, will provide further details, but these descriptions are not intended to limit the scope of the invention. Various modifications and improvements made to the technical solutions of this invention by those skilled in the art without departing from the inventive concept should fall within the protection scope of this invention.

[0063] Example 1

[0064] refer to Figures 1 to 5 This embodiment discloses a conformal, non-invasive neuromodulation device worn on the abdominal skin by the user, used to promote metabolism and control blood sugar in type 2 diabetes.

[0065] Figure 1 Typical use cases of neuromodulation devices are shown, such as Figure 1 As shown, the neuromodulation device is located within the T8 to T12 anatomical region according to standard human anatomical divisions, corresponding to... Figure 1 T8-T12. This location covers the portal venous system and its associated nerve plexus. The device employs a flexible wearable structure with a smooth, contoured exterior that conforms to the natural curvature of the abdomen or other anatomical sites. It can be securely fixed to the T8-T12 anatomical sites and remain in place during daily activities including sitting, standing, walking, and sleeping.

[0066] The wearable neuromodulation device provided in this embodiment adopts a single transducer system, including: a unit transducer, a unit circuit unit, and a flexible fixing component; wherein, as shown in the example... Figure 2 and Figure 3 As shown, the unit transducer uses a piezoelectric transducer probe 1, the unit circuit unit uses a flexible printed circuit board 4 integrated in a conformal patch, and an external flexible package 2 is configured on its outer side; the flexible fastener uses an adhesive skin interface pad 3 that is attached to the user's abdominal epidermis.

[0067] In one embodiment, the external flexible package 2 is made of thermoplastic polyurethane material and has an Ecoflex silicone resin overlay, providing a waterproof rating of not less than IPX7 to protect the internal flexible printed circuit board 4 assembly from moisture and mechanical stress.

[0068] IPX7 is a waterproof rating in the IP (Ingress Protection) rating standard established by the International Electrotechnical Commission (IEC). The "X" indicates that the dustproof rating is untested or unspecified, and "7" indicates a waterproof rating of 7. This rating requires the equipment to remain operational even after being immersed in static water at a depth of 1 meter for 30 minutes, with the water temperature difference from the sample temperature not exceeding 5°C.

[0069] like Figure 2 A close-up strabismus perspective view of the neuromodulation device provided in this embodiment is shown. Figure 3 Three main orthogonal views of the neuromodulation device provided in this embodiment are shown: a top view, a side view (outline), and a side view (northward). The flexible conformal structure of the neuromodulation device is shown in the side view (outline). The top view shows the circular or elliptical footprint of the neuromodulation device, with the piezoelectric transducer probe 1 centered for targeting the portal vein plexus in the abdomen. The skin interface pad 3 is used to fix the piezoelectric transducer probe 1 to the abdominal epidermis. The entire assembly is curved along a gentle arc, showing how it wraps around the curvature of the user's abdomen without delamination or loss of acoustic coupling.

[0070] Preferably, the operating frequency of the piezoelectric transducer probe 1 is between 0.5 MHz and 3.0 MHz.

[0071] To ensure user comfort when wearing the device under clothing, the wearable neuromodulation device provided in this embodiment has a thickness of approximately 6–10 mm at its center.

[0072] Essentially, a first through hole is provided on the skin interface pad 3, corresponding to the emitting surface of the piezoelectric transducer probe 1, so that the piezoelectric transducer probe 1 can be directly placed on the user's abdominal skin and deliver ultrasound pulses to the portal vein nerve plexus of the abdomen through its emitting surface when an activation signal is received.

[0073] In one embodiment, the outer periphery of the skin interface pad 3 is provided with a ring of vents.

[0074] Figure 2 and Figure 3 The piezoelectric transducer probe 1 shown is equipped with a semi-soft or rigid polymer housing to house the piezoelectric element. The piezoelectric transducer probe 1 is a disc of 1–3 composite PZT material or other composite material with a diameter of millimeters and has a downward-facing (towards the skin) planar emitting surface.

[0075] Figure 5 The image shows a cross-section of the piezoelectric transducer probe 1 deployed on the epidermis. This section cuts through the device and the underlying tissue layer, revealing the internal structure and the pathway of the ultrasonic energy. (Reference) Figure 5 The piezoelectric transducer probe 1 has a mechanically deformable supergel lens attached to its emitting surface. The supergel lens is deployed to fill the space between the piezoelectric transducer probe 1 and the user's skin, and its focal length changes upon mechanical compression or decompression, achieving longitudinally adjustable focal depth modulation of the piezoelectric transducer probe 1. The arrows or dashed lines in the figure indicate the relationship between supergel lens compression and focal depth adjustability.

[0076] Figure 5 The marked epidermis and transdermal region 10 is the outer layer of skin, in direct contact with the skin interface pad 3. The ultrasound pulse is depicted as a ray or wavefront emitted from the piezoelectric transducer probe 1, passing through the supergel lens 7 (in which it is refracted and focused), through the skin layer and the homogeneous soft tissue layer beneath the skin, and converging at a focal point located within the portal vein plexus 8.

[0077] The attachable supergel lens works on the principle of an adjustable acoustic convex lens, where the focal length... Due to the curvature of the lens interface Harmony and acoustic refractive index According to the formula Decide.

[0078] In its uncompressed state, the supergel lens exhibits a Fresnel-like surface texture with concentric rings and grooves, reaching a maximum height of several millimeters. Under mechanical compression, the thickness of the supergel lens decreases while its interfacial curvature increases, thus shortening the focal length; conversely, upon release of mechanical compression, the thickness increases and the interfacial curvature decreases, resulting in a longer focal length. This adjustability allows for the redistribution of acoustic energy to the vicinity of the target nerve plexus, thereby reducing tissue fatigue.

[0079] In this embodiment, the dynamic longitudinal focal length depth adjustment range provided by the piezoelectric transducer probe 1 and the supergel lens attached to its emitting surface is 1~100 mm.

[0080] In one embodiment, the supergel lens comprises a polyethyl acrylate or other similar polymer having a sparse but strongly entangled cross-linked network.

[0081] Preferably, the entanglement-crosslinking ratio is 50:1.

[0082] refer to Figure 4 In this embodiment, the flexible printed circuit board 4 is electrically connected to the piezoelectric transducer probe 1, and the piezoelectric transducer probe 1 is driven by a single pulse generator channel.

[0083] Flexible printed circuit board 4 in Figure 4 The image is shown as a thin, meandering substrate that can be bent and twisted without breaking, containing a set of driving electronics. Figure 4 This illustrates how the various electronic components are distributed on the flexible printed circuit board 4 to maintain conformality while delivering high-voltage pulses to the piezoelectric transducer probe 1.

[0084] In one embodiment, the flexible printed circuit board 4 specifically integrates a microcontroller 24, a data acquisition / analog front end 26, a high-voltage pulse generator 27, an ultra-low noise single-channel variable gain amplifier (VGA) 18, a power supply and power management circuit, a wireless transceiver (Bluetooth Low Energy), an impedance matching network, a feedback channel, etc., and is connected to the piezoelectric transducer probe 1 through a single-channel connector 25.

[0085] In one embodiment, the battery is a rechargeable battery that can provide several hours of continuous operation.

[0086] In this embodiment, the flexible printed circuit board 4 is directly fixedly connected to the skin interface pad 3, and a second through hole is opened on the flexible printed circuit board 4 to allow the piezoelectric transducer probe 1 to pass through.

[0087] In one embodiment, the flexible printed circuit board 4 and the piezoelectric transducer probe 1 are fixed in the following way: a first connecting part is provided on the side wall of the piezoelectric transducer probe 1, and a second connecting part that cooperates with the first connecting part is provided at the edge of the second through hole on the flexible printed circuit board 4. The first connecting part and the second connecting part are connected by snap-fit ​​or side suction, so as to achieve relative fixation between the piezoelectric transducer probe 1 and the flexible printed circuit board 4.

[0088] In a preferred embodiment, the flexible printed circuit board 4 is connected to the skin interface pad 3 via an annular fastener 6; specifically, the annular fastener 6 is provided with a third through hole of the same size as the second through hole; the lower surface of the annular fastener 6 is fixedly connected to or detachably connected to the skin interface pad 3; at the same time, its upper surface is fixedly connected to the flexible printed circuit board 4, for fixing and supporting the edge area of ​​the second through hole and the piezoelectric transducer probe 1 fixed in this area.

[0089] Preferably, the flexible printed circuit board 4 and the skin interface pad 3 are connected by detachable connectors such as snap-fit ​​devices to facilitate the replacement of the skin interface pad 3 with decreased adhesion.

[0090] In one embodiment, the flexible printed circuit board 4 further includes an onboard display-input interface unit, enabling the connection of various display peripherals without the need for additional adapter cards or independent graphics cards.

[0091] Example 2

[0092] Similar to Embodiment 1, this embodiment provides a neural modulation device employing a multi-transducer system. The difference from Embodiment 1 is that it includes multiple transducer elements, which are arranged in an orderly manner to form a transducer array.

[0093] In this embodiment, as Figure 4 and Figure 6 As shown, a flexible substrate 5 conforming to the abdominal curve contour is provided at the second through hole on the flexible printed circuit board 4 for mounting multiple transducer elements 1.I~1.IX; similarly, the side view (outline) shows the flexible conformal structure, in which the entire transducer array is bent to wrap around the user's abdominal curve.

[0094] In one embodiment, each individual transducer element is mounted on a flexible substrate 5 via a flexible hinge, allowing the transducer array to bend in accordance with complex body curves without breaking or losing electrical continuity.

[0095] In a preferred embodiment, the transducer array is arranged in a rectangular grid, a hexagonal grid, or a linear array.

[0096] When the neuromodulation device provided in this embodiment is worn on the T8 to T12 anatomical sites of the user's abdomen, multiple transducer elements are directly covered on the user's abdominal skin and together form a spherical focusing emission surface when an activation signal is received. Multiple ultrasound beams are focused and then deliver ultrasound pulses to the portal vein nerve plexus in the abdomen.

[0097] It should be noted that the method of fixing the flexible substrate 5 depends on whether its area can overlap with the edge area of ​​the fifth through hole. When the flexible substrate 5 overlaps with the edge area of ​​the fifth through hole, the overlapping part of the two is directly connected. When the flexible substrate 5 does not overlap with the edge area of ​​the fifth through hole, multiple connectors are evenly arranged around the periphery of the fifth through hole to achieve relative fixation of the flexible substrate 5, the transducer array and the flexible printed circuit board 4.

[0098] Figure 7 The figure shows a cross-section of a multi-transducer system deployed on the epidermis and transdermal region 10. The figure highlights two key features: the first is conformality, achieved through… Figure 7As can be seen, the worn neuromodulation device clearly wraps around the user's body curves, exhibiting a close-fitting conformal structure. The left and right edges of the device curve upwards or downwards to match the abdominal contour, maintaining close contact with the user's abdominal skin. The second aspect is dynamic focusing; multiple ultrasound beams forming the focal volume are shown emitted from each transducer element 1.1 to 1.1X. Each beam initiates tissue penetration, focusing independently, and the multiple ultrasound beams converge at a common focal point in the wrapping or curved areas, which is the target neural tissue site in this embodiment—the portal vein plexus.

[0099] Unlike the single transducer system in Example 1, which achieves focusing solely through a supergel lens, the multi-transducer system adds electronic beam steering and focusing.

[0100] The neuromodulation device provided in this embodiment is configured with a multi-element circuit unit adapted to a multi-transducer system, which includes a microcontroller, a pulse generator, a power supply and power supply circuit, and a multi-channel connector for connecting the transducer array; wherein, the pulse generator is configured with multiple pulse generator channels corresponding one-to-one with multiple transducer elements.

[0101] Essentially, the multi-element circuit unit also includes beamforming logic, phase delay circuitry for electronic beam steering, and individual impedance matching for each transducer element to achieve electronic beam steering and electronic focusing of the transducer array.

[0102] Meanwhile, in this embodiment, the microcontroller 24A independently manages the timing and amplitude of each channel via the multiplexer 17. By adjusting the phase and amplitude of the signal sent to each element, it enables the device to move the focus laterally (left and right) and axially (deeper and shallower) without any mechanical movement, so that the ultrasound beams delivered by multiple transducer elements converge at a common focus within the portal vein plexus, just as... Figure 7 The multiple converging rays depicted in the text have a considerable depth of focus and converge precisely at the location of the target neural tissue.

[0103] In a multi-transducer system, supergel lenses are an optional configuration. In a preferred embodiment, a supergel lens is attached to the emitting surface of each transducer in the transducer array. The supergel lens provides coarse axial focusing, while electron beamforming provides fine adjustment and lateral scanning.

[0104] In a specific embodiment, the supergel lens can be replaced with a mechanically deformable lens made of silicon-based composite material or hydrogel with tunable acoustic properties.

[0105] In this embodiment, the configurations of the data acquisition / analog front end 26, high voltage pulse generator 27, ultra-low noise single-channel variable gain amplifier (VGA) 18, power supply and power management circuit, low power Bluetooth 19, etc. are the same as in embodiment 1, and will not be described in detail here.

[0106] Example 3

[0107] Belonging to the same concept as Embodiments 1 and 2, this embodiment provides a neuromodulation system, including: the neuromodulation device described in Embodiment 1 or Embodiment 2 and an external mobile device located away from the neuromodulation device, wherein the external mobile device is configured to store the operating data and treatment parameters of the neuromodulation device in real time.

[0108] To address the communication issue between the neuromodulation device and the external mobile device, the neuromodulation system provided in this embodiment further includes a data communication module coupled to the microcontroller, used to realize data and signal interaction between the neuromodulation device and the external mobile device.

[0109] The data here includes: operational data and treatment parameters of the neuromodulation device.

[0110] In one embodiment, the data communication module includes a wired communication module and / or a wireless communication module.

[0111] Preferably, the data communication module uses Bluetooth Low Energy for wireless communication and employs data encryption to ensure security.

[0112] In this embodiment, the external mobile device includes a display interface and a user interface. The display interface presents the real-time telemetry operating data of the neuromodulation device. The user interface is configured to allow users to set treatment parameters, view treatment history, and export data for clinical review within FDA safety limits. The treatment parameters include one or more of the following: pulse repetition frequency, pulse duration, duty cycle, and sound pressure level. The operating data includes one or more of the following: battery level, connection status, thermal index, mechanical index, and the user's cumulative treatment time.

[0113] Preferably, the external mobile device is a smartphone or tablet that can run a customized mobile application. This application connects to the continuous glucose monitoring (CGM) system via a health data application programming interface (API) to continuously acquire the user's blood glucose values.

[0114] The neuromodulation device achieves closed-loop control through wireless communication and a customized mobile application. When the user's blood glucose level, as detected by the application, exceeds the user's preset blood glucose threshold, the application sends an activation command to the neuromodulation device. This causes the device to deliver ultrasound pulses to the portal vein plexus according to a pre-programmed periodic stimulation protocol, thereby regulating the user's blood glucose.

[0115] The pre-programmed periodic stimulation protocol includes a pulse repetition frequency of 1–100 Hz, a peak rarefaction pressure of 0–1 MPa, and a duration of 1–10 minutes.

[0116] In another embodiment, the customized mobile application is configured to activate a neuromodulation device when the user's blood glucose levels exceed a preset blood glucose threshold twice in a row, causing the device to deliver ultrasound pulses to the portal vein plexus according to a pre-programmed periodic stimulation protocol to regulate the user's blood glucose.

[0117] In one implementation, the blood glucose threshold is configured by the user between 120 mg / dL and 250 mg / dL.

[0118] In one implementation, the microcontroller is configured to deactivate after a predetermined time.

[0119] In this embodiment, the customized mobile application also includes a security interlock program and an emergency stop program.

[0120] The neuromodulation device is prohibited from activation when the thermal index exceeds 0.3, the mechanical index exceeds 0.7, or the battery power is insufficient to complete the periodic stimulation program. When the emergency stop procedure is triggered, the neuromodulation device is forcibly shut down.

[0121] Furthermore, the customized mobile application in this embodiment also includes a machine learning model composed of a recurrent neural network, a long short-term memory network, and a gradient boosting decision tree, which can provide the following functions:

[0122] The learning function learns the relationship between the user's treatment parameters and blood glucose levels.

[0123] The prediction function predicts future blood glucose levels based on the relationship between the user's treatment parameters and blood glucose levels, as well as the user's current treatment parameters.

[0124] The guidance function generates the optimal stimulation plan based on the user's current blood glucose level to prevent or correct hyperglycemia.

[0125] After obtaining the predicted blood glucose level and the recommended optimal stimulation plan, it is displayed through the user interface, and the user can confirm, modify or reject the optimal stimulation plan.

[0126] In one implementation, the machine learning model continuously updates its predicted blood glucose levels and recommended optimal stimulation protocols based on updated blood glucose values ​​and treatment parameters.

[0127] Example 4

[0128] Based on the neuromodulation system provided in Example 3, this example provides a method for treating type 2 diabetes mellitus and related metabolic disorders by applying a neuromodulation device to the portal vein glucose sensing system.

[0129] The method includes:

[0130] (a) Imaging of the portal vein region using an external high-resolution ultrasound system to identify anatomical landmarks, including the portal vein, hepatic artery, and xiphoid process;

[0131] (b) Mark the abdominal epidermis;

[0132] (c) Wear the neuromodulation device described in Example 1 or Example 2 on the marked site;

[0133] (d) Delivering low-intensity pulsed ultrasound with treatment parameters optimized for glycemic regulation.

[0134] Optionally, the above method operates in a closed-loop mode with continuous glucose monitor (CGM) feedback for autonomous, threshold-triggered glucose regulation.

[0135] Example 5

[0136] This embodiment provides the transformation process from research and development to clinical application of the neuromodulation device described in Embodiment 1 or Embodiment 2. Specifically, it provides an exemplary development workflow that incorporates artificial intelligence (AI) and machine learning (ML) models into the system architecture to achieve personalized adaptive adjustment of treatment parameters, blood glucose prediction, and automatic generation of non-invasive neurohypoglycemic prescriptions. The entire process complies with FDA 21 CFR 1020.33 safety standards for ultrasound therapy devices and endocrine clinical real-world research guidelines.

[0137] The overall development workflow is divided into four phases: device research and development, transformation and animal clinical validation, pipeline development, multidimensional data closed-loop modeling, and model verification and deployment. Each phase is implemented step by step in a standardized manner.

[0138] During the device development, translation, and animal clinical validation phases, key components of the neuromodulation device (especially various morphological factors and improved iteratively designed ultrasound transducers) were conceived, developed, and rigorously validated at multiple biological scales.

[0139] The first step was a feasibility study in rodents: 80 six-week-old db / db spontaneous T2DM mice, half male and half female, with fasting blood glucose ≥11.1 mmol / L, were acclimatized for 7 days, kept at a constant temperature of 22±1℃, and allowed free access to high-fat feed and water; a blank control group of 20 mice and an ultrasound intervention group of 60 mice were set up.

[0140] The purpose of the feasibility verification experiment is to gradually identify the effective ultrasound range for portal vein nerve activation, determine the dose-response relationship of baseline hypoglycemic agents, and eliminate ineffective combinations of acoustic parameters.

[0141] The gradient test parameters were: a gradient group of 1~100mm across the entire focal depth, five gradient groups of sound pressure 0.1~1.0MPa, a gradient of 1~100Hz pulse repetition frequency, and a gradient of 1~10min single treatment duration; the upper limit of daily intervention was fixed at 4 treatment cycles within 24 hours, with single variable control between groups.

[0142] The criteria for evaluation were: portal vein nerve electrophysiological potential acquisition, fasting blood glucose reduction 2 hours after intervention, inflammatory pathological sections of liver tissue, and sections of subcutaneous tissue thermal damage caused by ultrasound irradiation of the abdominal wall.

[0143] Feasibility verification experiments determined the optimal baseline parameters for mice to be: focal depth 12 mm, sound pressure 0.4 MPa, pulse repetition frequency 20 Hz, and single intervention time 6 min.

[0144] After the feasibility verification experiment is completed, data archiving should be carried out, and device sound field data, blood glucose time series data, and neural potential data should be entered synchronously. Effective / ineffective intervention labels should be marked to construct the first generation of rodent labeled dataset.

[0145] Next is the long-term safety conversion experiment in non-human primates: the subjects are 12 adult cynomolgus monkeys whose liver anatomy and portal venous plexus anatomy conform to human physiological structure, inducing a secondary T2DM model, with a long-term intervention period of 90 days.

[0146] The experiment involved reusing the optimal parameter range for rodents, adaptively fine-tuning the focal depth by combining the abdominal wall fat thickness and portal vein burial depth in primates, conducting daily compliant interventions, regularly collecting blood samples to test glycated hemoglobin and insulin resistance indicators, implanting a biocompatibility testing patch into the abdominal wall, and evaluating the long-term skin adhesion irritation of the transducer and its in vivo acoustic conduction stability.

[0147] The outputs of this stage include: parameter correction ranges adapted to human anatomy, conclusions of safety with no hepatotoxicity, renal toxicity, or neurotoxicity in long-term intervention, unified intervention labeling standards across species, and expansion of heterogeneous datasets.

[0148] Finally, there was the human stratified clinical validation trial: The inclusion criteria for the study included 126 patients aged 18-70 years diagnosed with type 2 diabetes mellitus, who were stratified to have complications such as hypertension, hyperlipidemia, and obesity, and excluded individuals with abdominal wall skin damage, a history of portal vein surgery, or contraindications for implantation of electronic devices.

[0149] The clinical validation trial adopted a two-stage intervention model: the first stage was the hospital-based supervised intervention stage, in which medical staff manually verified the ultrasound depth of focus and intervention duration, and collected full-dimensional physiological data in a closed loop; the second stage was the off-site dynamic autonomous intervention stage, in which the AI ​​model autonomously adapted to individualized treatment parameters, patients wore neuromodulation devices for autonomous use, and the platform remotely controlled acoustic safety thresholds.

[0150] Crucially, cohort validation parameters and patient-specific parameters were derived and processed in parallel. These parameters included: optimal depth of focus range of 1–100 mm, sound pressure working window of 0.1–1.0 MPa, pulse repetition frequency of 1–100 Hz, single treatment duration of 1–10 min, and a maximum of 4 intervention cycles in 24 hours. These parameters served as controllable variables input to the AI ​​model.

[0151] During these verification phases, the ultrasonic acoustic parameters were monitored in real time by the airborne system, and the thresholds were locked in real time. The over-limit device automatically shut down and alarmed: thermal index TI < 0.3, mechanical index MI < 0.7, and spatial peak time average intensity ISPTA < 720 mW / cm².

[0152] All ultrasonic transducer components underwent sound field calibration and focus offset verification in advance, with a calibration error of ≤±0.5mm. Water bath sound field calibration was completed before each day's test, and the original calibration data was archived and stored for subsequent machine learning dataset traceability.

[0153] During the pipeline development phase, a structured approach to incorporating artificial intelligence and machine learning models into the device architecture is described.

[0154] This pipeline can include three core building blocks: task loop, large dataset classification, and supervised learning positive and negative sample labeling rules.

[0155] In the task loop module, a closed-loop feedback loop is built: multidimensional physiological data input → model output of blood glucose prediction + ultrasound prescription → device execution of intervention → CGM blood glucose result feedback → model error correction. Each patient completes no less than 6 rounds of small closed-loop iterations per day to continuously optimize individualized adaptation capabilities.

[0156] In the big data classification module, four types of original heterogeneous data were collected: dynamic CGM minute-by-minute blood glucose time-series data, ultrasound telemetry operation data of neural modulation devices, patient APP self-reported dietary and lifestyle data, and hospital structured diagnosis and treatment data; data format was unified, timestamps were aligned, and device noise was removed, and the dataset was de-identified and encrypted, and divided into a training set of 70%, a validation set of 15%, and a test set of 15%.

[0157] In the supervised learning positive and negative sample labeling rule module, negative samples are labeled as: the time segment in which blood glucose is maintained within the TIR target range (3.9–10.0 mmol / L) for 24 hours without ultrasound intervention, without the need for exogenous intervention to stabilize blood glucose; positive samples are labeled as: the effective intervention time segment in which blood glucose returns to the target range and insulin sensitivity is improved within 60 minutes after targeted ultrasound intervention via a neural modulation device during hyperglycemic episodes (blood glucose > 10.0 mmol / L). This labeling distinction makes supervised learning possible, and supervised deep learning modeling is completed based on the labeled negative and positive samples.

[0158] Once the pipeline development is complete, the workflow can be carried out in the following stages:

[0159] Training: The model is trained on a combined historical dataset across animals and humans, and iterates and re-iterates through multiple periods to minimize prediction error.

[0160] Hyperparameter fine-tuning: Adjusting hyperparameters (learning rate, batch size, network architecture depth) to optimize model performance in specific use cases of glucose neural regulation. Specific optimizations are made to postprandial blood glucose fluctuation and circadian rhythm-based glucose control performance, avoiding overfitting.

[0161] Testing: Evaluate the trained model on a reserved test dataset not seen during training to assess its generalization ability. Validation: Validate the model on prospective data (new patient cohorts or real-world dynamic data) to confirm its real-world accuracy and reliability.

[0162] Final Deployment: The validated model will be deployed on mobile applications and / or cloud infrastructure for real-time patient use. The actual AI / ML model incorporated into the device system architecture enables user-defined T2DM management self-learning, glycemic prediction, and personalized ultrasound prescription delivery.

[0163] The data analyzed and processed can include the following categories:

[0164] User pathology / blood biomarkers: HbA1c (scenario analysis of prediabetes and confirmed patients), metabolic profile (such as fasting blood glucose, postprandial blood glucose, glycemic variability), lipid profile (total cholesterol, low-density lipoprotein, high-density lipoprotein, triglycerides);

[0165] Demographic data: age, body mass index (BMI), sex, race, and other relevant demographic factors;

[0166] Previous health / medical data: complications (hypertension, dyslipidemia, cardiovascular disease), medication history (metformin, GLP-1 receptor agonist, insulin, SGLT2 inhibitor) and previous diabetes complications; wearable parameters: sleep duration / deep sleep stages, average daily steps, activity metabolic equivalent, heart rate variability (HRV), resting heart rate (RHR), sampling frequency once every 5 minutes, synchronously linked with the built-in sensor data of the neuromodulation device.

[0167] All these data types can undergo data preprocessing, which includes handling missing values, outlier removal, normalization, and time alignment.

[0168] After preprocessing, the workflow proceeds to multidimensional feature extraction, including:

[0169] Statistical time-domain characteristics: intraday average blood glucose, blood glucose standard deviation, coefficient of variation, intraday extreme blood glucose difference, and time interval (TIR) ​​within the target blood glucose range;

[0170] Frequency domain time-series characteristics: blood glucose time-series power spectral density, intraday blood glucose main fluctuation oscillation frequency;

[0171] Symmetrical distribution characteristics: skewness and kurtosis of blood glucose values, asymmetry coefficients of hyperglycemia and hypoglycemia shifts;

[0172] Physiological rhythm characteristics: diurnal blood glucose rhythm phase, daytime fluctuation amplitude, postprandial blood glucose response fitting curve, and super-diurnal metabolic rhythm coefficient;

[0173] These features are used to construct an individualized physiological feature matrix through baseline characterization (defining each patient's normal blood glucose status) and multidimensional feature integration (combining features across domains to capture complex physiological interactions), which is then input into the AI / ML model training module, where the AI / ML model learns based on the structure identified from the preprocessed data.

[0174] The AI / ML model embeds a fine-tuned medical-specific LLM agent, exclusively performing four fixed functions: first, parsing unstructured or semi-structured data (clinical notes, user-reported dietary logs, etc.); second, generating natural language interpretations of model predictions; third, providing human-computer interaction responses via the APP: answering questions about device wearability and intervention frequency; and fourth, linking with the backend to complete intervention visualization, risk visualization plotting, and personalized treatment recommendations, simultaneously generating archiveable clinical intervention reports.

[0175] The model can then derive predictive models that can identify clinically relevant biomarkers, including but not limited to: HOMA-IR (homeostasis model assessment of insulin resistance), insulin resistance (IR) index, and other relevant biomarkers (QUICKI, Matsuda index, treatment index).

[0176] Through the predictive modeling process, a true baseline is established. The true baseline refers to the empirically validated, clinically confirmed outcome that the model predictions are compared to.

[0177] In this embodiment, the true baseline may include: laboratory-measured fasting blood glucose and insulin levels, clinically diagnosed changes in blood glucose status, physician-reviewed and confirmed responses to neuromodulation device therapy, and other indicators. The establishment of the true baseline employs a funnel-like process of evaluation, regression, and classification, where model outputs are continuously evaluated against actual results.

[0178] The finalized predictive model can be operated based on three core analytical functions:

[0179] Patient stratification analysis: Based on the sensitivity of ultrasound intervention, intraday blood glucose fluctuation type, and probability of hyperglycemic episodes, patients with type 2 diabetes mellitus (T2DM) were divided into high-response, intermediate-response, and low-response subgroups. Ultrasound energy and intervention frequency prescriptions were matched differently to achieve personalized treatment plans.

[0180] Model robustness assessment: Robustness metrics ensure reliable performance in dynamic environments and assess the model's robustness to changes in input data quality, missing data, and real-world artifacts (sensor detachment, use of irregular devices); among which, the robustness qualification standard is that the prediction error increase under disturbance is less than 12%, ensuring stable operation in complex home scenarios.

[0181] Feature sensitivity analysis: By analyzing the feature sensitivity of each input feature through gradient ablation, we can identify which input features (recent blood glucose trends, time of day, and previous treatments) have the greatest impact on model prediction.

[0182] In this embodiment, the timing of postprandial food intake, daily HRV heart rate parameters, previous hypoglycemic medication type, and morning basal blood glucose were determined to be the top four core influencing factors, empowering clinicians to make intervention decisions and providing information for the interpretability of the model.

[0183] Example 6

[0184] This embodiment provides the mechanism and process by which a user generates therapeutic function when using the neuromodulation device described in Embodiment 1 or Embodiment 2.

[0185] like Figure 8 As shown, this embodiment illustrates the mechanism by which a user utilizes a neuromodulation device to generate therapeutic functions by describing a sequence of biological events from ultrasound application to metabolic effects. This diagram is best understood as a physiological flowchart or schematic diagram, specifically including the following steps:

[0186] Step 1: Deliver ultrasound pulses to the target neural circuit.

[0187] The user lies supine and relaxed. The wearable neuromodulation device is attached to the right upper quadrant of the abdomen, where the portal vein is projected. An AI model matches the individualized depth of focus and emits compliant low-intensity pulsed ultrasound (LIPUS). Treatment parameters are controlled by an onboard safety chip, locking the safety thresholds of TI, MI, and ISPTA. The ultrasound beam precisely converges on the peripheral nerve fiber layer of the portal vein, without causing high-intensity radiation damage to the liver parenchyma.

[0188] Step 2: Activation of peripheral nerves via mechanical conduction.

[0189] The ultrasound beam converges on the portal venous plexus, a network of afferent nerve fibers surrounding the portal vein and hepatic artery. These nerve fibers express mechanosensitive ion channels (e.g., PIEZO1 / 2, TRPV family) that are mechanically gated by the acoustic radiation force of the ultrasound pulse.

[0190] After ultrasound exposure, these channels open, allowing sodium and calcium ions to flow in and generate action potentials that propagate along afferent nerve fibers to the central nervous system.

[0191] Step 3: Transmission of afferent signals to the brain for metabolic regulation.

[0192] Neural signals ascend via the vagus nerve and / or visceral nerves, targeting and transmitting to the nucleus tractus solitarius (NTS) in the brainstem for signal integration. After integration, the neural projection signals direct the action of the hypothalamus and other autonomic regulatory centers, especially the arcuate nucleus and paraventricular nucleus of the hypothalamus, linking with brain regions that regulate glucose metabolism to complete the collection of physiological signals. In the control group with vagus nerve transection, the same ultrasound intervention showed no downstream glucose regulation effect, verifying the necessity of the afferent neural pathway.

[0193] Step 4: Autonomic nerve efferent regulation / sympathetic nerve outflow.

[0194] The brain integrates incoming neural signals with other physiological inputs (circulatory hormones, nutrient levels, other sensory information), and in response, the brain sends signals to the surroundings through the vagus nerve and sympathetic nerves.

[0195] These signals can increase insulin sensitivity in the liver, muscle, and adipose tissue; inhibit hepatic glucose production; and enhance glucose uptake in peripheral tissues, among other things. The end result of this cascade reaction is a measurable decrease in blood glucose concentration.

[0196] The above embodiments are only for illustrating the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made in accordance with the spirit and essence of the present invention should be covered by the present invention.

Claims

1. A neural modulation device, characterized in that, The device includes: A transducer unit comprising one or more transducer elements for targeting the portal venous plexus in the abdomen, the transducer elements being configured to be directly applied to the user’s abdominal skin and to deliver ultrasound pulses to the portal venous plexus in the abdomen upon receiving an activation signal. The circuit unit electrically connected to the transducer unit includes a microcontroller, a pulse generator, a power supply and power circuitry, and a connector for connecting the transducer element.

2. The neural modulation device according to claim 1, characterized in that, The neuromodulation device also includes a flexible fixation element for fixing the transducer element to the abdominal skin; The flexible fastener has a first through hole that matches the size of the transducer unit so that the transducer unit can directly contact the abdominal skin. The circuit unit is a flexible printed circuit board integrated in a conformal patch, which is equipped with an external flexible package. The flexible printed circuit board is connected to the flexible fastener, and a second through hole corresponding to the position of the first through hole is formed on the flexible printed circuit board.

3. The neural modulation device according to claim 2, characterized in that, The transducer unit includes a single transducer element, the emitting surface of which is attached to a supergel lens, the supergel lens being configured to change its focal length upon compression or decompression, for achieving longitudinal focal length depth adjustment of the single transducer element. The longitudinal focal length depth adjustment range provided by the individual transducer element and the supergel lens is 1~100 mm; The individual transducer element is inserted into the second through hole; The transducer element has a first connecting portion on its sidewall, and the edge region of the second through hole has a second connecting portion that cooperates with the first connecting portion, for the purpose of fixing the transducer element to the flexible printed circuit board.

4. The neural modulation device according to claim 2, characterized in that, The transducer unit is a transducer array containing multiple transducer elements; The multiple transducer elements are mounted in an orderly manner on a flexible substrate that conforms to the contour of the abdominal curve. When the multiple transducer elements receive the activation signal, they together form a spherical focusing emission surface, and after the multiple ultrasonic beams are focused, they deliver ultrasonic pulses to the portal vein nerve plexus in the abdomen. The pulse generator is configured with multiple pulse generator channels that correspond one-to-one with the multiple transducer elements; The connector is a multi-channel connector.

5. The neural modulation device according to claim 4, characterized in that, Each of the transducer elements has a supergel lens attached to its emitting surface, the supergel lens being configured to change its focal length upon compression or decompression, for achieving longitudinal focal length depth adjustment of each of the transducer elements. The arrangement of the transducer array is selected from a rectangular grid, a hexagonal grid, or a linear array; The plurality of transducer elements are mounted on the flexible substrate via flexible hinges, and each transducer element does not lose electrical continuity when the flexible substrate is in a bent state. When there is an overlap between the flexible substrate and the edge region of the second through hole, the transducer array and the flexible printed circuit board are fixed relative to each other by direct contact connection; when there is no overlap between the flexible substrate and the edge region of the fifth through hole, the transducer array and the flexible printed circuit board are fixed relative to each other by connector.

6. The neural modulation device according to claim 4, characterized in that, The circuit unit also includes beamforming logic, a phase delay circuit for electronic beam steering, and individual impedance matching for each transducer element to achieve electronic beam steering and electronic focusing of the transducer array. The microcontroller independently manages the timing and amplitude of each pulse generator channel through a multiplexer. It controls the lateral focus adjustment of each transducer element by adjusting the phase and amplitude of the signal sent to each transducer element, so that the ultrasound beams delivered by the multiple transducer elements converge at a common focus in the portal vein plexus.

7. The neural modulation device according to claim 2, characterized in that, The flexible printed circuit board is fixedly connected to the flexible fastener, or can be detachably connected. The external flexible encapsulation is made of thermoplastic polyurethane material and has an Ecoflex silicone resin overlay, providing a waterproof rating of not less than IPX7. The circuit unit also includes: an onboard display-input interface unit; The battery is a rechargeable battery.

8. A neural modulation system, characterized in that, include: The neuromodulation device according to any one of claims 1 to 7; External mobile devices located away from the neuromodulation device; and A data communication module coupled to the microcontroller is used to realize data and signal interaction between the neuromodulation device and the external mobile device; The data includes: operational data and treatment parameters of the neuromodulation device; The external mobile device is also configured to store the operating data and treatment parameters of the neuromodulation device in real time.

9. The neural modulation system according to claim 8, characterized in that, The external mobile device includes a user interface configured to allow the user to set treatment parameters within FDA safety limits, including one or more of pulse repetition frequency, pulse duration, duty cycle, and sound pressure level. The external mobile device also includes a display interface for presenting the operating data of the neuromodulation device; The operational data includes one or more of the following: battery level, connection status, thermal index, mechanical index, and the user's cumulative treatment time; The external mobile device is configured to connect to an external continuous glucose monitoring system for continuously acquiring the user's blood glucose levels. When the user's blood glucose level exceeds the user's preset blood glucose threshold, the wearable neuromodulation device is activated, so that the neuromodulation device delivers ultrasound pulses to the portal vein plexus according to a pre-programmed periodic stimulation scheme, thereby regulating the user's blood glucose.

10. The neural modulation system according to claim 9, characterized in that, The external mobile device also includes a machine learning model that provides at least one of the following functionalities: The learning function learns the relationship between the user's treatment parameters and blood glucose levels. The prediction function predicts future blood glucose levels based on the relationship between the user's treatment parameters and blood glucose levels, and the user's current treatment parameters. as well as The guidance function generates the optimal stimulation plan based on the user's current blood glucose level to prevent or correct hyperglycemia; The machine learning model is configured to continuously update its predicted blood glucose levels and recommended optimal stimulation protocols based on updated blood glucose values ​​and treatment parameters. The user interface is configured to display the predicted blood glucose trajectory and the recommended optimal stimulation protocol, and to accept the user's confirmation, modification or rejection of the optimal stimulation protocol; The external mobile device also includes a safety interlock program: when the thermal index exceeds a first threshold, the mechanical index exceeds a second threshold, or the battery power is insufficient to complete the periodic stimulation program, the activation of the neuromodulation device is prohibited; wherein, the first threshold and the second threshold are preset values; The external mobile device also includes an emergency stop procedure that, when triggered, shuts down the neuromodulation device.

Images