Breath-synchronized extracorporeal phrenic nerve electrical stimulator

Abstract

The present application relates to a kind of respiratory synchronous diaphragm nerve electric stimulator in vitro, belong to medical equipment technical field.The technical scheme of a kind of diaphragm nerve electric stimulation based on real-time breathing signal synchronization trigger is proposed for the problems of high weaning failure rate and quality of life decline caused by VIDD.The technical scheme points include: detecting inspiration negative pressure signal through nasal cavity air pressure sampling tube, control unit dynamically adjusts trigger threshold according to negative pressure signal, when negative pressure is lower than preset value and waveform slope meets the standard, trigger electric stimulation module to output biphasic pulse, pulse parameters include positive and negative alternating waveform, width 0.1ms to 1ms, frequency 10Hz to 50Hz and current intensity 5mA to 30mA, and strictly limit electric stimulation to be only output in inspiration cycle, and expiration cycle is completely blocked stimulation by hardware switch circuit.It is mainly suitable for intensive care and rehabilitation therapy scene, effectively prevent and treat VIDD, improve patient weaning success rate and quality of life.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of medical equipment, and particularly relates to a breathing-synchronized extracorporeal phrenic nerve electric stimulator. BACKGROUND

[0002] Mechanical ventilation is a key means to save patients' lives in intensive care, but its long-term use can lead to ventilator-induced diaphragmatic dysfunction (VIDD). Studies have shown that more than 80% of mechanically ventilated survivors have diaphragmatic atrophy or hypofunction, which significantly increases the weaning failure rate and can cause complications such as pulmonary infection and multiple organ failure, ultimately increasing the rate of disability and mortality. The core mechanism of VIDD is diaphragmatic disuse atrophy, and its pathological process involves molecular-level changes such as mitochondrial function damage, increased oxidative stress, and activated protein degradation pathway. Although mechanical ventilation itself is a necessary measure for treatment, the diaphragmatic dysfunction caused by it has become a difficult problem that needs to be solved in clinical practice.

[0003] Existing phrenic nerve electric stimulation technology is divided into two categories: invasive and non-invasive. The invasive method includes transvenous electrode implantation or surgical implantation of a stimulation device, which can directly act on the phrenic nerve, but has significant drawbacks: invasive procedures can cause complications such as pneumothorax, infection, and thrombosis, and the maintenance of implanted devices is complex, limiting their widespread use in critically ill patients. Non-invasive percutaneous phrenic nerve electric stimulation avoids surgical risks, but its technical bottleneck is the inability to achieve precise synchronization with the patient's breathing cycle. Existing non-invasive devices mostly use fixed frequency or manual trigger mode, which leads to the following problems:

[0004] 1. Asynchronous between man and machine: the electric stimulation does not match the patient's actual inspiration cycle, which may occur during the exhalation phase, causing mis-triggered stimulation, diaphragmatic muscle abnormal contraction, and aggravation of respiratory muscle fatigue or even damage;

[0005] 2. Stimulation parameters are rigid: fixed intensity current output cannot adapt to individual differences (such as skin impedance fluctuations and changes in breathing patterns), leading to insufficient stimulation (which cannot effectively activate the diaphragm) or excessive stimulation (which can cause muscle spasms or pain);

[0006] 3. Limitations of signal detection: traditional respiratory signal detection relies on chest and abdominal movement sensors or single threshold judgment, which is easily disturbed by coughing, body position changes, and other factors, resulting in high misjudgment rate;

[0007] 4. Lack of dynamic adaptability: the device cannot automatically adjust the stimulation duration or intensity when the patient's breathing rhythm changes, such as from calm breathing to deep breathing, resulting in decreased treatment effectiveness.

[0008] The root of the above problems lies in the insufficient accuracy of capturing and analyzing the respiratory signal in the prior art, and the lack of intelligent dynamic control logic. For example, non-invasive devices often determine the inspiration starting point through chest wall movement or a simple negative pressure threshold, but the chest wall movement signal is easily disturbed by limb activity, and a single negative pressure threshold cannot distinguish between real inspiration and temporary airway pressure fluctuations (such as swallowing action). In addition, the static setting of electrical stimulation parameters (such as fixed pulse width, frequency) cannot adapt to the physiological state changes of patients at different stages of the disease, especially ICU patients who often have complex conditions such as unstable respiratory rate and increased airway secretions.

[0009] In the process of trying to solve these problems, researchers have faced the following technical difficulties:

[0010] Accurate respiratory signal extraction: how to capture the start and end signals of the patient's inspiration action with high sensitivity under non-invasive conditions, while filtering out environmental noise and physiological interference;

[0011] Real-time dynamic adjustment: how to dynamically adjust the electrical stimulation parameters (such as pulse length, intensity) according to real-time respiratory data to ensure that the stimulation is strictly synchronized with the respiratory rhythm;

[0012] Safety redundancy design: how to implement a double protection mechanism at the hardware and software level to completely avoid mis-stimulation during the exhalation period, and prevent the device from entering an uncontrollable state due to signal loss or abnormalities.

[0013] Although existing research has tried to improve signal detection accuracy through multi-sensor fusion or complex algorithms, the results are still limited by computational delay, increased device power consumption, and other problems, making it difficult to meet the stringent requirements for real-time and reliability in intensive care scenarios. Therefore, developing a non-invasive diaphragmatic electrical stimulation system that can accurately synchronize with the respiratory cycle, adaptively adjust parameters, and be safe and redundant is still a key unmet clinical need. SUMMARY

[0014] One object of the present application is to solve the problem that non-invasive diaphragmatic electrical stimulation devices cannot be synchronized with the patient's respiratory cycle, and to avoid diaphragmatic muscle damage caused by human-machine maladjustment.

[0015] Solve the problem of traditional air pressure sampling tube falling off or unstable signal, improve the reliability and wearing comfort of nasal negative pressure detection.

[0016] Solve the problem that the fixed threshold trigger mechanism cannot adapt to the respiratory differences of individual patients, and reduce the risk of false triggering or missed triggering.

[0017] Solve the problem of air pressure signal interference by environmental noise, and improve the accuracy of inspiration trigger determination.

[0018] Solve the problem of rigid electrical stimulation parameters in dynamic respiratory patterns, and achieve adaptive matching of stimulation duration and respiratory cycle.

[0019] Solve the problem of overload or deficiency of electric stimulation current caused by changes in skin impedance, ensure the safety and effectiveness of stimulation.

[0020] Solve the problem of stimulation mismatch caused by respiratory mutations such as apnea or rapid breathing, and enhance the adaptability of the device to abnormal breathing.

[0021] Solve the risk of false triggering of electric stimulation during the expiratory period, and provide a hardware-level safety protection mechanism.

[0022] Solve the problem of tissue damage caused by charge accumulation of traditional electric stimulation waveform, and optimize pulse parameters to reduce side effects.

[0023] Solve the risk of device out-of-control caused by loss or interference of respiratory signal, and improve the fault tolerance capability of the system.

[0024] The present application provides a respiratory synchronized extracorporeal diaphragm nerve electric stimulator, comprising: an air pressure sampling tube placed in the patient's nasal cavity for detecting the negative pressure signal caused by inspiration; an air pressure sensor connected with the air pressure sampling tube for converting the negative pressure signal into an electrical signal; a control unit receiving the air pressure sensor signal and triggering electric stimulation when detecting that the negative pressure value is lower than a preset threshold Ph; an electric stimulation module outputting electric stimulation pulses to the diaphragm nerve, the envelope duration of the electric stimulation pulses being less than or equal to the inspiration duration of natural breathing, or automatically matching the inspiration period of natural breathing; the pulse waveform output by the electric stimulation module includes a positive-negative biphasic pulse, the pulse width is 0.1ms to 1ms, the frequency is 10Hz to 50Hz, and the current intensity is 5mA to 30mA; the electric stimulation pulses are prohibited from being output during the expiratory period.

[0025] Preferably, the respiratory synchronized extracorporeal diaphragm nerve electric stimulator of the present application, the air pressure sampling tube is made of flexible silica gel material, the inner diameter is 1mm to 3mm, and the outer wall is provided with anti-skid lines; the range of the air pressure sensor is-50Pa to 50Pa, and the accuracy is ±1Pa; the control unit judges the inspiration starting point by detecting the falling rate of the negative pressure signal, and triggers electric stimulation when detecting that the negative pressure value is lower than the preset threshold Ph and the negative pressure change rate exceeds 0.5Pa / ms.

[0026] Preferably, in the respiratory synchronized external phrenic nerve electrical stimulator of the present invention, the air pressure sampling tube is made of biocompatible medical silicone with an inner diameter of 1mm to 3mm and the outer wall surface is covered with a non-slip coating for stable fixation in the patient's nasal cavity; the negative pressure signal detected by the air pressure sensor is processed by a dynamic threshold algorithm, and the threshold algorithm automatically adjusts the trigger threshold Ph according to the negative pressure data of the patient's 5 consecutive respiratory cycles, with an adjustment range of -5Pa to -25Pa, and a step size of ±1Pa; the control unit monitors the waveform slope of the negative pressure signal in real time, and when it detects that the negative pressure value is lower than the trigger threshold Ph and the absolute value of the waveform slope exceeds 0.3Pa / ms, it is determined to be an effective inhalation trigger signal, and the electrical stimulation pulse is triggered.

[0027] Preferably, the respiratory synchronized external phrenic nerve electrical stimulator of the present invention has a pressure sensor range of -50Pa to 50Pa, a detection accuracy of ±0.5Pa, and a sampling frequency of 100Hz to 200Hz; the electrical signal output by the pressure sensor is processed by a baseline drift correction algorithm, and the algorithm deducts the influence of ambient air pressure fluctuations in real time, and the correction period is 1s to 3s; the control unit extracts the main frequency component of the inspiratory negative pressure signal of 0.1Hz to 2Hz through spectrum analysis, and filters out high-frequency noise ≥10Hz; the trigger threshold Ph is determined in combination with the negative pressure amplitude and the waveform slope of the negative pressure signal. When the negative pressure value is lower than -10Pa and the absolute value of the waveform slope is ≥0.2Pa / ms, the electrical stimulation pulse is triggered.

[0028] Preferably, the respiratory synchronized external phrenic nerve electrical stimulator of the present invention, the control unit analyzes the negative pressure signal in real time through a dynamic window algorithm, the algorithm comprising the following steps: calculating the mean and standard deviation of the negative pressure in three consecutive respiratory cycles, dynamically setting the trigger threshold Ph to the mean minus 1.5 times the standard deviation, and the threshold range is -8Pa to -20Pa; detecting the interval with the absolute value of the slope ≥0.25Pa / ms in the negative pressure signal waveform, and determining it as an effective inspiratory trigger window; triggering electrical stimulation only when the negative pressure value is lower than Ph within the window, and shielding interference signals in the rest of the time; the envelope duration of the electrical stimulation pulse is based on the duration of the real-time inspiratory window t 窗口 Dynamically adjust the duration of the inhalation window t 窗口 It is the time interval from the moment when the absolute value of the negative pressure signal waveform slope reaches or exceeds 0.25 Pa / ms for the first time to the moment when the negative pressure signal recovers from its peak value to less than 10% of its maximum negative pressure value. The adjustment method is:

[0029] Based on the duration of the current inspiratory time window t 窗口 , calculate the envelope duration of the next cycle through the proportional-integral control algorithm t 包络 , whose range ist 窗口 ×(1±5%);

[0030] The PWM signal is generated by the timer module to control the pulse duration output by the electrical stimulation module, ensuring that the adjustment error is less than ±5%;

[0031] The electrical stimulation module is connected to the patient's body surface through the electrode lead and updates the envelope duration parameter after each inspiration trigger.

[0032] Preferably, the respiratory synchronized extracorporeal phrenic nerve electrical stimulator of the present application, the pulse waveform output by the electrical stimulation module is a biphasic square wave containing alternating positive and negative pulses, the positive pulse width is 0.2ms to 0.5ms, the negative pulse width is 0.1ms to 0.3ms, and the pulse interval is 1ms to 5ms; the current intensity of the electrical stimulation module is dynamically adjusted according to the patient's skin impedance, and the adjustment logic is: with a preset reference voltage of 10V to 50V, divided by the real-time detected skin impedance value of 500Ω to 3000Ω, multiplied by a safety factor of 0.8 to 1.2, the output current range is 5mA to 30mA; the electrical stimulation module has an overcurrent protection circuit, which automatically cuts off the output when the real-time current exceeds 120% of the set value, and restores detection within 100ms.

[0033] Preferably, the respiratory synchronized extracorporeal phrenic nerve electrical stimulator of the present application, the envelope duration of the electrical stimulation pulse is dynamically adjusted by real-time detection of the inspiration cycle of natural respiration, the adjustment range is 0.5 seconds to 2 seconds, and the adjustment method is: the control unit extracts the inspiration duration of the continuous 3 respiratory cycles based on the air pressure sensor signal, calculates the average inspiration duration as the reference value; the envelope duration of the current respiratory cycle is set to 95% to 105% of the reference value, and does not exceed ±5% of the real-time inspiration duration; dynamic adjustment is realized by sliding window average algorithm, the window length is 5 to 10 respiratory cycles, and the weight coefficient decays with the respiratory sequence; when the patient's respiratory pattern mutation is detected, and the inspiration duration changes more than 20% of the reference value, the control unit suspends dynamic adjustment and switches to the fixed envelope duration mode, and the fixed value is set to 80% to 120% of the historical average value.

[0034] Preferably, the respiratory-synchronized extracorporeal phrenic nerve electrical stimulator of the present application, the control unit prohibits the electrical stimulation output during the expiratory period through a double-checking mechanism, specifically including: inspiration termination determination: when detecting that the negative pressure signal drops below 10% of the peak value, it is determined that the inspiration period is over; expiration signal confirmation: within a time window of 50 ms to 200 ms after the inspiration termination, if nasal positive pressure is detected, the output of the electrical stimulation module is locked; a hardware switch circuit is connected in series in the output circuit of the electrical stimulation module, controlled by the control unit, physically cutting off the stimulation pulse transmission path during the expiratory period; the locked state lasts until the next inspiration period trigger signal arrives, and during this period the control unit ignores all interference signals including noise, cough and / or swallowing action.

[0035] Preferably, the respiratory-synchronized extracorporeal phrenic nerve electrical stimulator of the present application, in the positive and negative biphasic pulse waveform output by the electrical stimulation module, the positive pulse width is 0.2 ms to 0.6 ms, the negative pulse width is 0.1 ms to 0.4 ms, and the positive pulse width is always 1.2 times to 2.0 times the negative pulse width; the pulse frequency is dynamically adjusted according to the real-time breathing frequency of the patient, and the adjustment rule is: if the breathing frequency is ≤12 times / minute, the stimulation frequency is 20 Hz to 30 Hz; if the breathing frequency is >12 times / minute, the stimulation frequency is 30 Hz to 40 Hz; the initial value of the current intensity is 10 mA to 15 mA, and within the first 5 minutes after the start of stimulation, it is gradually increased to a target intensity ≤30 mA in a manner of increasing 1 mA to 2 mA every 30 s, in order to avoid sudden increase of current causing muscle spasm; the charge balance error of the biphasic pulse is less than ±5%, and the positive and negative pulse width ratio is dynamically corrected by real-time monitoring of the pulse integral charge.

[0036] Preferably, the respiratory-synchronized extracorporeal phrenic nerve electrical stimulator of the present application, the duration of the locked state is controlled by a dynamic time window, specifically: within the time window from the start of the expiratory period to the arrival of the next inspiration trigger signal, the control unit continuously shields all non-inspiration signals; the maximum allowed duration of the time window is 1.2 times to 1.5 times the current breathing period, if the inspiration trigger signal is not detected within the time, it switches to a safe standby mode and stops all electrical stimulation output; the identification criteria of the interference signals include: a. noise components with a signal frequency higher than 5 Hz; b. cough or swallowing action with a signal amplitude less than 15% of the peak value of the current breathing period; the control unit delays for 50 ms to 100 ms after the inspiration trigger signal arrives for redundant checking, and if the checking is passed, the locked state is released and the electrical stimulation output is restored.

[0037] Beneficial effects:

[0038] The present application ensures that the diaphragm nerve stimulation is strictly synchronized with the patient's inspiration cycle by detecting the nasal cavity negative pressure signal in real time and dynamically adjusting the electric stimulation parameters, and avoids diaphragm muscle damage caused by man-machine dislocation; the double-phase pulse design reduces charge residue and improves treatment safety.

[0039] The flexible silicone material and anti-skid line design of the present application enhance the wearing stability of the air pressure sampling tube, reduce signal interruption caused by falling off or displacement, improve the reliability of negative pressure detection with high-precision air pressure sensors, and reduce the risk of false triggering.

[0040] The dynamic threshold algorithm of the present application combines with the individual respiratory data of the patient, adaptively adjusts the triggering condition, and adapts to the difference in respiratory intensity of different patients; the waveform slope determines the accurate identification of effective inspiration signals, and reduces the interference of coughing or swallowing.

[0041] The baseline drift correction and spectrum filtering technology of the present application effectively eliminate environmental air pressure fluctuations and high-frequency noise, and improve signal purity; the multi-parameter joint determination mechanism enhances the accuracy of inspiration triggering and reduces the false positive rate.

[0042] The dynamic window algorithm of the present application limits the electric stimulation to be triggered only during the effective inspiration period, avoids invalid stimulation, and adjusts the envelope length with a sliding average to ensure that the stimulation is smoothly matched with the respiratory rhythm and reduces the influence of parameter mutation on the patient.

[0043] The current dynamic adjustment mechanism based on skin impedance of the present application avoids overload or insufficient stimulation, and improves treatment safety; the overcurrent protection circuit provides hardware-level redundant protection to prevent accidental injury caused by equipment failure.

[0044] The respiratory mutation detection and fixed mode switching function of the present application enhances the adaptability of the device to abnormal breathing, avoids stimulation mismatch caused by apnea or rapid breathing, and optimizes the stability of parameter adjustment with a sliding window algorithm.

[0045] The double-checking mechanism and hardware switch circuit of the present application completely block the inspiration period stimulation, eliminate the risk of man-machine dislocation, and ensure the absolute safety of treatment during the locking period.

[0046] The asymmetric pulse waveform design of the present application reduces tissue polarization damage, and the charge balance calibration improves the safety of long-term treatment; the respiratory frequency linkage adjustment mechanism optimizes the diaphragm nerve activation efficiency.

[0047] The dynamic time window control and redundant verification mechanism of the present application prevents the device from entering an uncontrollable state due to signal loss; the quantitative interference identification standard reduces the risk of misoperation and improves the system fault tolerance. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 : Schematic diagram of artificially setting the envelope length of electric stimulation in a natural breathing mode;

[0049] Figure 2 : Schematic diagram of the automatic matching of the electrical stimulation envelope duration to the inspiratory cycle in natural breathing mode;

[0050] Figure 3 : Electrical stimulation waveform of external diaphragm pacemaker;

[0051] Figure 4 : Figure 3 A local magnified view of the waveform;

[0052] Figure 5 : Figure 3 The corresponding relationship diagram between the envelope duration of the electrical stimulation pulse (tc) and the respiratory cycle (T);

[0053] Figure 6 : Schematic diagram of the circuit principle of an external diaphragm pacemaker with inspiratory synchronization. DETAILED DESCRIPTION

[0054] The present invention will be described in further detail below in conjunction with the accompanying drawings so that those skilled in the art can implement the invention with reference to the description.

[0055] like Figures 1-6 As shown, an air pressure sampling tube is placed in the patient's nasal cavity. The sampling tube is connected to the air pressure sensor on the host to detect the negative pressure caused by inhalation.

[0056] When the negative pressure in the nasal cavity is detected to be lower than a certain threshold value Ph, electrical stimulation of the patient's phrenic nerve is initiated. The envelope duration of the electrical stimulation pulse can be set in a variety of ways, for example: the envelope duration of the electrical stimulation pulse needs to be set manually and must be less than the inhalation duration (see Figure 1 ), or the envelope duration of the electrical stimulation pulse automatically matches the inspiratory cycle of natural breathing (see Figure 2 The electrical stimulation waveform of the external diaphragm pacemaker is as follows Figure 3 The partial enlarged diagram of the waveform is shown in Figure 4 . Use positive and negative pulse electrical stimulation. Where τ is the pulse width; τ j - interval between adjacent positive and negative pulses; f - pulse frequency. Figure 3 The pulse envelope diagram is Figure 5 , where tc is the envelope duration of the electrical stimulation pulse and T is the respiratory cycle. The schematic circuit diagram of the external diaphragm pacemaker with inspiratory synchronization is shown in the figure. Figure 6 .

[0057] According to one embodiment of the present application, the inner diameter of the air pressure sampling tube can be selected as 1.0 mm, 1.5 mm, 2.0 mm or 3.0 mm, and the depth of the anti-skid pattern on the outer wall can be set in the range of 0.1 mm to 0.3 mm. The range of the air pressure sensor can be selected as -50 Pa to 50 Pa, and the accuracy can be selected as ±0.5 Pa, ±1.0 Pa or ±1.5 Pa.

[0058] The material of the air pressure sampling tube can be selected as biocompatible medical silica gel, and the anti-skid coating on the outer wall can be added with polyurethane or silica-based anti-skid material. The air pressure sensor can be selected as a MEMS micro air pressure sensor, and the packaging shell thereof can be selected as ABS plastic or aluminum alloy.

[0059] The air pressure sampling tube is assembled at the vestibule of the patient's nasal cavity and connected with the host through a hose; the air pressure sensor is assembled at the front end of the device main control board and close to the signal input interface. During the working process, the nasal cavity generates negative pressure when the patient inhales, and the air pressure sampling tube transmits the pressure change to the air pressure sensor, which converts the analog signal into a digital signal and transmits it to the control unit.

[0060] The value of the preset threshold Ph can be set as -5 Pa, -10 Pa, -15 Pa or -20 Pa, and the trigger judgment condition can be combined with the negative pressure amplitude being lower than the threshold and the waveform slope being ≥0.2 Pa / ms, 0.3 Pa / ms or 0.4 Pa / ms. The control unit can be selected as an ARM Cortex-M4 microcontroller, and the built-in ADC sampling frequency can be set as 100 Hz, 150 Hz or 200 Hz.

[0061] The hardware of the control unit can be integrated into the main control board close to the output end of the air pressure sensor; the software algorithm can deploy a dynamic threshold calculation module to update the threshold through a sliding window average method. During the working process, the control unit receives the air pressure sensor data in real time, and sends a trigger instruction to the electric stimulation module when detecting that the negative pressure signal meets the threshold and waveform slope conditions.

[0062] In the positive and negative biphasic waveform of the electric stimulation pulse, the positive pulse width can be selected as 0.2 ms, 0.4 ms or 0.6 ms, and the negative pulse width can be selected as 0.1 ms, 0.2 ms or 0.3 ms; the frequency can be set as 10 Hz, 20 Hz, 30 Hz or 40 Hz; the current intensity can be set as 5 mA, 10 mA, 15 mA or 20 mA. The envelope length adjustment range can be set as 0.5 s, 1.0 s or 1.5 s.

[0063] The circuit of the electric stimulation module can include an H-bridge driving chip, and the output electrode can be an Ag / AgCl patch electrode. The module is assembled at the output end of the device and connected to the surface projection area of the phrenic nerve of the neck of the patient through a wire. During operation, the module outputs a biphasic pulse according to the set parameters after receiving the instruction of the control unit, and cuts off the output path through the MOSFET switch circuit during the exhalation period.

[0064] According to another embodiment of the application, the inner diameter of the air pressure sampling tube can be selected to be 1.0 mm, 1.5 mm, 2.0 mm or 3.0 mm, and the depth of the anti-skid pattern on the outer wall can be set to be 0.1 mm, 0.2 mm or 0.3 mm. The material can be selected to be a biocompatible medical silicone, and the anti-skid coating on the outer wall can be added with a polyurethane or a silicone-based anti-skid material.

[0065] The air pressure sampling tube can be assembled at the vestibule of the nasal cavity of the patient and connected to the air pressure sensor interface of the main host of the device through a hose. During operation, the nasal cavity generates negative pressure when the patient inhales, and the air pressure sampling tube transmits the pressure change to the air pressure sensor, which converts the analog signal into a digital signal and transmits it to the control unit.

[0066] The range of the air pressure sensor can be selected to be -50 Pa to 50 Pa, and the accuracy can be selected to be ±0.5 Pa, ±1.0 Pa or ±1.5 Pa. The sensor can be selected to be a MEMS micro air pressure sensor, and the packaging shell can be selected to be ABS plastic or aluminum alloy material.

[0067] The air pressure sensor can be assembled at the front end of the main control board of the device, close to the signal input interface. During operation, the sensor detects the nasal cavity negative pressure signal in real time and transmits the electrical signal to the control unit for subsequent analysis.

[0068] One of the triggering conditions of the control unit can be set to be that the negative pressure change rate exceeds 0.5 Pa / ms. The specific value can be selected to be 0.4 Pa / ms, 0.5 Pa / ms or 0.6 Pa / ms. The control unit can be selected to be an ARM Cortex-M4 microcontroller, and the built-in ADC sampling frequency can be set to be 100 Hz, 150 Hz or 200 Hz.

[0069] The hardware of the control unit can be integrated into the main control board, and the software algorithm can deploy a real-time slope calculation module. During operation, the control unit continuously receives the data of the air pressure sensor, and when it is detected that the negative pressure value is lower than the preset threshold Ph and the falling rate of the negative pressure signal exceeds the set threshold, the control unit immediately sends a triggering instruction to the electric stimulation module.

[0070] The envelope duration adjustment range of the electric stimulation pulse can be set to be 0.5 s, 1.0 s, 1.5 s or 2.0 s. The dynamic adjustment algorithm can be realized by a sliding window average method, and the window length can be selected to be 5 to 10 breathing cycles.

[0071] The dynamic adjustment function can be integrated in the software of the control unit. In the working process, the control unit dynamically adjusts the envelope duration of the electric stimulation pulse according to the real-time detected inspiration cycle, so as to match it with the current breathing rhythm of the patient.

[0072] Through the above-mentioned embodiments, the stable wearing and signal acquisition of the air pressure sampling tube can be ensured, the high-precision sensor improves the reliability of negative pressure detection, the dynamic trigger condition reduces the risk of misjudgment, the envelope duration self-adaptive adjustment enhances the synchronization of treatment and breathing cycle, and the safety and effectiveness of the equipment are improved.

[0073] Through the above-mentioned embodiments, stable detection and high-precision triggering of the inspiration negative pressure signal can be realized, and the electric stimulation pulse and the breathing cycle of the patient are strictly synchronized; the adjustable range of the biphasic pulse parameter adapts to the needs of different patients, and the hardware switch circuit completely blocks the expiratory period mis-stimulation, thereby improving the safety and effectiveness of the treatment.

[0074] According to another embodiment of the application, the adjustment range of the trigger threshold Ph can be set to -5 Pa, -10 Pa, -15 Pa or -20 Pa, and the adjustment step can be selected as ±1 Pa or ±2 Pa. The dynamic threshold algorithm can be realized by a sliding window method, and the window length can be selected as 5 consecutive breathing cycles.

[0075] The algorithm can be deployed in the software module of the control unit. In the working process, the control unit calculates the mean and standard deviation according to the negative pressure data of the last 5 breaths of the patient, and automatically updates the trigger threshold Ph to adapt to the changes of individual respiratory intensity.

[0076] The determination threshold of the waveform slope can be set to 0.2 Pa / ms, 0.3 Pa / ms or 0.4 Pa / ms. The control unit can use an ARM Cortex-M4 microcontroller with a built-in real-time differential calculation module, and the sampling frequency can be set to 100 Hz, 150 Hz or 200 Hz.

[0077] The hardware of the control unit can be integrated into the main control board, and the software algorithm can monitor the slope change of the negative pressure signal in real time. In the working process, when the absolute value of the waveform slope is detected to exceed the set threshold, the control unit determines that it is an effective inspiration signal, and immediately triggers the electric stimulation module to output a pulse.

[0078] Through the above-mentioned embodiments, the biocompatible material and the anti-slip coating ensure the stable fixation of the air pressure sampling tube and reduce signal interruption; the dynamic threshold algorithm adaptively adjusts the trigger condition according to the individual breathing data of the patient, thereby reducing the risk of false triggering; the waveform slope determination accurately distinguishes between real inspiration and interference signals, thereby improving the triggering accuracy.

[0079] According to another embodiment of the application, the range of the air pressure sensor can be selected as -50 Pa to 50 Pa, for example -50 Pa, 0 Pa, or 50 Pa. The detection accuracy can be set as ±0.5 Pa, for example, when the actual detection value is -10.5 Pa, the system determines that it is within the range of -10.0 Pa to -11.0 Pa. The sampling frequency can be selected as 100 Hz, 150 Hz, or 200 Hz, for example, the air pressure signal is continuously sampled at 150 Hz. The sensor can be selected as a MEMS micro air pressure sensor, the packaging material can be selected as ABS plastic or aluminum alloy, and the internal sensing element can be based on the piezoresistive principle. The air pressure sensor can be assembled near the front signal input interface of the device main control board and connected to the control unit through a flexible circuit. During operation, the sensor collects the nasal negative pressure signal in real time, updates the data every 1 ms to 10 ms, and transmits it to the control unit through the SPI or I2C interface.

[0080] The electrical signal output by the air pressure sensor is processed by a baseline drift correction algorithm, and the correction period is 1 s to 3 s.

[0081] The baseline drift correction algorithm can subtract the environmental air pressure fluctuation in real time, for example, by using the sliding window average method, and the window length can be set as 1 s, 2 s, or 3 s. During correction, the algorithm calculates the average value of the air pressure signal in the current window as the baseline, for example, if the baseline is -2 Pa, the subsequent signal value is subtracted from the baseline. The correction period can be set as updating the baseline value every 1 s, 2 s, or 3 s. The algorithm can be embedded in the firmware of the control unit, for example, a filter function written in C language. During operation, the control unit recalculates the baseline value and updates the correction parameters after receiving 100 sampling points (about 0.67 s, 150 Hz sampling rate), to ensure signal stability.

[0082] The control unit extracts the main frequency component 0.1 Hz to 2 Hz of the inhalation negative pressure signal through spectral analysis and filters out high frequency noise ≥10 Hz.

[0083] The spectral analysis can be based on Fast Fourier Transform (FFT), the frequency resolution can be set as 0.1 Hz, and the analysis window length can be selected as 5 s or 10 s. The main frequency extraction range is limited to 0.1 Hz to 2 Hz, for example, the breathing fundamental frequency of 1.2 Hz is identified. The high frequency noise filtering can use a digital low pass filter, the cutoff frequency can be set as 10 Hz, and the filter type can be selected as a Butterworth second-order filter. The control unit can be selected as an ARM Cortex-M4 microcontroller, which has a built-in DSP library to implement FFT operation. During operation, the control unit performs spectral analysis every 5 s to extract the current breathing frequency and filter high frequency interference signals (such as device electromagnetic noise) in real time.

[0084] The determination of the triggering threshold Ph combines the negative pressure amplitude and the waveform slope, and the electric stimulation pulse is triggered when the negative pressure value is lower than -10 Pa and the absolute value of the waveform slope is greater than or equal to 0.2 Pa / ms.

[0085] The triggering threshold Ph can be set to -10 Pa, for example, the amplitude condition is met when the detected negative pressure is -12 Pa. The waveform slope determination can be based on the difference calculation of adjacent sampling points, for example, if the air pressure of two consecutive sampling points decreases by 4 Pa and the time interval is 20 ms, the slope is -0.2 Pa / ms. The control unit can calculate the absolute value of the slope in real time, and the triggering condition is met when the absolute value reaches 0.2 Pa / ms, 0.3 Pa / ms or 0.4 Pa / ms. The logical relationship between the threshold and the slope is an “and” condition, which needs to be met at the same time. During operation, the control unit detects the negative pressure value and the slope every 1 ms, and if both meet the standard, a triggering signal is sent to the electric stimulation module.

[0086] By using a high-precision air pressure sensor and a baseline drift correction algorithm, the influence of environmental interference on the signal is reduced, and the stability of the negative pressure detection is improved. The spectrum analysis and low-pass filtering technology effectively distinguish the respiratory signal from high-frequency noise, and reduce the risk of false triggering. The multi-condition determination mechanism combining threshold and slope enhances the accuracy of inspiration triggering and avoids single parameter misjudgment. The overall scheme can improve the synchronization accuracy of electric stimulation and respiratory cycle, and ensure the safety and effectiveness of treatment.

[0087] According to another embodiment of the application, the dynamic window algorithm calculates the mean and standard deviation of the negative pressure based on three consecutive respiratory cycles, and sets the triggering threshold Ph to be the mean minus 1.5 times the standard deviation, with a threshold range of -8 Pa to -20 Pa.

[0088] The number of consecutive respiratory cycles can be set to 3, for example, the negative pressure data of the last 3 respiratory cycles is calculated. The mean calculation can be based on the sliding window method, for example, the mean is updated every 3 breaths. The standard deviation calculation can use the sample standard deviation formula, and the specific value of the triggering threshold Ph can be the mean minus 1.5 times the standard deviation, for example, the threshold range can be selected as -8 Pa, -12 Pa, -15 Pa or -20 Pa. The control unit can use an ARM Cortex-M4 microcontroller, the packaging material can be ABS plastic, and the internal memory can store respiratory cycle data. The control unit can be assembled in the central processing area of the device main control board and connected with the air pressure sensor through a flat cable. During operation, the control unit recalculates the mean and standard deviation every 3 respiratory cycles, and dynamically updates the threshold Ph, for example, if the mean is -10 Pa and the standard deviation is 2 Pa, Ph is set to -13 Pa.

[0089] The interval with an absolute slope value greater than or equal to 0.25 Pa / ms in the negative pressure signal waveform is determined as an effective inspiration triggering window.

[0090] The slope detection can be based on the difference between adjacent sampling points, for example, when the time interval is 1 ms, the air pressure change of 0.25 Pa meets the condition. The slope threshold can be set to 0.25 Pa / ms, 0.3 Pa / ms or 0.35 Pa / ms. The control unit can have a built-in differential calculation module, for example, by comparing the air pressure difference between the adjacent two sampling points in real time. The slope determination interval can cover the entire inspiration stage, for example, from the inspiration starting point to the peak. The control unit can be assembled in the signal processing area of the main control board, close to the ADC conversion module. During operation, the control unit calculates the slope every 1 ms, and when the absolute value of the slope of multiple consecutive sampling points is ≥0.25 Pa / ms, it is marked as an effective trigger window.

[0091] Only in the effective trigger window and the negative pressure value is lower than Ph, the electric stimulation is triggered, and the rest of the period is shielded from interference signals.

[0092] The trigger logic is an "and" condition, for example, the negative pressure needs to be lower than -12 Pa and the slope is ≥0.25 Pa / ms. Interference signal shielding can be achieved through software flag bits, for example, all signals are ignored in the non-trigger window. The shielding period can cover the expiration cycle and the inspiration invalid stage. The control unit can integrate a digital switch circuit, for example, by controlling the enable end of the electric stimulation module through GPIO. During operation, when the effective window is detected and the negative pressure meets the standard, the control unit sends a trigger signal to the electric stimulation module; otherwise, the output is maintained in the closed state.

[0093] The envelope duration of the electric stimulation pulse is adjusted according to the duration of the real-time inspiration window t 窗口 Dynamic adjustment, the duration of the inspiration window t 窗口 The time interval from the start time when the absolute value of the waveform slope of the negative pressure signal first reaches or exceeds 0.25 Pa / ms to the time when the negative pressure signal rises from the peak to below 10% of its maximum negative pressure value, the adjustment error is less than ±5%, and the adjustment method is:

[0094] Based on the duration of the current inspiration time window t 窗口 The envelope duration of the next cycle is calculated by a proportional-integral control algorithm t 包络 Its range is t 窗口 ×(1±5%);

[0095] The PWM signal is generated by the timer module to control the pulse duration output by the electric stimulation module, ensuring that the adjustment error is less than ±5%;

[0096] The electric stimulation module is connected to the patient's body surface through electrode leads and updates the envelope duration parameter after each inspiration trigger.

[0097] The envelope duration adjustment can be based on the duration of the real-time inhalation window, for example, if the duration of the inhalation window t 窗口 is 1.2 s, the envelope duration is set to 1.14 s to 1.26 s (±5%). The dynamic adjustment algorithm can use proportional-integral control, for example, to correct the output according to the difference between the current and historical inhalation durations. The control unit can have a built-in timer module, for example, to control the duration of the electrical stimulation pulse through a PWM signal. The electrical stimulation module can be mounted at the output of the device and connected to the patient's body surface through electrode leads. During operation, the control unit measures the duration of the inhalation window after each inhalation trigger and adjusts the envelope duration for the next cycle accordingly t 包络 , ensuring that the error does not exceed ±5%.

[0098] Technical effects: The dynamic window algorithm adjusts the threshold value adaptively based on real-time respiratory data, improving the adaptability of the trigger condition to individual differences. The combined slope and threshold determination mechanism reduces the risk of false triggering and ensures precise synchronization of stimulation with the effective inhalation phase. Dynamic adjustment of the envelope duration and error control optimize the timeliness of the phrenic nerve stimulation, avoiding premature or late termination of stimulation. The overall scheme enhances the compatibility of the device for different breathing patterns, improving treatment safety and effectiveness.

[0099] According to another embodiment of the present application, the pulse waveform output by the electrical stimulation module is a biphasic square wave, containing alternating positive and negative pulses, with a positive pulse width of 0.2 ms to 0.5 ms, a negative pulse width of 0.1 ms to 0.3 ms, and a pulse interval of 1 ms to 5 ms. The positive pulse width can be selected as 0.2 ms, 0.3 ms, or 0.5 ms, and the negative pulse width can be selected as 0.1 ms, 0.2 ms, or 0.3 ms, for example, with a positive pulse of 0.3 ms and a negative pulse of 0.2 ms output alternately. The pulse interval can be set to 1 ms, 3 ms, or 5 ms, for example, with a delay of 3 ms after the end of the positive pulse to output the negative pulse. The electrical stimulation module can use an H-bridge driving circuit, and the core chip can support biphasic waveform output. The pulse generation circuit can be mounted at the output of the device main control board, close to the electrode interface. During operation, the control unit sends timing signals to the H-bridge to drive the electrodes to output alternating positive and negative pulses according to the set parameters, and accurately controls the pulse width and interval through the timer.

[0100] The current intensity is dynamically adjusted according to the patient's skin impedance, with the adjustment logic being the preset reference voltage of 10 V to 50 V divided by the real-time detected skin impedance value of 500 Ω to 3000 Ω, multiplied by a safety factor of 0.8 to 1.2, with the output current range being 5 mA to 30 mA.

[0101] The reference voltage can be selected as 10V, 30V or 50V, for example, 30V is preset as the initial voltage. The skin impedance detection can select a four-electrode impedance measurement module, and the real-time detection range covers 500Ω to 3000Ω, for example, when the impedance is detected as 1500Ω, the current is calculated as 30V / 1500Ω*1.0=20mA. The safety factor can be set to 0.8, 1.0 or 1.2, for example, when the impedance fluctuates, the output current is reduced by 0.9 factor. The current adjustment circuit can be integrated in the electrical stimulation module, including a voltage conversion and feedback control unit. During operation, the control unit reads the impedance value every 100ms, recalculates the target current, and adjusts the output voltage through the DAC module to ensure that the current is stable within the set range.

[0102] The overcurrent protection circuit in the electrical stimulation module automatically cuts off the output when the real-time current exceeds 120% of the set value, and restores the detection within 100ms.

[0103] The overcurrent threshold can be set to 120% of the set current, for example, when the target current is 20mA, the trigger threshold is 24mA. The protection circuit can select a fast recovery fuse or a MOSFET switch, and the response time is less than 10ms. The detection recovery period can be set to 100ms, for example, after cutting off the output, the current state is detected again after a delay of 100ms. The overcurrent protection circuit can be assembled on the power input side of the electrical stimulation module and connected in series in the output loop. During operation, the real-time current is converted into a voltage signal by a sampling resistor, which is monitored by a comparator circuit. If the current exceeds the limit, the comparator triggers the MOSFET to disconnect the circuit, and resets after 100ms to allow output again.

[0104] Technical effects: The dual-phase square wave design reduces charge accumulation and reduces the risk of tissue damage. The dynamic current adjustment mechanism adapts to the differences in skin impedance of different patients, avoiding overload or insufficient stimulation. The overcurrent protection circuit provides hardware-level safety protection to prevent discomfort caused by accidental current exceeding the standard. The overall scheme ensures the treatment effect while significantly improving the safety and reliability of the device.

[0105] According to another embodiment of the application, the control unit extracts the inhalation duration of the last three consecutive breathing cycles based on the air pressure sensor signal, and calculates the average inhalation duration as the reference value.

[0106] The number of continuous breathing cycles can be set to 3 times, for example, the latest 3 inhalation durations of 1.2 s, 1.3 s and 1.1 s are extracted. The mean value calculation can use the arithmetic mean method, for example, the reference value 1.2 s is calculated. The control unit can use an ARM Cortex-M4 microcontroller, the packaging material can use ABS plastic, and the internal memory can store breathing cycle data. The control unit can be assembled in the central processing area of the device main control board and connected with the air pressure sensor through the wire. During the working process, the control unit extracts the inhalation start and end time points from the air pressure signal after completing 3 breathing cycles, calculates the inhalation duration of each time and updates the reference value.

[0107] The envelope duration of the current breathing cycle is set to 95% to 105% of the reference value, and does not exceed ±5% of the real-time inhalation duration.

[0108] The envelope duration adjustment range can be set to 95% (such as 1.14 s) to 105% (such as 1.26 s) of the reference value, for example, when the reference value is 1.2 s, the envelope duration is set to 1.14 s to 1.26 s. The ±5% limit of the real-time inhalation duration can ensure synchronization with the current actual breathing, for example, when the real-time inhalation duration is 1.3 s, the envelope duration does not exceed 1.365 s or is lower than 1.235 s. The dynamic adjustment algorithm can be integrated into the firmware of the control unit, for example, by modifying the output duration through proportional control. During the working process, the control unit calculates the final envelope duration through linear interpolation according to the current reference value and the real-time inhalation duration, and sends it to the electric stimulation module for execution.

[0109] The dynamic adjustment is realized through the sliding window average algorithm, the window length is 5 to 10 breathing cycles, and the weight coefficient decays with the breathing sequence.

[0110] The sliding window length can be selected as 5, 7 or 10 breathing cycles, for example, 5 breathing cycles as the window. The weight coefficient can decay with time, for example, the weight of the latest breathing is 0.4, the previous one is 0.3, and so on. The algorithm can be embedded in the DSP module of the control unit, for example, using the weighted average function implemented in C language. During the working process, the control unit removes the oldest data in the window and recalculates the weighted average value as the new reference value every time a breathing cycle data is added, and updates the envelope duration parameter.

[0111] When the inhalation duration change is detected to exceed 20% of the reference value, switch to the fixed envelope duration mode, and the fixed value is 80% to 120% of the historical average value.

[0112] The mutation threshold can be set to 20% of the reference value, for example, when the reference value is 1.2s, the mutation determination threshold is ±0.24s. The fixed envelope length can be set to 80% (such as 0.96s) or 120% (such as 1.44s) of the historical average value. The control unit can be built-in abnormality detection module, for example, by comparing the difference between the current inspiration time and the reference value. During operation, if it is detected that the current inspiration time exceeds 1.44s or is lower than 0.96s, the control unit suspends dynamic adjustment, and fixes the envelope length to the average value of the last 10 breath cycles (such as 1.15s), until the breathing pattern recovers to stable.

[0113] Technical effects: The reference value is calculated by continuous breath cycle data, which improves the matching degree of envelope length and actual patient breathing. The sliding window algorithm and weight decay mechanism optimize the smoothness of dynamic adjustment, reducing parameter mutation interference. The breath mutation detection and fixed mode switching enhance the adaptability of the device to abnormal conditions, avoiding the risk of stimulation out of step. The overall scheme ensures that the electric stimulation pulse remains accurately synchronized when the breath cycle changes, ensuring treatment safety and effectiveness.

[0114] According to another embodiment of the application, the inspiration termination determination is achieved by detecting that the negative pressure signal drops below 10% of the peak value.

[0115] The peak determination can be based on the maximum negative pressure value in the current breath cycle, for example, when the peak value is -20Pa, the termination threshold is -2Pa. The control unit can use an ARM Cortex-M4 microcontroller, and the packaging material can be ABS plastic. The internal ADC module can sample the air pressure signal in real time. The control unit can be assembled in the signal processing area of the device main control board, close to the output end of the air pressure sensor. During operation, the control unit continuously tracks the negative pressure peak value, and when it detects that the real-time negative pressure value rises above 10% of the peak value (such as from -20Pa to -2Pa), it determines that the inspiration cycle is over and marks the beginning of the expiration phase.

[0116] The expiration signal confirmation detects the nasal positive pressure within a time window of 50ms to 200ms after inspiration termination.

[0117] The delay time window can be selected as 50ms, 100ms or 200ms, for example, set to 100ms. The positive pressure detection threshold can be set to +1Pa to +5Pa, for example, when +3Pa is detected, it is confirmed to enter the expiration phase. The air pressure sensor can be a MEMS micro sensor with a range of -50Pa to +50Pa and a detection accuracy of ±1Pa. The sensor can be assembled at the connection port of the nasal sampling tube and connected to the main control board through a hose. During operation, the control unit starts a timer after inspiration termination, and detects the nasal pressure after a delay of 100ms. If the pressure turns positive and lasts more than the threshold, the expiration confirmation signal is triggered, and the electric stimulation output is locked.

[0118] A hardware switch circuit is connected in series in the output circuit of the electrical stimulation module, physically cutting off the transmission path of the stimulation pulse during the exhalation period.

[0119] The hardware switch circuit can be a MOSFET or a relay module, for example, an N-channel MOSFET is used as the switching element. The on-resistance of the switch circuit can be less than 0.1Ω, and the withstand voltage value covers 30V to 60V. The circuit can be assembled at the output end of the electrical stimulation module, connected in series between the electrode lead and the driving chip. During operation, the control unit sends a turn-off signal to the MOSFET after confirming exhalation, making it in the off state, completely blocking the transmission path of the electrical stimulation pulse. Before the next inhalation trigger signal arrives, the switch circuit remains open, ensuring no stimulation output during the exhalation period.

[0120] The lock state lasts until the next inhalation trigger signal arrives, during which the control unit ignores interference signals such as noise, coughing, or swallowing.

[0121] Interference signal shielding can be achieved through software filtering algorithms, such as setting amplitude thresholds (e.g., less than 15% of the current breathing cycle peak) and frequency thresholds (e.g., higher than 5Hz). The control unit can have a built-in digital filter module, such as an IIR filter to filter out high-frequency noise. The release condition of the lock state is to detect a valid inhalation trigger signal (e.g., negative pressure below -10Pa and slope ≥0.2Pa / ms). During operation, the control unit continuously monitors the signal during the lock period and only responds to signals that meet the inhalation trigger conditions, while other interference signals are directly discarded, ensuring strict synchronization of stimulation output.

[0122] Technical effects: The dual-determination mechanism (peak drop and positive pressure detection) improves the accuracy of exhalation phase recognition, avoiding premature or late state switching. The hardware switch circuit provides physical-level signal blocking, completely eliminating the risk of exhalation period mis-stimulation. The interference signal shielding algorithm enhances the system's anti-interference ability, ensuring the absolute safety of treatment during the lock period. The overall solution realizes high-precision synchronization of electrical stimulation and breathing cycles, ensuring the safety and effectiveness of the patient's treatment process.

[0123] According to another embodiment of the present application, in the positive and negative biphasic pulse waveform output by the electrical stimulation module, the positive pulse width is 0.2ms to 0.6ms, the negative pulse width is 0.1ms to 0.4ms, and the positive pulse width is 1.2 times to 2.0 times the negative pulse width.

[0124] The positive pulse width can be selected as 0.3 ms, 0.4 ms or 0.5 ms, and the negative pulse width can be selected as 0.2 ms, 0.3 ms or 0.4 ms, for example, the combination of a positive pulse of 0.4 ms and a negative pulse of 0.3 ms meets the 1.33 times ratio. The electric stimulation module can select an H-bridge driving circuit, and the core chip can support adjustable pulse width output. The driving circuit can be assembled on the output end of the device main control board, close to the electrode interface. During operation, the control unit generates a timing signal through the timer module to accurately control the width and interval of the positive and negative pulses, for example, the positive pulse is switched to the negative pulse of 0.3 ms immediately after 0.4 ms, to ensure that the ratio meets the set range.

[0125] The pulse frequency is dynamically adjusted according to the real-time breathing frequency of the patient, and the stimulation frequency is 20 Hz to 30 Hz when the breathing frequency is ≤12 times / minute, and 30 Hz to 40 Hz when the breathing frequency is >12 times / minute.

[0126] The breathing frequency detection can be based on the air pressure sensor signal, for example, by calculating the time difference between the starting points of two consecutive inhalations. The stimulation frequency can be selected as 20 Hz, 25 Hz or 30 Hz (when the breathing frequency is ≤12 times / minute), or 30 Hz, 35 Hz, 40 Hz (when the breathing frequency is >12 times / minute). The control unit can integrate frequency switching logic, for example, when the breathing frequency is detected as 10 times / minute, the stimulation frequency is set to 25 Hz. During operation, the control unit updates the breathing frequency every 10 s, and adjusts the output frequency of the electric stimulation module according to the preset rule, to ensure that the stimulation parameters match the breathing rhythm.

[0127] The initial value of the current intensity is 10 mA to 15 mA, and it is increased by 1 mA to 2 mA every 30 s within the first 5 minutes after the start of stimulation, and gradually increased to a target intensity of ≤30 mA.

[0128] The initial current can be set to 10 mA, 12 mA or 15 mA, for example, starting from 12 mA and gradually increasing. The increment step can be selected as 1 mA or 2 mA, for example, increasing by 1 mA every 30 s, from 12 mA to 20 mA within 5 minutes. The current adjustment circuit can integrate a voltage control module, for example, by adjusting the output voltage through a DAC module. During operation, the control unit starts a timer after starting, and sends an increment instruction to the electric stimulation module every 30 s, for example, gradually increasing from 12 mA to 20 mA, until reaching the preset upper limit of 30 mA, to avoid sudden increase of current causing muscle spasm.

[0129] The charge balance error of the biphasic pulse is less than ±5%, which is dynamically corrected by real-time monitoring of the integrated charge amount of the pulse to adjust the positive and negative pulse width ratio.

[0130] The charge balance detection can be based on an integration circuit, for example, the positive and negative pulse charge amounts are calculated by a current sampling resistor and an integrator. The dynamic correction algorithm can be embedded in the control unit firmware, for example, when the positive pulse charge amount is detected to exceed the negative pulse by 5%, the negative pulse width is automatically increased. The correction period can be set to update once every 10 pulses, for example, the positive pulse is adjusted from 0.4ms to 0.38ms, and the negative pulse is adjusted from 0.3ms to 0.32ms. During operation, the control unit continuously monitors the charge difference, adjusts the pulse width by proportion, and ensures that the charge accumulation error is controlled within ±5% under long-term stimulation.

[0131] Technical effect: The asymmetric biphasic pulse design reduces the risk of tissue polarization and improves long-term treatment safety. The dynamic frequency adjustment linked to the breathing frequency optimizes the diaphragmatic nerve activation efficiency and adapts to different breathing states. The current progressive increment mechanism reduces the risk of muscle spasm and ensures a smooth patient adaptation process. The real-time charge balance calibration reduces electrochemical damage and ensures treatment safety. The overall scheme improves the effect of electrical stimulation while significantly reducing the risk of side effects.

[0132] According to another embodiment of the present application, the duration of the locked state is controlled by a dynamic time window, with a maximum allowed duration of 1.2 to 1.5 times the current breathing cycle, and if no inspiration trigger signal is detected within the maximum allowed duration, the system switches to a safe standby mode.

[0133] The multiple of the dynamic time window can be 1.2, 1.35 or 1.5, for example, when the breathing cycle is 3s, the maximum allowed duration is 3.6s to 4.5s. The control unit can be an ARM Cortex-M4 microcontroller, the packaging material can be ABS plastic, and the internal timer module can set the dynamic time window. The control unit can be assembled in the central processing area of the device main control board and connected to the air pressure sensor through the bus. During operation, the control unit starts timing at the beginning of the expiration cycle, and if no valid inspiration signal is detected within the maximum allowed duration (e.g. negative pressure below -10Pa and slope ≥0.2Pa / ms), the system switches to a safe standby mode, stops electrical stimulation output and enters a low-power state.

[0134] The identification criteria for interference signals include noise components with a frequency higher than 5Hz and cough or swallowing actions with an amplitude less than 15% of the peak value of the current breathing cycle.

[0135] The frequency threshold can be set to 5 Hz or higher, for example, by analyzing the signal spectrum through FFT, and filtering out noise above 5 Hz. The amplitude threshold can be set to 15% of the peak value of the current respiratory cycle, for example, when the peak value is -20 Pa, the amplitude of the interference signal needs to be less than -3 Pa. The control unit can integrate a digital filter module, for example, a Butterworth low-pass filter, with a cutoff frequency of 5 Hz. The filter circuit can be installed in the signal preprocessing area of the main control board, close to the ADC conversion module. During operation, the control unit analyzes the signal frequency and amplitude in real time, and if high-frequency noise or low-amplitude transient signals (such as coughing) are detected, it is marked as interference and shielded.

[0136] The control unit delays 50 ms to 100 ms after the inspiration trigger signal arrives to perform a redundancy check, and after the check is passed, it unlocks and restores the electrical stimulation output.

[0137] The delay time can be selected as 50 ms, 75 ms or 100 ms, for example, set to 75 ms. The redundancy check conditions can include a second confirmation that the negative pressure value is below the threshold (such as -10 Pa) and the waveform slope meets the standard (such as ≥0.2 Pa / ms). The control unit can have a built-in timer and logic judgment module, for example, through GPIO control of the verification process. During operation, when the inspiration trigger signal is first detected, the control unit starts the delay timer, and after the expiration, the air pressure signal is reacquired. If the second detection still meets the trigger conditions, it is unlocked and the stimulation output is restored; otherwise, the lock state is maintained to avoid false triggering.

[0138] Technical effect: Dynamic time window control avoids long-term locking of the device due to signal loss, enhancing the system fault tolerance. Quantitative interference identification criteria reduce the risk of misjudgment, ensuring that only invalid signals are shielded during locking. The redundancy check mechanism improves the reliability of the trigger signal, preventing accidental noise from causing false operations. The overall scheme can still maintain high reliability in complex physiological environments, ensuring the safety and stability of the treatment process.

[0139] Embodiment:

[0140] 1. Air pressure sampling tube: A biocompatible medical silicone tube with an inner diameter of 2.0 mm is used, with an anti-slip pattern on the outer wall with a depth of 0.2 mm, fixed in the patient's nasal vestibule.

[0141] 2. Air pressure sensor: A MEMS micro sensor with a range of -50 Pa to 50 Pa is selected, with a detection accuracy of ±1 Pa and a sampling frequency of 150 Hz, installed in the front signal input interface of the main control board.

[0142] 3. Control unit: Based on ARM Cortex-M4 microcontroller, with built-in dynamic threshold algorithm, set the trigger threshold Ph to -10 Pa. When the negative pressure signal is detected to be lower than -10 Pa and the waveform slope is ≥0.3 Pa / ms, the electrical stimulation is triggered.

[0143] 4. Electric stimulation module: output biphasic pulse, positive pulse width 0.3 ms, negative pulse width 0.2 ms, frequency 20 Hz, initial current 10 mA. The pulse envelope duration dynamically matches the inspiration period, ranging from 0.8 s to 1.2 s.

[0144] 5. Hardware switch circuit: N-channel MOSFET is connected in series in the electric stimulation output circuit to physically cut off the current transmission path during the expiration period.

[0145] Working process:

[0146] When the patient inhales, the negative pressure in the nasal cavity is transmitted to the air pressure sensor through the sampling tube. The control unit analyzes the signal in real time, and when the negative pressure is lower than -10 Pa and the slope meets the standard, the electric stimulation module is triggered to output a biphasic pulse to stimulate the diaphragm to contract. After the expiration period begins, the MOSFET switch is turned off, completely blocking the stimulation output.

[0147] Comparative example:

[0148] 1. Air pressure sampling tube: a common silica gel tube with an inner diameter of 4.0 mm is used, without anti-skid design, which is easy to cause signal interruption due to displacement.

[0149] 2. Air pressure sensor: range -30 Pa to 30 Pa, accuracy ±2 Pa, sampling frequency 50 Hz, signal noise is large.

[0150] 3. Control unit: uses a fixed trigger threshold Ph=-15 Pa, without combining waveform slope determination, resulting in false triggering when coughing or short-term negative pressure fluctuation.

[0151] 4. Electric stimulation module: outputs monophasic pulse, width 0.5 ms, fixed frequency 30 Hz, current intensity fixed at 15 mA, unable to adapt to individual differences.

[0152] 5. Missing hardware switch circuit: only through software control to stop stimulation output, there is still a risk of pulse leakage during expiration period.

[0153] Problem analysis:

[0154] High false triggering rate: due to fixed threshold and no slope determination, false stimulation is easily caused by coughing or body position change.

[0155] Human-machine asynchronization: monophasic pulse and fixed parameters cannot match the changes in respiratory period, diaphragm contraction is out of sync with inspiration action.

[0156] Safety risk: software control is difficult to completely block the output during expiration period, which may cause abnormal contraction of diaphragm.

[0157] Effect comparison:

[0158] Embodiment: false trigger rate is reduced by dynamic threshold and slope determination; hardware switch completely blocks expiratory stimulation; dual-phase pulse adapts to breathing cycle, and patient weaning success rate is improved.

[0159] Comparative example: false trigger leads to treatment interruption; single-phase pulse causes charge accumulation, which may cause tissue damage in long-term use; expiratory stimulation leakage increases clinical risk.

[0160] Although embodiments of the present application have been disclosed in connection with the above description and the accompanying drawings, it will be apparent to those skilled in the art that numerous modifications can be made without departing from the general principles set forth herein. Accordingly, the present application is not to be limited to the specific details and examples described above but is to be given the full scope embodied in the appended claims and equivalents thereof.

Claims

1. A breath-synchronized extrathoracic phrenic nerve electrical stimulator, comprising: The method comprises the following steps: an air pressure sampling tube is placed in the patient's nasal cavity to detect the negative pressure signal caused by inspiration; an air pressure sensor is connected to the air pressure sampling tube to convert the negative pressure signal into an electrical signal; an electrical stimulation module outputs an electrical stimulation pulse to the phrenic nerve within the envelope duration of the electrical stimulation pulse, and the envelope duration of the electrical stimulation pulse is less than or equal to the inspiration duration of natural respiration or automatically matches the inspiration cycle of natural respiration; the pulse waveform output by the electrical stimulation module includes a positive-negative biphasic pulse, the pulse width is 0.1ms to 1ms, the frequency is 10Hz to 50Hz, and the current intensity is 5mA to 30mA; the electrical stimulation pulse is prohibited from being output during the expiration cycle; a control unit receives the air pressure sensor signal, and the control unit analyzes the negative pressure signal in real time through a dynamic window algorithm, the dynamic window algorithm comprising the following steps: calculate the mean and standard deviation of the negative pressure in the last 3 consecutive breathing cycles, dynamically set the trigger threshold Ph as the mean minus 1.5 times the standard deviation, and the threshold range is-8Pa to-20Pa; detect the interval in the negative pressure signal waveform where the waveform slope absolute value is greater than or equal to 0.25Pa / ms to determine the effective inspiration trigger window; only in the effective inspiration trigger window and when the negative pressure value is lower than the trigger threshold Ph, the electrical stimulation is triggered, and the remaining period is shielded from interference signals; an envelope duration of the electrical stimulation pulses is adapted to a duration of a real-time inhalation window t 窗口 dynamically adjusted, the duration of the inhalation window t 窗口 for the time interval from the start of the negative pressure signal when the absolute value of the waveform slope first reaches or exceeds 0.25 Pa / ms to the time when the negative pressure signal falls back to less than 10% of its maximum negative pressure value, the adjustment method is: Based on the duration of the current inhalation window t 窗口 , the envelope length of the next cycle is calculated by a proportional-integral control algorithm t 包络 , t 包络 = t 窗口 ×(1±5%) generate a PWM signal through a timer module to control the pulse duration output by the electrical stimulation module, and ensure that the adjustment error is less than ±5%.

2. The respiratory synchronized extracorporeal phrenic nerve electrical stimulator according to claim 1, wherein: the control unit determines the inspiration starting point by detecting the negative pressure signal drop rate, and triggers the electrical stimulation when the negative pressure value is lower than the trigger threshold Ph and the negative pressure change rate exceeds 0.5Pa / ms.

3. The respiratory synchronized extracorporeal phrenic nerve electrical stimulator according to claim 1, wherein: the air pressure sampling tube is made of silica gel, with an inner diameter of 1mm to 3mm and an outer wall surface covered with a non-slip coating for stable fixation in the patient's nasal cavity; the control unit monitors the waveform slope of the negative pressure signal in real time, and determines the effective inspiration trigger signal when the negative pressure value is lower than the trigger threshold Ph and the waveform slope absolute value exceeds 0.3Pa / ms, triggering the electrical stimulation.

4. The respiratory synchronized extracorporeal phrenic nerve electrical stimulator according to claim 1, wherein: the air pressure sensor has a range of-50Pa to 50Pa, a detection accuracy of ±0.5Pa, and a sampling frequency of 100Hz to 200Hz; the electrical signal output by the air pressure sensor is processed through a baseline drift correction algorithm, which real-time deducts the influence of environmental air pressure fluctuations, and the correction period is 1s to 3s; the control unit extracts the main frequency component 0.1Hz to 2Hz of the inspiration negative pressure signal through frequency spectrum analysis and filters out high frequency noise greater than or equal to 10Hz.

5. The respiratory synchronized extracorporeal phrenic nerve electrical stimulator according to claim 1, wherein: the pulse waveform output by the electrical stimulation module is a biphasic square wave, including positive-negative alternating pulses, the positive pulse width is 0.2ms to 0.5ms, the negative pulse width is 0.1ms to 0.3ms, and the pulse interval is 1ms to 5ms. The current intensity of the electric stimulation module is dynamically adjusted according to the skin impedance of the patient, and the adjustment method is: a preset reference voltage of 10V to 50V is divided by the real-time detected skin impedance value of 500Ω to 3000Ω, and then multiplied by a safety factor of 0.8 to 1.2, and the output current range is 5mA to 30mA; The electric stimulation module is built-in with an over-current protection circuit, which automatically cuts off the output when the real-time current exceeds 120% of the set value, and restores the detection within 100ms.

6. The breath-synchronized extracorporeal phrenic nerve electric stimulator of claim 1, wherein: The envelope duration of the electric stimulation pulse is dynamically adjusted by real-time detection of the inspiratory period of natural breathing, and the adjustment method is: The control unit extracts the inspiratory duration of the last 3 consecutive breath cycles based on the air pressure sensor signal, calculates the average inspiratory duration as a reference value; The envelope duration of the current breath cycle is set to 95% to 105% of the reference value, and does not exceed ±5% of the real-time inspiratory duration; Dynamic adjustment is achieved through a sliding window average algorithm, with a window length of 5 to 10 breath cycles, and a weight coefficient that decays with the breath sequence; When a sudden change in the patient's breathing pattern is detected, and the inspiratory duration changes by more than 20% of the reference value, the control unit suspends dynamic adjustment and switches to a fixed envelope duration mode, with a fixed value set to 80% to 120% of the historical average value.

7. The breath-synchronized extracorporeal phrenic nerve electric stimulator of claim 1, wherein: The control unit prohibits electric stimulation output during the expiratory period through a double-checking mechanism, specifically including: Inspiratory termination determination: when the negative pressure signal is detected to drop below 10% of the peak value, the inspiratory period is determined to be over; Expiratory signal confirmation: within a time window of 50ms to 200ms after the inspiratory termination, if nasal positive pressure is detected, the electric stimulation module output is locked; The output circuit of the electric stimulation module is connected in series with a hardware switch circuit, which is controlled by the control unit and physically cuts off the stimulation pulse transmission path during the expiratory period; The lock state lasts until the next inspiratory cycle trigger signal arrives, and during this period the control unit ignores all interference signals including noise, cough and / or swallowing action.

8. The breath-synchronized extracorporeal phrenic nerve electric stimulator of claim 1, wherein: In the positive and negative biphasic pulse waveform output by the electric stimulation module, the positive pulse width is 0.2ms to 0.6ms, and the negative pulse width is 0.1ms to 0.4ms, and the positive pulse width is always 1.2 to 2.0 times the negative pulse width; The pulse frequency is dynamically adjusted according to the real-time breathing frequency of the patient, and the adjustment method is: if the breathing frequency ≤12 times / minute, the stimulation frequency is 20Hz to 30Hz; if the breathing frequency >12 times / minute, the stimulation frequency is 30Hz to 40Hz; The initial value of the current intensity is 10mA to 15mA, and within the first 5 minutes after the start of stimulation, it is gradually increased to a target intensity ≤30mA at a rate of 1mA to 2mA every 30s, in order to avoid sudden increase of current causing muscle spasm; The charge balance error of the biphasic pulse is less than ±5%, which is dynamically corrected by real-time monitoring of the integral charge amount of the pulse to adjust the positive and negative pulse width ratio.

9. The breath-synchronized extracorporeal diaphragm electric neuro-stimulator of claim 7, wherein: the duration of the lock state is controlled by a dynamic time window, specifically: the control unit continuously masks all non-inhalation signals from the start of the expiratory phase until the time window of the arrival of the next inhalation trigger signal; the maximum allowed duration of the time window is 1.2 to 1.5 times the current respiratory cycle, if the inhalation trigger signal is not detected within the time window, the system switches to a safe standby mode and stops all electric stimulation output; the identification criteria of the interference signal include: a. noise components with a signal frequency higher than 5 Hz; b. cough or swallowing action with a signal amplitude less than 15% of the peak value of the current respiratory cycle; the control unit performs a redundant check 50 to 100 ms after the arrival of the inhalation trigger signal, if the check is passed, the lock state is released and the electric stimulation output is resumed.

Images