Respiratory support system and method with in-vitro negative pressure cooperating with electrical stimulation

Abstract

The invention discloses a respiratory support system and method with in-vitro negative pressure cooperating with electrical stimulation. The method comprises the steps that a signal detection module collects body surface respiratory muscle electromyography, thoracic posture, thoracico-abdominal deformation and respiratory airflow signals of a patient in real time; the negative pressure ventilation module generates negative pressure outside the thorax of the patient through a thoracic nail, a vacuum pump, a pressure reducing valve and a two-position three-way valve to reduce the pressure of a pleural cavity and pulmonary alveoli; the constant-current electrical stimulation module applies constant-current electrical stimulation to the diaphragm to enhance the contractility of the diaphragm; the control module receives the respiratory physiological signal, judges the respiratory intention of a patient through a built-in respiratory intention recognition algorithm, and controls the negative pressure ventilation and electrical stimulation module to realize triggering of an inspiration phase and an expiration phase and man-machine synchronization; through the synergistic effect of negative pressure ventilation and electrical stimulation, the respiratory support effect and man-machine synchronism are improved, and the device is suitable for adjuvant therapy of respiratory dysfunction patients.

Description

Technical Field

[0001] This invention relates to the field of medical system technology, specifically to a respiratory support system and method using external negative pressure combined with electrical stimulation. Background Technology

[0002] Respiratory support equipment is an important medical device used to help patients with respiratory diseases breathe. A ventilator is a widely used respiratory support device in clinical practice. It delivers air to the lungs by increasing airway pressure to support or replace the patient's respiratory function. Ventilators can be broadly classified into two categories based on their ventilation method: positive pressure ventilation (PPV) ventilators and negative pressure ventilation (NPV) ventilators. While PPV ventilators primarily use increased airway pressure to force air into the lungs, the mechanical force applied to the lungs can damage sensitive lung tissue, such as causing ventilator-induced lung injury. This type of injury can lead to a series of complications, such as tension pneumothorax, pulmonary edema, and oxygen toxicity, thereby worsening the condition and endangering the patient's life.

[0003] Negative pressure ventilation (NPV) machines simulate physiological spontaneous breathing and are an alternative to traditional positive pressure ventilation. They work by creating negative pressure around the chest, reducing pressure in the pleural cavity and alveoli, and promoting airflow into the patient's lungs. However, existing NPV machines have some shortcomings in practical use: because the gas flow rate cannot precisely match the patient's ventilation needs, insufficient ventilation or patient-ventilator asynchrony often occurs. This not only prolongs the patient's treatment time but also reduces the efficiency and effectiveness of treatment. Summary of the Invention

[0004] Purpose of the invention: The purpose of this invention is to provide a respiratory support system and method with external negative pressure combined with electrical stimulation to solve the problems of insufficient ventilation, patient-ventilator asynchrony and lung injury in existing respiratory support systems.

[0005] Technical solution: According to one aspect of the present invention, an external negative pressure combined with electrical stimulation respiratory support system is provided, comprising: The signal detection module is used to collect the patient's respiratory physiological signals in real time, including at least the electromyography of the respiratory muscles on the body surface, chest posture, chest and abdominal deformation, and respiratory airflow signals. The negative pressure ventilation module is used to generate negative pressure outside the patient's chest cavity to reduce the pressure in the pleural cavity and alveoli, and promote the inflow of air into the lungs; The constant current electrical stimulation module is used to apply constant current electrical stimulation to the phrenic nerve to promote diaphragmatic contraction and enhance diaphragmatic contractile force. The control module is used to receive respiratory physiological signals collected by the signal detection module, and to determine the patient's breathing intention based on the built-in breathing intention recognition algorithm, and control the operation of the negative pressure ventilation module and the constant current electrical stimulation module to achieve the triggering of the inspiratory and expiratory phases and human-machine synchronization.

[0006] Furthermore, the negative pressure ventilation module consists of a chest armor, a pressure sensor, a two-position three-way valve, a vacuum pump, and a pressure reducing valve.

[0007] Furthermore, the pressure reducing valve is used to adjust the pressure generated by the vacuum pump, and the two-position three-way valve is used to switch and regulate the pressure inside the chest armor.

[0008] Furthermore, the constant current electrical stimulation module includes a constant current drive circuit, an isolation and protection circuit, and a power amplifier circuit.

[0009] Furthermore, the controller module includes a controller, an A / D converter, and a D / A converter.

[0010] According to another aspect of the present invention, a respiratory support method using external negative pressure combined with electrical stimulation is provided, comprising: The signal detection module collects the patient's respiratory physiological signals in real time and sends them to the control module. The control module preprocesses the received signals through a built-in respiratory intention recognition algorithm and reconstructs the respiratory timing waveform to determine the patient's respiratory intention. Based on the respiratory intention, the control timing of the negative pressure ventilation module and the constant current electrical stimulation module is synchronized with the patient's inspiratory and expiratory phases. in, If the intention to inhale is detected, the controller module sends a trigger signal to make the negative pressure ventilation module and the constant current electrical stimulation module work simultaneously. If the system detects an intention to exhale, the controller module shuts off the vacuum pump, switches the two-position three-way valve to the right position, and stops the electrical stimulation.

[0011] Furthermore, if the intention to inhale is determined, the vacuum pump is activated, the pressure reducing valve is adjusted to the preset pressure, and the two-position three-way valve is switched to the left position, forming a negative pressure cavity between the thoracic armor and the surface of the chest and abdomen, generating a local vacuum to reduce the pressure in the pleural cavity and alveoli, assisting lung expansion, and allowing external air to flow into the lungs. At the same time, the constant current electrical stimulation module sends an electrical stimulation signal to the phrenic nerve, causing the diaphragm to contract and increase the negative pressure in the pleural cavity.

[0012] Furthermore, the preprocessing includes: splicing the time-frequency domain information of surface respiratory muscle electromyography, chest posture, chest and abdominal deformation, and respiratory airflow signals from different sources to construct a high-dimensional feature matrix, and using principal component analysis to reduce the dimensionality and fuse the spliced ​​high-dimensional feature matrix.

[0013] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: 1. During the inspiratory phase, the negative pressure ventilation module, in conjunction with the constant current electrical stimulation module, stimulates the diaphragm to increase the diaphragm contractile force, further improving the respiratory ventilation, so that the ventilation provided by the ventilator matches the ventilation demand; 2. By reconstructing the respiratory sequence and extracting respiratory parameters through the respiratory intention recognition method of multi-source physiological information fusion, the triggering of the two respiratory phases and the synchronization between the patient and the ventilator are realized, reducing patient-ventilator asynchrony and improving the therapeutic efficacy and efficiency; Compared with the existing positive pressure ventilation ventilator, it avoids mechanical damage to lung tissue and reduces ventilator-induced lung injury and related complications; Compared with the existing negative pressure ventilator, it solves the problem of insufficient ventilation and improves patient-ventilator coordination. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of an external negative pressure combined with electrical stimulation respiratory support system architecture provided in an embodiment of the present invention;

[0015] Figure 2 A schematic diagram illustrating the respiratory support control principle of the negative pressure ventilation module and the constant current electrical stimulation module working together as provided in an embodiment of the present invention;

[0016] Figure 3 A schematic diagram of reconstructed respiratory timing and a schematic diagram of parameter extraction provided for embodiments of the present invention;

[0017] Figure 4 This is a schematic diagram of the triggering rules for a respiratory support system provided in an embodiment of the present invention;

[0018] Figure 5 This is a schematic diagram of a respiratory support method using external negative pressure combined with electrical stimulation, provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0020] Traditional positive pressure ventilation machines apply mechanical force to the lungs, which can damage sensitive lung tissue and often lead to ventilator-induced lung injury. Although negative pressure ventilation machines simulate physiological spontaneous breathing, the gas they produce cannot meet the patient's actual ventilation needs, resulting in insufficient ventilation and patient-ventilator asynchrony, which prolongs the patient's treatment time and reduces efficacy.

[0021] In response, this application creatively proposes an extracorporeal negative pressure combined with electrical stimulation respiratory support system. This system uses a signal detection module to collect the patient's respiratory physiological signals in real time, a negative pressure ventilation module to generate negative pressure outside the chest cavity, a constant current electrical stimulation module to apply constant current electrical stimulation to the diaphragm, and a control module to determine the patient's breathing intention based on a built-in respiratory intention recognition algorithm and control the operation of the negative pressure ventilation module and the constant current electrical stimulation module to achieve triggering and synchronization of the inspiratory and expiratory phases with the patient.

[0022] For ease of understanding, the following explains some key terms in this embodiment: Respiratory physiological signals encompass multidimensional physiological indicators of a patient's respiratory activity, such as surface respiratory muscle electromyography, chest posture, chest and abdominal deformation, and respiratory airflow signals. These signals are used to comprehensively assess a patient's respiratory status.

[0023] The diaphragm is the main respiratory muscle in the human body, located between the thoracic and abdominal cavities. Its contraction and relaxation directly affect the ventilation of the lungs.

[0024] The inhalation and exhalation phases constitute a complete respiratory cycle in the human body. The inhalation phase refers to the process of air entering the lungs, and the exhalation phase refers to the process of air being expelled from the lungs.

[0025] Example 1 Please refer to Figure 1 This embodiment provides a respiratory support system with external negative pressure combined with electrical stimulation, the specific structure and working principle of which are as follows: The signal detection module is used to collect the patient's respiratory physiological signals in real time, including at least the electromyography of the respiratory muscles on the body surface (electrical myographism of the diaphragm and the electromyography of the intercostal muscles on the body surface), chest posture, chest and abdominal deformation, and respiratory airflow signals.

[0026] Specifically, the signal detection module includes multiple sensors capable of simultaneously acquiring surface respiratory muscle electromyography (EMG) signals, chest posture signals, chest and abdominal deformation signals, and respiratory airflow signals. Surface respiratory EMG signals are acquired using Ag / AgCl (silver chloride) surface electrodes attached to the patient's skin. EMG signals are very weak bioelectrical signals, and effective feature extraction requires multi-stage amplification and filtering circuits. The surface EMG signals are guided to a front-end amplifier via surface electrodes. After primary amplification, the main amplifier filters out unwanted interference and signals using a high-pass filter (cutoff frequency 20Hz) and a low-pass filter (cutoff frequency 1000Hz). After secondary amplification, the signals are further filtered by a power frequency notch filter (50Hz) to remove power frequency interference. Flexible strain gauges are used to acquire chest and abdominal deformation signals, and a Wheatstone bridge is used to measure the strain gauges; the differential voltage generated by this circuit is proportional to the resistance. An IMU sensor, placed below the sternum, is used to measure chest posture signals. Human respiratory airflow was collected using a bidirectional flow sensor; a Butterworth high-pass filter was used to filter out low-frequency artifact noise (0~20Hz) caused by sensor movement to improve the signal-to-noise ratio; different signals were normalized and the dataset was divided into training and testing sets to complete the preprocessing of initial electromyography signals, pose signals, chest and abdominal deformation, and respiratory airflow signals.

[0027] The negative pressure ventilation module is used to generate negative pressure outside the patient's chest cavity to reduce the pressure in the pleural cavity and alveoli, thereby promoting the flow of air into the lungs.

[0028] Specifically, the negative pressure ventilation module consists of a chest armor, a pressure sensor, a two-position three-way valve, a vacuum pump, and a pressure reducing valve.

[0029] Please refer to Figure 2During the inspiratory phase, the control module sends a trigger signal to control the simultaneous operation of the negative pressure ventilation module and the constant current electrical stimulation module. At this time, the vacuum pump is turned on, and the pressure is adjusted via the pressure reducing valve. The two-position three-way valve is switched to the left position. A negative pressure cavity is formed between the thoracic cavity and the thoracic and abdominal surfaces. This local vacuum reduces the pressure in the pleural cavity and alveoli. Air flows into the lungs through the mouth, nose, and airways under atmospheric pressure. Simultaneously, an electrical stimulation signal is sent to the phrenic nerve to promote diaphragmatic contraction, further increasing the negative pressure in the pleural cavity and promoting air inflow into the lungs. Therefore, under the combined action of the negative pressure cavity in the thoracic cavity and the electrical stimulation of the phrenic nerve, the amount of air flowing into the lungs increases, thus increasing respiratory ventilation. During the expiratory phase, the control module shuts off the vacuum pump, the two-position three-way valve switches to the right position, and electrical stimulation stops. At this time, the thoracic cavity is connected to the atmosphere, and the intrapulmonary pressure is higher than the external atmospheric pressure, forming an outward pressure gradient. Under the influence of diaphragmatic relaxation and the reverse traction of the pleural cavity and lung tissue, the gas in the lungs is expelled through the expiratory tract, completing the exhalation process. When the respiratory support system is working, the control timing of the vacuum pump, 2-position 3-way valve, and electrical stimulation module remains consistent during both inhalation and exhalation. During inspiration, the control module provides a high-level signal to the vacuum pump, 2-position 3-way valve, and electrical stimulation module; during exhalation, the control module sends a low-level signal to the vacuum pump, 2-position 3-way valve, and electrical stimulation module.

[0030] For example, the circumferential dimensions of the thoracic armor are determined using a 3D reconstruction model of the human chest and abdomen from a CT scan. A negative pressure thoracic armor made of transparent acrylic is then fabricated using a 3D printing rapid prototyping process. Medical-grade silicone is used to fix the edges of the thoracic armor to ensure wearing comfort and cavity sealing. Traditional negative pressure ventilation devices may use a standard-sized thoracic armor, leading to discomfort or poor sealing. By customizing the thoracic armor size based on a 3D reconstruction model of the human chest and abdomen from a CT scan, a perfect fit between the thoracic armor and the patient's thoracic cavity can be ensured, minimizing air leakage and improving the efficiency of negative pressure ventilation. 3D printing technology can quickly and accurately manufacture customized thoracic armors with complex shapes. Transparent acrylic material is lightweight, durable, and easy to observe. The medical-grade silicone edge fixation further enhances the seal while improving wearing comfort and preventing skin damage. These design and manufacturing details together ensure the effective formation and maintenance of the negative pressure cavity, improving treatment outcomes and patient compliance.

[0031] The constant current electrical stimulation module is used to apply constant current electrical stimulation to the phrenic nerve to promote diaphragmatic contraction and enhance diaphragmatic contractile force.

[0032] Specifically, the constant current electrical stimulation module includes a constant current drive circuit, an isolation and protection circuit, and a power amplifier circuit. The constant current drive circuit outputs a stable, constant current, unaffected by changes in load impedance. The isolation and protection circuit protects patient safety and includes overcurrent and overvoltage protection functions. The power amplifier circuit amplifies the control signal to a sufficient power level to drive the electrodes to apply electrical stimulation to the diaphragm. The electrical stimulation electrodes are attached to the corresponding body surface location on the diaphragm, transmitting the electrical stimulation signal through the skin to the diaphragm. This module applies constant current electrical stimulation to the diaphragm, stimulating the diaphragmatic nerve, enhancing diaphragmatic contractility, and improving respiratory efficiency.

[0033] The control module is used to receive respiratory physiological signals collected by the signal detection module, and to determine the patient's breathing intention based on the built-in breathing intention recognition algorithm, and control the operation of the negative pressure ventilation module and the constant current electrical stimulation module to achieve the triggering of the inspiratory and expiratory phases and human-machine synchronization.

[0034] Specifically, the control module includes a controller, an A / D converter, and a D / A converter. The controller employs a high-performance microprocessor with a built-in respiratory intention recognition algorithm, capable of processing multiple physiological signals in real time. The A / D converter converts the analog signals acquired by the signal detection module into digital signals for the controller to process and analyze. The D / A converter converts the digital control signals output by the controller into analog signals to drive the negative pressure ventilation module and the constant current electrical stimulation module. The controller receives the respiratory physiological signals acquired by the signal detection module and analyzes the patient's breathing pattern and respiratory intention through the built-in respiratory intention recognition algorithm, identifying the timing of the transition between the inspiratory and expiratory phases.

[0035] The system provided in this invention achieves precise assistance and human-machine synchronization in the patient's breathing process by collecting multi-source respiratory physiological signals, the synergistic effect of external negative pressure ventilation and diaphragmatic constant current electrical stimulation, and combining it with an intelligent respiratory intention recognition algorithm. Therefore, this system can effectively solve problems such as insufficient ventilation, patient-machine asynchrony, and potential lung injury in existing respiratory support protocols, improving the physiological compatibility and treatment efficiency of respiratory support.

[0036] In some embodiments described above in this application, a negative pressure ventilation module is proposed to generate negative pressure outside the thoracic cavity to reduce the pressure in the pleural cavity and alveoli and promote air inflow into the lungs. Furthermore, this application proposes that the negative pressure ventilation module comprises a chest armor, a pressure sensor, a two-position three-way valve, a vacuum pump, and a pressure reducing valve.

[0037] The thoracic armor is a device that fits snugly against the outside of the human thoracic cavity, creating a relatively closed negative pressure chamber. Pressure sensor P 腔This valve is used to monitor pressure changes within the negative pressure chamber in real time and convert the pressure signal into an electrical signal output. A two-position three-way valve is a valve with two working positions and three ports used to control the flow direction of fluid; here, it is used to switch the connection status of the negative pressure chamber with the vacuum pump or the atmosphere. A vacuum pump is a device used to generate negative pressure by evacuating air from the negative pressure chamber to reduce the pressure inside. A pressure reducing valve is used to adjust the magnitude of the negative pressure generated by the vacuum pump to ensure that the pressure within the negative pressure chamber is maintained within a preset safe and effective range.

[0038] In some embodiments described above in this application, a negative pressure ventilation module is proposed to generate negative pressure outside the thoracic cavity to reduce the pressure in the pleural cavity and alveoli, and to promote air inflow into the lungs. This application further clarifies the specific structure and function of the aforementioned negative pressure ventilation module, wherein a pressure reducing valve is used to regulate the pressure generated by the vacuum pump, and a two-position three-way valve is used to switch and regulate the pressure within the thoracic cavity.

[0039] By clearly defining the functions of the pressure-reducing valve and the two-position three-way valve, the negative pressure ventilation module can precisely control the negative pressure value inside the thoracic cavity and achieve rapid and accurate pressure switching between the inspiratory and expiratory phases. The pressure-reducing valve's ability to regulate the vacuum pump pressure allows the system to provide appropriate negative pressure according to the patient's actual needs, avoiding discomfort or insufficient ventilation caused by excessively high or low negative pressure. The switching and control capabilities of the two-position three-way valve ensure smooth respiratory cycles, avoid patient-ventilator asynchrony, and improve ventilation efficiency and patient comfort.

[0040] In some of the solutions described above in this application, a constant current electrical stimulation module is proposed to apply constant current electrical stimulation to the diaphragm to enhance the diaphragm's contractile force. In this regard, this application further proposes that the constant current electrical stimulation module includes a constant current drive circuit, an isolation and protection circuit, and a power amplifier circuit.

[0041] In some of the embodiments described above in this application, a controller module is proposed to receive respiratory physiological signals and control the negative pressure ventilation module and the constant current electrical stimulation module according to the breathing intention. In this regard, this application further proposes that the controller module includes a controller, an A / D converter and a D / A converter.

[0042] Example 2 Existing respiratory support systems often face problems of insufficient ventilation and patient-ventilator asynchrony during ventilation, leading to reduced treatment efficiency and prolonged patient recovery time. To address this, this application provides a respiratory support method using external negative pressure combined with electrical stimulation. Please refer to [link to relevant documentation]. Figure 5 Specifically, it includes the following steps: S1. The signal detection module collects the patient's respiratory physiological signals in real time and sends them to the control module.

[0043] S2. The control module uses a built-in breathing intention recognition algorithm to preprocess the received signal and reconstruct the breathing timing waveform to accurately determine the patient's breathing intention. Based on the determination result, the control module dynamically adjusts the working timing of the negative pressure ventilation module and the constant current electrical stimulation module to ensure synchronization with the patient's inspiratory and expiratory phases.

[0044] S3. When the intention to inhale is detected, the control module sends a trigger signal to make the negative pressure ventilation module and the constant current electrical stimulation module work simultaneously; when the intention to exhale is detected, the control module controls the vacuum pump to shut down, the two-position three-way valve to switch to the right position, and stops the electrical stimulation.

[0045] Specifically, when the system determines that the patient intends to inhale, the control module immediately sends a trigger signal, causing the negative pressure ventilation module and the constant current electrical stimulation module to start working simultaneously. Specifically, the vacuum pump begins to pump air, the pressure reducing valve automatically adjusts to the preset negative pressure value, and the two-position three-way valve quickly switches to the left position, forming a closed negative pressure cavity between the thoracic cavity and the thoracic and abdominal surfaces. This creates a local vacuum environment within the negative pressure cavity, effectively reducing the pressure in the pleural cavity and alveoli, providing external driving force for lung expansion, and facilitating the smooth inflow of external air into the lungs. Simultaneously, the constant current electrical stimulation module precisely locates the phrenic nerve and sends electrical stimulation signals of a specific frequency and intensity to that area. The electrical stimulation signals act directly on the phrenic nerve, causing active contraction of the diaphragm, further increasing the negative pressure within the pleural cavity, creating a synergistic effect with the external negative pressure, significantly improving inspiratory efficiency.

[0046] When the system determines that the patient intends to exhale, the control module immediately adjusts the operating status of each module. The controller module issues a stop command, the vacuum pump shuts down and stops pumping air, and the two-position three-way valve switches to the right position, connecting the negative pressure chamber to the outside atmosphere and eliminating the external negative pressure environment. At the same time, the constant current electrical stimulation module stops sending electrical stimulation signals to the phrenic nerve, the diaphragm relaxes naturally, the thoracic cavity pressure returns to normal, and the lungs complete the exhalation process under their own elastic recoil force.

[0047] For example, when the controller in the control module accurately identifies the patient's breathing intention using a built-in breathing intention recognition algorithm (which can be understood as a deep learning algorithm), the controller immediately outputs a digital signal "1" when an inspiratory intention is detected. This "1" signal is simultaneously sent to the negative pressure ventilation module and the constant current electrical stimulation module, instructing them to start working immediately, such as the vacuum pump starting to pump air and the electrical stimulation starting to deliver pulses. When an expiratory intention is detected, the controller outputs a digital signal "2," instructing both modules to immediately stop working. Between inspiratory and expiratory intentions, if there is no clear triggering event, the controller outputs a digital signal "0," maintaining the current state to ensure stable system operation and avoid false triggering.

[0048] The technical solution provided in this embodiment involves a signal detection module continuously monitoring changes in the patient's respiratory status throughout the respiratory support process. The control module dynamically adjusts the negative pressure intensity and electrical stimulation parameters based on real-time feedback, ensuring that the support effect remains synchronized with the patient's actual needs. Through continuous learning and optimization, the system gradually improves the accuracy and response speed of respiratory intention recognition.

[0049] The core innovation of this embodiment lies in the precise capture of the patient's breathing intention by combining real-time detection of multi-source respiratory physiological signals with a breathing intention recognition algorithm through dynamic feedback. Because the breathing intention recognition algorithm can reconstruct the respiratory timing waveform based on preprocessed signals, it accurately identifies the inspiratory and expiratory initiations. Therefore, the control module can promptly trigger or stop relevant modules, avoiding insufficient ventilation and patient-machine asynchrony caused by timing discrepancies in traditional systems. Furthermore, the synergistic operation of the negative pressure ventilation module and the constant current electrical stimulation module enhances the negative pressure in the thoracic cavity and the diaphragmatic contraction force during inspiration, significantly increasing ventilation; while during expiration, the negative pressure is completely released and stimulation stops, allowing the patient to exhale naturally.

[0050] In some of the solutions described above in this application, it is proposed that the negative pressure ventilation module and the constant current electrical stimulation module work simultaneously to increase ventilation volume when the intention to inhale is detected. In this regard, this application further proposes that when the intention to inhale is detected, the vacuum pump is activated, the pressure reducing valve is adjusted to a preset pressure, and the two-position three-way valve is switched to the left position, forming a negative pressure cavity between the thoracic turbinate and the thoracic and abdominal surfaces, generating a local vacuum to reduce the pressure in the pleural cavity and alveoli, assisting lung expansion, allowing external air to flow into the lungs. Simultaneously, the constant current electrical stimulation module sends an electrical stimulation signal to the phrenic nerve, causing the diaphragm to contract and increasing the negative pressure in the pleural cavity.

[0051] In some of the solutions described above in this application, preprocessing is proposed to process multi-source respiratory physiological signals in order to identify respiratory intention. In this regard, this application further proposes that the above preprocessing includes: splicing the time-frequency domain information of surface respiratory muscle electromyography, chest posture, chest and abdominal deformation and respiratory airflow signals from different sources to construct a high-dimensional feature matrix, and using principal component analysis to reduce the dimensionality and fuse the spliced ​​high-dimensional feature matrix.

[0052] Specifically, for electromyography (EMG) signals, the absolute mean (MAV), variance (VAR), root mean square (RMS), envelope, and fixed sample entropy of the intercostal and diaphragmatic EMG signals are extracted using a moving window. Frequency domain information includes median frequency (MF) and average power frequency (MPF). The root mean square (RMS) and variance (VAR) are used to extract time-domain features of pose and deformation signals. Frequency domain information extraction methods include Fourier transform and short-time Fourier transform. The time-frequency domain information from different sources (EMG, pose, and deformation) is concatenated to construct a high-dimensional feature matrix. Principal component analysis (PCA) is then used to reduce the dimensionality and fuse the concatenated high-dimensional feature matrix. These features retain the main information of the original data while reducing redundancy and noise.

[0053] Through the above technical solution, this application effectively solves the problems of high-dimensional redundancy and noise interference in multi-source respiratory physiological signal processing. By comprehensively stitching together the time-frequency domain information of surface respiratory muscle electromyography, chest posture, chest and abdominal deformation, and respiratory airflow signals, complete capture of respiratory-related information is ensured, avoiding feature omissions or local interference that may be caused by a single signal source. Subsequently, a high-dimensional feature matrix is ​​constructed to provide a unified structure for data processing. Furthermore, principal component analysis is used to reduce and fuse the high-dimensional feature matrix, which not only significantly reduces the feature dimension and computational complexity, but also effectively filters out noise by extracting the main components in the data, making the feature set more refined and representative. This provides high-quality, low-redundancy input for subsequent respiratory intention recognition algorithms, thereby significantly improving the accuracy and real-time performance of respiratory intention recognition. Finally, this optimized preprocessing method ensures that the control module can more accurately judge the patient's respiratory intention, thereby achieving precise synchronization between the negative pressure ventilation module and the constant current electrical stimulation module and the patient's inspiratory and expiratory phases, greatly reducing human-machine aggression and improving the therapeutic effect and patient comfort of the respiratory support system.

[0054] In some of the solutions described above in this application, a deep learning-based approach is proposed for predicting respiratory timing, using a Long Short-Term Memory (LSTM) recurrent neural network model to reconstruct the respiratory timing waveform. Please refer to [reference needed]. Figure 3 Breathing pattern parameters (respiratory rate, inspiratory time, expiratory time, and respiratory start and end points) are extracted using methods such as threshold detection and peak extraction.

[0055] Specifically, the LSTM model design process is as follows: Input the time series of multi-source fused features; then use one or more LSTM units to capture temporal dependencies; add a fully connected layer after the LSTM layer to map the hidden states to the dimension of the respiratory time series waveform; and output the reconstructed respiratory time series waveform. During model training, mean squared error (MSE) is used as the loss function to measure the difference between the reconstructed waveform and the real respiratory airflow waveform, and the model parameters are optimized using gradient descent.

[0056] Among them, the LSTM model excels at processing time-series data, capturing complex long-term dependencies in respiratory physiological signals, thus achieving accurate reconstruction of respiratory time-series waveforms. By combining the reconstructed waveforms with methods such as threshold detection and peak extraction, key respiratory pattern parameters such as the onset of inspiration, the onset of expiration, respiratory rate, inspiratory time, and expiratory time can be accurately identified. This deep learning-based prediction and parameter extraction method overcomes the dependence on signal feature engineering and sensitivity to individual differences inherent in traditional methods, significantly improving the accuracy and robustness of respiratory intention recognition. It ensures that the system can precisely synchronize with the patient's spontaneous breathing rhythm, thereby reducing human-machine aggression and improving treatment outcomes. Using MSE as the loss function and gradient descent to optimize model parameters ensures the effectiveness of model training and the accuracy of the reconstructed waveforms.

[0057] Please refer to Figure 4 The controller establishes operational triggering rules for the negative pressure ventilation module and the electrical stimulation module, using digital encoding based on respiratory parameters (inspiratory initiation and expiratory initiation). An inspiratory initiation of "1" indicates that the respiratory support system sends signals to both the negative pressure ventilation module and the constant current electrical stimulation module; an expiratory initiation of "2" indicates that the respiratory support system stops sending digital signals to these modules; and "0" represents no response, meaning the controller does not change its current operating state. In summary, the respiratory support system controls the negative pressure ventilation module and the electrical stimulation module based on multi-source respiratory physiological signals to achieve human-machine synchronization.

[0058] By translating inspiratory and expiratory intentions into explicit digital codes and defining 0 as a non-responsive state, the controller enables precise, real-time triggering and stopping of the negative pressure ventilation and electrical stimulation modules. This clear, discrete control logic avoids fuzzy judgments and delays, ensuring the immediacy and accuracy of the system's response. It allows negative pressure ventilation and electrical stimulation to be highly synchronized with the patient's spontaneous breathing intentions, significantly reducing patient-ventilator asynchrony and improving the comfort and effectiveness of respiratory support. Simultaneously, the explicit non-responsive state ensures the system's stability during non-triggered periods, avoiding unnecessary intervention.

[0059] In this regard, this application proposes a computer device including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, and when the programs are executed by the processors, they perform the steps as described in any of the above-described respiratory support methods.

[0060] In some of the embodiments described above in this application, addressing the problems of insufficient ventilation and human-machine asynchrony in existing respiratory support equipment, this embodiment achieves automated and precise control of the respiratory support process through deep integration of hardware and software. Specifically, one or more processors provide real-time computing capabilities to process multi-source respiratory physiological signals and execute respiratory intention recognition algorithms, ensuring rapid response and accurate decision-making to the patient's respiratory status; the memory stores the respiratory intention recognition algorithm, control parameters, and historical respiratory data, supporting stable system operation and persistent storage of key information; one or more programs contain an instruction set for implementing the respiratory support method, which, when executed by the processor, automatically completes the acquisition of respiratory physiological signals, the recognition of respiratory intentions, and the coordinated control of the negative pressure ventilation module and the constant current electrical stimulation module. For example, when the inspiratory initiation is detected, a high-level signal is sent to start the vacuum pump, the two-position three-way valve, and the electrical stimulation module, and a low-level signal is sent to stop the relevant operations when the expiratory initiation is detected.

[0061] Through the above technical solutions, computer equipment can accurately match the patient's respiratory needs based on real-time analysis of multi-source respiratory physiological signals. During the inspiratory phase, it can simultaneously enhance negative pressure ventilation and electrical stimulation to increase ventilation volume, and during the expiratory phase, it can promptly release negative pressure and stop electrical stimulation, thereby effectively solving the problems of insufficient ventilation and patient-ventilator asynchrony, and significantly improving the treatment efficiency and patient comfort of respiratory support.

[0062] Traditional respiratory support systems often face challenges such as patient-ventilator asynchrony and insufficient ventilation in clinical applications, leading to reduced treatment efficiency and patient discomfort. To address this issue, embodiments of this application propose a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned respiratory support method.

[0063] The computer program is configured to perform data acquisition via a signal detection module, processing via a respiratory intention recognition algorithm, and coordinated control of the negative pressure ventilation module and the constant current electrical stimulation module. Specifically, when the processor executes the program, it first acquires real-time signals of the patient's respiratory muscle electromyography, chest posture, chest and abdominal deformation, and respiratory airflow through the signal detection module. Then, it uses the respiratory intention recognition algorithm to extract and fuse features from the multi-source signals to identify inspiratory and expiratory intentions. Finally, based on the recognition results, it automatically controls the operating states of the negative pressure ventilation module and the constant current electrical stimulation module. For example, upon detecting an inspiratory intention, the program triggers the simultaneous operation of the negative pressure ventilation module and the constant current electrical stimulation module, creating a negative pressure cavity through the thoracic cavity and applying electrical stimulation to enhance diaphragmatic contraction. Upon detecting an expiratory intention, the program stops the operation of the relevant modules, allowing the thoracic cavity to communicate with the atmosphere to complete the expiratory process. This automated execution mechanism avoids human intervention errors and significantly improves respiratory synchronization accuracy.

[0064] Furthermore, this computer-readable storage medium provides a persistent and portable program storage solution, enabling flexible deployment of respiratory support methods across different hardware platforms. Through the aforementioned technical solution, the system achieves precise triggering and human-machine synchronization of the respiratory support process, effectively enhancing human-machine coupling and increasing ventilation, thereby resolving the problems of human-machine asynchrony and insufficient ventilation in existing technologies, and improving the efficacy and efficiency of respiratory support therapy.

[0065] Example 3 The following example will provide a more detailed explanation of the above technical solution: Consider a patient with a neuromuscular disease causing respiratory muscle weakness, or a patient in the recovery phase of acute respiratory distress syndrome (ARDS) who has insufficient spontaneous breathing ability but still has some spontaneous breathing drive. Traditional positive pressure ventilation may worsen lung injury due to the mechanical pressure it exerts on the lungs, while existing negative pressure ventilation machines cannot provide sufficient ventilation and are prone to patient-ventilator asynchrony.

[0066] The signal detection module acquires the respiratory physiological signals of patient A in real time. Specifically, the patient wears a customized chestplate, which is sealed tightly to the outside of the rib cage using medical-grade silicone. Simultaneously, surface electrodes are attached to the patient's skin to collect electromyographic signals of the respiratory muscles, an IMU sensor is fixed below the sternum to measure thoracic posture signals, flexible strain gauges are attached to the chest and abdomen to monitor deformation signals, and a bidirectional flow sensor is connected to the patient's mouth and nose to monitor respiratory airflow in real time. These multi-source physiological signals are transmitted to the control module in real time.

[0067] After receiving these raw signals, the control module first performs preprocessing. For electromyography (EMG) signals, interference and noise are filtered out using multi-stage amplification and filtering circuits (such as high-pass filters, low-pass filters, and power frequency notch filters). For chest and abdominal deformation signals, a Wheatstone bridge circuit is used for measurement. For pose and respiratory airflow signals, a Butterworth high-pass filter is used to filter out low-frequency artifact noise, and all signals are normalized. After preprocessing, the control module performs feature extraction on different signals. For example, for EMG signals, time-domain features such as absolute mean, variance, root mean square (RMS), envelope, and fixed sample entropy are extracted, as well as frequency-domain information such as median frequency and average power frequency. For pose and deformation signals, time-domain features such as RMS and variance are extracted, and frequency-domain information is extracted using Fourier transform and short-time Fourier transform. Subsequently, the control module performs feature concatenation on these time-frequency domain information from different sources to construct a high-dimensional feature matrix. Principal component analysis is then used to reduce the dimensionality and fuse the concatenated high-dimensional feature matrix in order to retain the main information of the original data and reduce redundancy.

[0068] The control module utilizes a built-in breathing intention recognition algorithm to process the fused feature data. This algorithm, based on a Long Short-Term Memory Recurrent Neural Network (LSTM) model, reconstructs the respiratory timing waveform. By performing threshold detection and peak extraction on the reconstructed waveform, the control module can accurately determine the patient's breathing intention, identify the start and end points of the inspiratory and expiratory phases, and extract respiratory pattern parameters such as respiratory rate, inspiratory time, and expiratory time.

[0069] When the control module determines that patient A has an intention to inhale (for example, it identifies the inspiratory start and encodes it as 1), it immediately sends a trigger signal to make the negative pressure ventilation module and the constant current electrical stimulation module work simultaneously.

[0070] When the negative pressure ventilation module starts working, the vacuum pump activates, the pressure reducing valve adjusts to the preset negative pressure value, and the two-position three-way valve switches to the left position. At this time, a thoracic shield made of transparent acrylic material with medical-grade silicone edges is tightly attached to the outside of the patient's chest, forming a sealed negative pressure cavity between the thoracic shield and the chest and abdominal surface. The negative pressure generated by the vacuum pump creates a local vacuum within this cavity, thereby reducing the pressure in the patient's pleural cavity and alveoli. Under the influence of external atmospheric pressure, air flows into the lungs through the mouth, nose, and airways, assisting in lung expansion.

[0071] Simultaneously, the constant current electrical stimulation module also begins to operate. This module includes a constant current drive circuit, isolation and protection circuits, and a power amplifier circuit. It applies a constant current electrical stimulation signal to the patient's phrenic nerve. This electrical stimulation causes the diaphragm to contract, further increasing the negative pressure in the thoracic cavity, thereby working in conjunction with the negative pressure ventilation module to promote more airflow into the lungs. Through this synergistic effect of external negative pressure and electrical stimulation of the diaphragm, patient A's respiratory ventilation is significantly improved, matching their actual ventilation needs. This effectively solves the problem of insufficient ventilation compared to existing respiratory support methods that rely solely on external negative pressure.

[0072] When the control module determines that patient A intends to exhale (e.g., it identifies the start of exhalation and encodes it as 2), it shuts off the vacuum pump, switches the two-position three-way valve to the right position, and simultaneously stops electrical stimulation. At this time, the thoracic cavity is open to the atmosphere, and the intrapulmonary pressure is higher than the external atmospheric pressure, forming an outward pressure gradient. Under the relaxation of the diaphragm and the opposing traction of the thoracic cavity and lung tissue, the gas in the lungs is expelled through the expiratory tract, completing the exhalation process.

[0073] Throughout the respiratory support process, the control module uses multi-source respiratory physiological signals to determine the patient's breathing intention in real time and precisely controls the timing of the negative pressure ventilation module and the constant current electrical stimulation module to ensure synchronization with the patient's inspiratory and expiratory phases. This breathing intention recognition method based on multi-source information fusion serves as the trigger for the respiratory support system, achieving human-machine synchronization during respiratory support, effectively reducing human-machine asynchrony, improving the efficacy and efficiency of treatment, and overcoming the serious human-machine asynchrony problem in existing technologies.

[0074] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A respiratory support system with external negative pressure combined with electrical stimulation, characterized in that, include: The signal detection module is used to collect the patient's respiratory physiological signals in real time, including at least the electromyography of the respiratory muscles on the body surface, chest posture, chest and abdominal deformation, and respiratory airflow signals. The negative pressure ventilation module is used to generate negative pressure outside the patient's chest cavity to reduce the pressure in the pleural cavity and alveoli, and promote the inflow of air into the lungs; The constant current electrical stimulation module is used to apply constant current electrical stimulation to the phrenic nerve to promote diaphragmatic contraction and enhance diaphragmatic contractile force. The control module is used to receive respiratory physiological signals collected by the signal detection module, and to determine the patient's breathing intention based on the built-in breathing intention recognition algorithm, and control the operation of the negative pressure ventilation module and the constant current electrical stimulation module to achieve the triggering of the inspiratory and expiratory phases and human-machine synchronization.

2. The respiratory support system according to claim 1, characterized in that, The negative pressure ventilation module consists of a chest armor, a pressure sensor, a two-position three-way valve, a vacuum pump, and a pressure reducing valve.

3. The respiratory support system according to claim 2, characterized in that, The pressure reducing valve is used to adjust the pressure generated by the vacuum pump, and the two-position three-way valve is used to switch and regulate the pressure inside the thoracic armor.

4. The respiratory support system according to claim 1, characterized in that, The constant current electrical stimulation module includes a constant current drive circuit, an isolation and protection circuit, and a power amplifier circuit.

5. The respiratory support system according to claim 1, characterized in that, The controller module includes a controller, an A / D converter, and a D / A converter.

6. A respiratory support method using external negative pressure combined with electrical stimulation, characterized in that, include: The signal detection module collects the patient's respiratory physiological signals in real time and sends them to the control module. The control module preprocesses the received signals through a built-in respiratory intention recognition algorithm and reconstructs the respiratory timing waveform to determine the patient's respiratory intention. Based on the respiratory intention, the control timing of the negative pressure ventilation module and the constant current electrical stimulation module is synchronized with the patient's inspiratory and expiratory phases. in, If the intention to inhale is detected, the controller module sends a trigger signal to make the negative pressure ventilation module and the constant current electrical stimulation module work simultaneously. If the system detects an intention to exhale, the controller module shuts off the vacuum pump, switches the two-position three-way valve to the right position, and stops the electrical stimulation.

7. The respiratory support method according to claim 6, characterized in that, If the intention to inhale is detected, the vacuum pump is activated, the pressure reducing valve is adjusted to the preset pressure, and the two-position three-way valve is switched to the left position, forming a negative pressure cavity between the thoracic armor and the surface of the chest and abdomen, generating a local vacuum to reduce the pressure in the pleural cavity and alveoli, assisting lung expansion, and allowing external air to flow into the lungs. At the same time, the constant current electrical stimulation module sends an electrical stimulation signal to the phrenic nerve, causing the diaphragm to contract and increase the negative pressure in the pleural cavity.

8. The respiratory support method according to claim 6, characterized in that, The preprocessing includes: splicing the time-frequency domain information of surface respiratory muscle electromyography, chest posture, chest and abdominal deformation, and respiratory airflow signals from different sources to construct a high-dimensional feature matrix, and using principal component analysis to reduce the dimensionality and fuse the spliced ​​high-dimensional feature matrix.

9. A computer device, characterized in that, It includes one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the programs, when executed by the processors, implement the steps of the respiratory support method as described in any one of claims 6-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the respiratory support method as described in any one of claims 6-8.

Images