Breathing physiotherapy device

Abstract

The invention relates to a breathing physical therapy device, belongs to the technical field of medical rehabilitation equipment, and solves the problem that existing breathing training equipment cannot intelligently adjust resistance according to the real-time breathing effort degree of a user. The device comprises a mouth-nose mask, an airflow resistance adjusting valve, a diaphragm myoelectricity sensor and a pressure difference sensor, a control module monitors the pressure difference rising rate and the current myoelectricity integral value in real time in the treatment stage based on a myoelectricity integral reference value calculated in the calibration stage, and accordingly resistance adjusting instructions including resistance increasing, resistance reducing and emergency resistance reducing instructions are generated; and an instruction priority and execution logic are set, so that timely intervention is ensured when the breathing of the user is abnormal. The device is mainly used for respiratory muscle strength training and rehabilitation treatment of patients with chronic respiratory diseases, and the training safety and the individuation level can be effectively improved.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation equipment technology. More specifically, this invention relates to respiratory physiotherapy devices. Background Technology

[0002] Most devices used in respiratory rehabilitation training currently employ fixed resistance or manually adjustable resistance. These devices cannot dynamically adjust according to the user's real-time breathing effort, resulting in a lack of personalized adaptation during training. Since the user's breathing effort changes dynamically during training, fixed resistance settings are difficult to match the user's actual physiological state, potentially leading to undertraining or overtraining.

[0003] Some devices attempt to automatically adjust resistance by monitoring changes in respiratory airflow or pressure, but relying solely on air pressure parameters often fails to accurately reflect the actual exertion of the user's respiratory muscles. The correlation between respiratory effort and air pressure signals is not entirely linear, especially in individuals with weak respiratory muscles or abnormal breathing patterns. Therefore, relying solely on air pressure signals for resistance regulation has certain limitations.

[0004] To improve the accuracy of regulation, some studies have proposed introducing electromyography (EMG) signals as a direct indicator of respiratory effort. Diaphragmatic EMG signals can directly reflect the respiratory center's drive and the intensity of diaphragmatic contraction. However, in practical applications, effectively combining EMG signals with barometric pressure signals for real-time control remains challenging. On the one hand, EMG signals are susceptible to noise interference and have poor signal stability; on the other hand, when multiple signals participate in regulation, command conflicts or response lags can easily occur.

[0005] Furthermore, existing devices lack effective prioritization mechanisms when abnormal situations such as excessively rapid breathing rates or sudden changes in effort occur. Determining the execution order to ensure user safety when multiple adjustment commands are generated simultaneously is also an unresolved issue. These factors limit the widespread clinical application and effectiveness of breathing training devices. Summary of the Invention

[0006] This invention provides a respiratory physiotherapy device that can effectively address abnormal situations such as excessively rapid breathing rate or sudden changes in effort level. By establishing a clear instruction priority processing mechanism, it solves the problem of determining the execution order when multiple adjustment instructions are generated simultaneously, thereby significantly improving the safety and clinical applicability of the device.

[0007] This invention provides a respiratory physiotherapy device, comprising: The mouth and nose mask is equipped with a one-way air inlet and a one-way air outlet; An airflow resistance regulating valve is located at the front end of the one-way air inlet; The diaphragmatic electromyography sensor contains two surface electrodes and is attached 3-5 cm below the xiphoid process of the user. A differential pressure sensor is installed on the mouth and nose mask, and its first and second pressure taps are connected to the inside of the mask and the outside atmosphere, respectively. The control module, which is connected to the airflow resistance regulating valve, the diaphragm electromyography sensor, and the differential pressure sensor, is configured as follows: During the calibration phase, when the user takes multiple calm breaths, the average value of the diaphragm electromyography integral value in each inspiratory phase is calculated based on the signal output by the diaphragm electromyography sensor and set as the electromyography integral reference value E0. During each inspiratory phase of the treatment, the rate of change of pressure difference d△P / dt is monitored by a differential pressure sensor, and the current electromyographic integral value E1 is calculated. If d△P / dt>5cmH2O / s, the first adjustment command is generated: control the airflow resistance regulating valve to increase the resistance by 3-5cmH2O; If E1 < 120%E0, a second adjustment command is generated: increase the resistance by 2-4 cmH2O; If E1 > 150%E0, a resistance reduction instruction is generated: after reducing the resistance by 4-6 cmH2O in the subsequent expiratory phase, the resistance reduction instruction is cleared. The respiratory rate is calculated based on the signal output by the differential pressure sensor. If the average respiratory rate of multiple consecutive respiratory cycles is >25 breaths / min, an emergency resistance reduction command is generated: reduce the resistance to 5 cmH2O and maintain it for 60 seconds. If the first adjustment command, the second adjustment command, and the emergency drag reduction command are generated simultaneously, then one operation is executed in the order of the emergency drag reduction command, the first adjustment command, and the second adjustment command. If a drag reduction instruction is generated during the execution of any instruction, the drag reduction instruction is ignored; if an emergency drag reduction instruction or a first adjustment instruction is generated after the drag reduction instruction is generated but before its execution, the drag reduction instruction is abandoned, and one operation is executed in the order of the emergency drag reduction instruction and the first adjustment instruction.

[0008] Preferably, in the respiratory therapy device, the control module is further configured as follows: After the respiratory rate is calculated based on the signal output by the differential pressure sensor during the calibration phase, the fluctuation of the respiratory rate is judged by calculating the range or standard deviation of the frequency over 3-5 consecutive respiratory cycles. The control module calculates the peak flow rate of the intake phase based on data monitored by the differential pressure sensor; The control module records a breath as a valid breath if it simultaneously meets the conditions that the respiratory rate fluctuation is less than the preset frequency fluctuation threshold and the peak inspiratory flow is within the preset flow range, and counts the number of valid breaths. The control module completes the calculation of the electromyographic integral baseline value during the calibration phase, and the following conditions must be met simultaneously: The respiratory rate fluctuation is less than the preset frequency fluctuation threshold for 3-5 consecutive respiratory cycles; The peak inspiratory flow rate during each inspiratory phase is within the preset flow rate range for 3-5 consecutive respiratory cycles; And the number of effective breaths collected reached the preset minimum number; Only when all of the above conditions are met will the control module set the average value of the electromyographic integral of this group of effective breathing as the electromyographic integral reference value E0.

[0009] Preferably, in the respiratory physiotherapy device, before the control module calculates the current electromyographic integral value E1 during the treatment phase, it filters the raw signal collected by the diaphragm electromyographic sensor, with a filter passband frequency range of 10-500Hz.

[0010] Preferably, in the respiratory therapy device, the control module is further configured as follows: During the treatment phase, if the current electromyographic integral value E1 remains below 80% of the electromyographic integral benchmark value E0 for several respiratory cycles, an alarm signal is generated and the execution of the first regulation command, the second regulation command, and the resistance reduction command is suspended; the generation and execution of the emergency resistance reduction command are not affected.

[0011] Preferably, in the respiratory therapy device, the control module is further configured as follows: Perform signal quality monitoring during the treatment phase; Signal quality monitoring includes analyzing the raw signals acquired by the diaphragm electromyography sensor. The analysis methods include at least one of calculating the signal amplitude, assessing the noise level, or detecting abnormal waveforms. If the analysis results indicate that the signal quality is lower than the preset quality threshold for multiple consecutive respiratory cycles, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be paused, and a signal abnormality warning will be generated. If the signal amplitude output by the diaphragm electromyography sensor is lower than the preset minimum threshold for 5 consecutive seconds, an electrode detachment warning is generated, and the system switches to pure pneumatic feedback regulation mode. In pure pneumatic feedback regulation mode, the first regulation command and emergency resistance reduction command are generated only based on the signal output by the differential pressure sensor.

[0012] Preferably, in the respiratory physiotherapy device, before the control module generates the first adjustment command, the second adjustment command, or the resistance reduction command during the treatment phase, it also performs signal mutation collaborative verification: synchronously monitoring the slope K of the differential pressure sensor signal change. p Slope K of the diaphragm electromyography sensor signal change e ; When K p With K e The Pearson correlation coefficient is >0.7, and |K p |>5cmH2O / s, while Ke >When the slope of the diaphragm electromyography sensor signal change obtained from the statistics of calm breathing during the calibration phase reaches the upper limit of the normal range, it is determined to be an effective deep breathing effort; When K p With K e The Pearson correlation coefficient is ≤0.3, or K p With K e When the signs of the positive and negative symbols are opposite, it is determined to be motion artifact interference; The dynamic threshold adjustment algorithm is executed as follows: when motion artifact interference is detected, starting from the initial safety threshold Th0 = 5 cmH2O / s, the threshold is adjusted according to the formula Th... current = Th0 + 0.5T to calculate the current safety threshold, where T is the duration of the interference in seconds. When |K is detected continuously for 2 seconds... p When all values ​​are below the current safety threshold, the signal is considered to have stabilized. The judgment process of generating the first adjustment command, the second adjustment command, and the drag reduction command will continue only when the deep breathing effort is determined to be effective. If it is determined to be motion artifact interference, or K p With K e If the Pearson correlation coefficient is ≤0.5, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be suspended, and the signal output by the diaphragm electromyography sensor will be marked as needing recalibration; the generation and execution of the emergency resistance reduction command will not be affected by this. When K p In the inspiratory phase, the instantaneous flow rate exceeds 10 cmH2O / s, and K p With K e When the Pearson correlation coefficient is <0.3, it is determined to be choking, and an emergency choking resistance reduction command is generated: reduce the resistance to 5cmH2O and maintain it for 60s. The emergency choking resistance reduction command is executed immediately and has the same priority as the emergency resistance reduction command.

[0013] Preferably, in the respiratory therapy device, the control module is further configured as follows: During the treatment phase, if two or more first or second regulatory commands are generated within three consecutive respiratory cycles, the execution of subsequent first and second regulatory commands shall be suspended, and the resistance shall be maintained unchanged in the next expiratory phase until the respiratory rate recovers to below 20 breaths / min and E1 is within 120%-130% of E0. If an emergency drag reduction command is detected during the drag maintenance period, the emergency drag reduction command shall be executed immediately.

[0014] Preferably, in the respiratory therapy device, the control module is further configured as follows: If the E1 value is below 70% or above 130% of E0 for three consecutive respiratory cycles during the treatment phase, a transition phase lasting at least five respiratory cycles will be initiated immediately. Or, after executing the emergency decompression command or exiting the resistance maintenance period, enter a transition phase lasting at least 5 respiratory cycles; During the transition phase, the generation of the first adjustment command, the second adjustment command, and the drag reduction command will be suspended; the generation and execution of the emergency drag reduction command will not be affected by the transition phase. During the transition phase, the integral value of electromyography (EMG) for each complete respiratory cycle is calculated in real time. At the end of the transition phase, the arithmetic mean of all calculated EMG integral values ​​is taken and set as Em. 0过渡 And this is used to temporarily replace the electromyographic integral baseline value E0; After the transition phase ends, a gradual change process for the electromyographic integral baseline value is initiated; During the gradual change of the electromyographic integral baseline value, at the beginning of each subsequent respiratory cycle, the electromyographic integral baseline value used for comparison is changed from E... 0过渡 Adjust the value to the original electromyography integral baseline value E0 once until the difference between the two is less than the preset threshold or reaches the original electromyography integral baseline value E0. During the gradual change of the electromyographic integral baseline value, the following instruction generation logic is resumed: monitor the rate of increase of differential pressure d△P / dt and calculate the current electromyographic integral value E1, and determine whether to generate the first adjustment instruction, the second adjustment instruction, or the resistance reduction instruction based on this.

[0015] Preferably, in the respiratory physiotherapy device, the control module is further configured to: During the gradual change of the electromyographic integral reference value. While calculating the current electromyographic integral value E1 in each respiratory cycle, the short-term rate of change δ of the current electromyographic integral value E1 relative to the electromyographic integral value of the previous respiratory cycle is also calculated. If the absolute value of the short-term change rate δ of the electromyography integral value for two consecutive respiratory cycles is greater than the preset change rate threshold, the gradual change process of the electromyography integral benchmark value is immediately stopped, and the current electromyography integral benchmark value used for comparison is locked as the current electromyography integral value E1, which is used as the new electromyography integral benchmark value. The generation and execution of emergency drag reduction commands are not affected by the abort of the gradual change process.

[0016] The present invention has at least the following beneficial effects: This invention significantly improves the safety and effectiveness of breathing training through dual-signal monitoring and intelligent priority adjustment mechanisms, resulting in enhanced respiratory muscle strength and reduced subjective fatigue in patients after training.

[0017] This invention solves the problem of inaccurate baseline value calculation caused by respiratory instability during the calibration phase. It ensures that the baseline value comes from a period of stable breathing through a multi-condition verification mechanism, providing a reliable and personalized comparison benchmark for the treatment phase.

[0018] This invention solves the problem of execution conflict between emergency drag reduction command and drag reduction command when an emergency drag reduction command is issued. It ensures that safety commands are executed first by clear command priority logic, thus protecting user safety.

[0019] This invention solves the problem of inaccurate integral calculation caused by noise interference in the diaphragm electromyography signal. By filtering, the signal-to-noise ratio is improved, ensuring that the integral value truly reflects the degree of respiratory effort.

[0020] This invention solves the problem that resistance adjustment may cause discomfort when the user's breathing effort is insufficient. It avoids overload and improves training safety by continuously monitoring and pausing the adjustment mechanism.

[0021] This invention solves the problem of control failure caused by deterioration of sensor signal quality or electrode detachment. It maintains basic safety functions and improves equipment robustness through signal quality monitoring and fault response mechanisms.

[0022] This invention solves the problem of falsely triggering commands due to signal mutations. It effectively distinguishes between deep breathing, motion interference, and coughing through a signal mutation collaborative verification mechanism, thus avoiding false adjustments.

[0023] This invention solves the problem of excessively rapid breathing rate or abnormal effort caused by continuous resistance increase. It prevents the resistance from rising indefinitely through trend judgment and intervention mechanisms, and provides a recovery and adaptation period.

[0024] This invention solves the problem that the benchmark value becomes inapplicable due to changes in the user's state. It intelligently adjusts the evaluation criteria through a dynamic benchmark management mechanism to maintain the accuracy of resistance adjustment.

[0025] This invention solves the problem of inappropriate baseline adjustment caused by sudden changes in breathing effort during the gradual change of baseline value. It achieves rapid adaptation and improves control accuracy through short-term rate of change monitoring and dynamic termination mechanism.

[0026] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating the use of a respiratory physiotherapy device according to an embodiment of the present invention. Detailed Implementation

[0028] The present invention will now be described in further detail so that those skilled in the art can implement it based on the description.

[0029] like Figure 1 As shown, the present invention provides a respiratory physiotherapy device, comprising: A mouth and nose mask, which has a one-way air inlet and a one-way air outlet; The airflow resistance regulating valve is located at the front end of the one-way air inlet, and the resistance value can be adjusted within the range of 5-30 cmH2O. Diaphragmatic electromyography sensor, which includes two surface electrodes, is attached 3-5 cm below the xiphoid process of the user; A differential pressure sensor is installed on the mouth and nose mask. Its first pressure tap and second pressure tap are respectively connected to the internal cavity of the mask and the external atmosphere, and are used to monitor the pressure difference between the inside and outside of the mask in real time. The control module, which is connected to the airflow resistance regulating valve, the diaphragm electromyography sensor, and the differential pressure sensor, is configured as follows: During the calibration phase, when guiding the user to perform multiple calm breaths, the average value of the diaphragm electromyography integral value in each inspiratory phase is calculated based on the signal output by the diaphragm electromyography sensor and set as the electromyography integral reference value E0. During each inspiratory phase of the treatment, the rate of change of pressure difference d△P / dt is monitored by a differential pressure sensor, and the current electromyographic integral value E1 is calculated. If d△P / dt>5cmH2O / s, then the first adjustment command is generated: control the airflow resistance regulating valve to increase the resistance by 3-5cmH2O; Compare E1 with 120%E0 and 150%E0: If E1 < 120%E0, then a second adjustment command is generated: increase the resistance by 2-4 cmH2O; If E1 > 150%E0, a resistance reduction instruction is generated: after reducing the resistance by 4-6 cmH2O in the subsequent expiratory phase, the resistance reduction instruction is cleared. If 120%E0≤E1≤150%E0, then the resistance remains unchanged; The respiratory rate f is calculated based on the signal output by the differential pressure sensor. If the average respiratory rate f over multiple consecutive respiratory cycles is... avg >25 times / min, generate emergency resistance reduction command: reduce resistance to 5cmH2O and maintain for 60s; If the first adjustment command, the second adjustment command, and the emergency drag reduction command are generated simultaneously, one operation shall be executed in the following order: emergency drag reduction command first, first adjustment command second, and second adjustment command last. If a drag reduction instruction is generated during the execution of any instruction, the drag reduction instruction is ignored; if an emergency drag reduction instruction or a first adjustment instruction is generated after the drag reduction instruction is generated but before its execution, the drag reduction instruction is abandoned, and an operation is executed in the order of priority of the emergency drag reduction instruction and the first adjustment instruction.

[0030] Most devices used in respiratory rehabilitation training currently employ fixed resistance or manually adjustable resistance. These devices cannot dynamically adjust resistance based on the user's real-time respiratory effort, resulting in a lack of personalized adaptation during training. Some devices attempt to automatically adjust resistance by monitoring changes in airflow or pressure, but relying solely on air pressure parameters often fails to accurately reflect the actual exertion of the user's respiratory muscles. The correlation between respiratory effort and air pressure signals is not entirely linear, especially in individuals with weak respiratory muscles or abnormal breathing patterns. Therefore, resistance regulation based solely on air pressure signals has certain limitations.

[0031] The respiratory therapy device provided in this solution includes a nasal mask, an airflow resistance regulating valve, a diaphragm electromyography (EMG) sensor, a differential pressure sensor, and a control module. During the calibration phase, the control module guides the user through multiple calm breaths. Based on the signal output from the diaphragm EMG sensor, it calculates the average value of the diaphragm EMG integral during each inspiratory phase, setting this as the EMG integral baseline value E0. The purpose of this step is to establish a personalized baseline reflecting the user's respiratory effort in a calm state, providing a comparative standard for subsequent treatment phases. During each inspiratory phase in the treatment phase, the device monitors the rate of pressure rise (dΔP / dt) using the differential pressure sensor and calculates the current EMG integral value E1. Monitoring the rate of pressure rise allows for timely detection of changes in airflow, while calculating the EMG integral value directly reflects the actual contraction strength of the diaphragm.

[0032] If d△P / dt > 5 cmH2O / s, a first adjustment command is generated: the airflow resistance regulating valve is increased by 3-5 cmH2O. The purpose of this step is to appropriately increase resistance when excessive airflow is detected, preventing the user from reducing effective respiratory muscle training due to excessive airflow. If E1 < 120%E0, a second adjustment command is generated: the resistance is increased by 2-4 cmH2O. The effect of this step is to increase resistance when the user's breathing effort is insufficient, promoting sufficient training of the respiratory muscles. If E1 > 150%E0, a resistance reduction command is generated: the resistance is reduced by 4-6 cmH2O in the subsequent expiratory phase, and then the resistance reduction command is cleared. The purpose of this step is to reduce resistance when the user's breathing effort is excessive, preventing overload leading to muscle fatigue or discomfort.

[0033] The respiratory rate f is calculated based on the signal output by the differential pressure sensor. If the average respiratory rate f over multiple consecutive respiratory cycles is... avgIf the breathing rate exceeds 25 breaths / min, an emergency drag reduction command is generated to lower the resistance to 5 cmH2O and maintain this level for 60 seconds. This step rapidly reduces resistance when excessively rapid breathing is detected, ensuring user breathing safety and preventing hyperventilation. When multiple commands are generated simultaneously, the device executes one operation in the order of emergency drag reduction command, first adjustment command, and second adjustment command, ensuring that safety commands take priority. If a drag reduction command is generated during the execution of any command, it is ignored; if an emergency drag reduction command or first adjustment command is generated after the drag reduction command is generated but before its execution, the drag reduction command is abandoned and one operation is executed according to priority. This step aims to resolve conflicts between multiple commands, ensuring consistent system response and safety.

[0034] This solution significantly improves the device's responsiveness and safety in multi-command scenarios by introducing a command priority processing mechanism. When an emergency drag reduction command occurs between the generation and execution of the original command, the system immediately abandons the planned drag reduction command and executes the emergency drag reduction command in the current inspiratory phase. This approach ensures that in cases of excessively high respiratory rate or other emergencies, the device prioritizes drag reduction, preventing further respiratory burden due to resistance mismatch and ensuring user safety during training.

[0035] The purpose of this mechanism is to clearly define the execution order when multiple commands coexist, avoiding operational delays or misoperations caused by command stacking or conflicts. Its effectiveness lies in its ability to identify the urgency of commands in real time and promptly interrupt the execution of existing commands when a higher-priority command is detected, responding instead to the most urgent control request. By immediately implementing emergency drag reduction during the current inspiratory phase, the system can alleviate the user's respiratory pressure in the shortest possible time, preventing the situation from worsening, while maintaining the continuity and consistency of control, thus improving device reliability and user trust.

[0036] Example 1 A 52-year-old male patient with chronic obstructive pulmonary disease (COPD) underwent training using this respiratory therapy device. The patient was 1.70m tall, weighed 65kg, and pulmonary function tests showed an FEV1 / FVC ratio of 58%. Before training began, the patient's resting respiratory rate was 22 breaths / min, and his oxygen saturation was 95%. During the device calibration phase, the diaphragmatic electromyography (EMG) signals of the patient's diaphragm during five quiet breaths were recorded, and the calculated E0 integral value was 48 μV / s. After the treatment phase began, the device monitored the rate of pressure rise during the patient's third inspiration, dΔP / dt, which was 4.2 cmH2O / s, and simultaneously monitored the current E1 integral value, which was 42 μV / s. Since E1 < 120%E0 but dΔP / dt did not exceed 5 cmH2O / s, the device generated a second adjustment command, increasing the resistance from the initial 5 cmH2O to 8 cmH2O.

[0037] When the training had been underway for 5 minutes, the device detected an E1 value of 50 μVs (approximately 104% E0). Since E1 < 120% E0 and dΔP / dt is 4.8 cmH2O / s, which does not exceed 5 cmH2O / s, the device generated a second adjustment command, increasing the resistance from 8 cmH2O to 10 cmH2O.

[0038] At 8 minutes into the training, the device detected that the E1 value had risen to 52 μVs (approximately 108% E0), while dΔP / dt reached 5.3 cmH2O / s. At this point, E1 < 120% E0, which should have triggered the second adjustment command. However, since dΔP / dt was already greater than 5 cmH2O / s, the device prioritized generating the first adjustment command. Simultaneously, the device detected a respiratory rate of 19 breaths / min, not meeting the emergency resistance reduction condition. According to the command priority rule, the device prioritized executing the first adjustment command, further increasing the resistance from 10 cmH2O to 13 cmH2O, while abandoning the possible execution of the second adjustment command. Throughout the adjustment process, the device continuously monitored various parameters to ensure a smooth adjustment process.

[0039] The entire training process lasted 20 minutes. At the end, the patient's respiratory rate was 16 breaths / min, blood oxygen saturation remained at 96%, and the subjective fatigue score was 3. Post-training testing showed that the patient's maximum inspiratory pressure increased from 65 cmH2O to 72 cmH2O, indicating effective enhancement of respiratory muscle strength. Throughout the training process, the device performed four resistance adjustments in an orderly manner based on dual-signal monitoring and priority logic, ensuring both training effectiveness and patient safety and comfort.

[0040] Comparative Example 1 The same patient underwent training using a traditional respiratory training device that relies solely on pressure signals for adjustment. The initial resistance was set to 10 cmH2O. After training began, due to the patient's unstable breathing pattern, the device detected significant pressure fluctuations and automatically adjusted the resistance to 18 cmH2O within 2 minutes. The patient immediately experienced shortness of breath, with the respiratory rate rising to 28 breaths / min. At 6 minutes of training, the device again lowered the resistance to 12 cmH2O based on the pressure signal, but by this time the patient had developed significant respiratory distress, with oxygen saturation dropping to 93% and a subjective fatigue score of 7. Training was forced to stop at 8 minutes. Subsequent testing showed no improvement in the patient's maximum inspiratory pressure; instead, it decreased to 62 cmH2O due to fatigue.

[0041] Comparative Example 2 The same patient underwent training using a conventional respiratory trainer with fixed resistance. Following the standard training protocol, the resistance was set at 15 cmH2O. Initially, the patient was able to maintain normal breathing, with a respiratory rate of 20 breaths / min. However, as training progressed, by the 5th minute, the patient's respiratory muscles began to fatigue, the respiratory rate decreased to 12 breaths / min, and the tidal volume decreased from 550 mL to 380 mL. By the 10th minute, the patient's oxygen saturation had dropped to 92%, exhibiting significant respiratory compensation. Training was forced to stop at the 12th minute. Subsequent assessments showed that the patient experienced respiratory muscle soreness for 24 hours after training, and the maximum inspiratory pressure decreased from 65 cmH2O to 60 cmH2O.

[0042] Comparative Example 3 The same patient used a breathing training device regulated solely by diaphragmatic electromyography (EMG) signals. During the calibration phase, the device was identical to that in Example 1. After five quiet breaths, the system calculated a baseline E0 of 48 μVs. The device's regulation logic was set as follows: during each inspiratory phase of the treatment, the current E1 was calculated in real-time. If E1 < 120%E0 (i.e., 57.6 μVs), a "resistance increase command" was generated, controlling the airflow resistance regulating valve to increase the resistance by 3 cmH2O; if E1 > 150%E0 (i.e., 72 μVs), a "resistance decrease command" was generated, decreasing the resistance by 4 cmH2O; if the E1 value was between 120% and 150%E0, the current resistance was maintained. At the start of training, the device resistance was set to 5 cmH2O. Initially, the patient's breathing was stable, and the E1 value monitored in the first respiratory cycle was 49 μVs (approximately 102%E0), below the 120% threshold. The device then performed its first resistance increase, raising the resistance to 8 cmH2O. In the second respiratory cycle, the E1 value was 50 μVs (approximately 104% E0), still below the threshold. The device executed a second resistance increase, raising the resistance to 11 cmH2O. At the 4-minute mark of training, the patient experienced airway narrowing. The pressure signal showed airflow obstruction, and the rate of pressure gradient increase, dΔP / dt, decreased to 3.2 cmH2O / s. However, the device, unable to process the pressure signal, only monitored an E1 value of 52 μVs (approximately 108% E0) due to compensatory effort. According to the algorithm, this value was still below the 120% resistance increase threshold, so the device generated another resistance increase command, raising the resistance from 11 cmH2O to 14 cmH2O. At the 6-minute mark, airway obstruction worsened, and the E1 value rose to 55 μVs (approximately 115% E0) due to severe compensatory diaphragmatic contraction. The device continued to determine that this value was below the threshold and generated a resistance increase command, raising the resistance to 17 cmH2O. The patient's respiratory rate subsequently increased significantly to 25 breaths / min, and oxygen saturation dropped to 93%. However, because the E1 value never exceeded the drag reduction threshold of 150%E0, the device maintained a high resistance state, ultimately forcing the training to be terminated at the 14th minute due to severe respiratory distress. Post-training assessment showed that the patient's maximum inspiratory pressure decreased from 65 cmH2O to 62 cmH2O, the subjective fatigue score reached 7, and the recovery time exceeded 48 hours.

[0043] Comparative Example 4 A device employing non-prioritized command processing was used. This device simultaneously monitors electromyography and barometric pressure signals, but processes multiple commands in parallel when they are generated simultaneously. At the 7-minute mark of training, the device simultaneously monitored dΔP / dt at 5.5 cmH2O / s and E1 at 49 μV / s, generating both a first and a second adjustment command. The device initially increased the resistance by 4 cmH2O, then by another 3 cmH2O, resulting in a rapid increase from 10 cmH2O to 17 cmH2O within a short period. This abrupt change caused the patient's respiratory pattern to become disordered, with respiratory rate fluctuating between 15-26 breaths / min and oxygen saturation dropping to 94%. Although the 20-minute training was ultimately completed, the patient's subjective fatigue score was 5, and the maximum inspiratory pressure only increased from 65 cmH2O to 67 cmH2O after training, indicating a significantly lower training effect than in Example 1.

[0044] Results: The dual-signal monitoring and intelligent priority adjustment mechanism used in Example 1 significantly improved the safety and effectiveness of breathing training. Patients experienced enhanced respiratory muscle strength and lower subjective fatigue after training. In contrast, Comparative Examples 1 to 4 exposed the limitations of traditional methods such as single-signal dependence, fixed resistance mode, and lack of priority processing. These methods may lead to problems such as disordered breathing patterns, decreased blood oxygen saturation, and excessive fatigue, thus demonstrating the comprehensive advantages of this invention in achieving personalized adaptation and safety assurance.

[0045] In another embodiment, the control module of the respiratory therapy device is further configured as follows: After the respiratory rate is calculated based on the signal output by the differential pressure sensor during the calibration phase, the fluctuation of the respiratory rate is judged by calculating the range or standard deviation of the frequency over 3-5 consecutive respiratory cycles. The control module calculates the peak flow rate of the intake phase based on data monitored by the differential pressure sensor; The control module records a breath as a valid breath if it simultaneously meets the conditions that the respiratory rate fluctuation is less than the preset frequency fluctuation threshold and the peak inspiratory flow is within the preset flow range, and counts the number of valid breaths. The control module completes the calculation of the electromyographic integral baseline value during the calibration phase, and the following conditions must be met simultaneously: The respiratory rate fluctuation is less than the preset frequency fluctuation threshold for 3-5 consecutive respiratory cycles; The peak inspiratory flow rate during each inspiratory phase is within the preset flow rate range for 3-5 consecutive respiratory cycles; And the number of effective breaths collected reached the preset minimum number; Only when all of the above conditions are met will the control module set the average value of the electromyographic integral of this group of effective breathing as the electromyographic integral reference value E0.

[0046] Currently, respiratory rehabilitation training equipment often uses a simple averaging algorithm for calibration, which directly uses the characteristic values ​​of respiratory signals collected over a period of time to calculate the baseline value. This method has low requirements for the user's respiratory stability. If the user experiences occasional deep breathing, coughing, or changes in breathing rhythm during the calibration phase, abnormal signal values ​​will be generated and included in the calculation, resulting in the final baseline value failing to accurately reflect the true level under calm conditions. Some devices have attempted to improve this problem by extending the acquisition time or manually removing obvious outliers, but extending the acquisition time increases the user's preparation burden, while manual intervention reduces the automation and operability of the equipment.

[0047] During the calibration phase, the respiratory therapy device provided in this solution first calculates the respiratory rate based on the signal output by the differential pressure sensor. It then determines the degree of respiratory rate fluctuation by calculating the range or standard deviation of the frequency over 3-5 consecutive respiratory cycles. This step identifies whether the respiratory rhythm is stable. Simultaneously, the control module calculates the peak inspiratory flow rate using data monitored by the differential pressure sensor. This parameter reflects the depth and intensity of each inspiratory breath. The device records effective breaths as those that simultaneously meet the conditions of respiratory rate fluctuation being less than a preset frequency fluctuation threshold and peak inspiratory flow rate being within a preset flow range. It also counts the number of effective breaths. This step ensures that only respiratory cycles with a stable breathing pattern and normal inspiratory depth are adopted.

[0048] When the control module calculates the electromyographic integral baseline value during the calibration phase, three conditions must be met simultaneously: the respiratory rate fluctuation for 3-5 consecutive respiratory cycles must be less than a preset frequency fluctuation threshold; the peak inspiratory flow rate of each inspiratory phase within the same time period must be within a preset flow rate range; and the number of effective breaths collected must reach a preset minimum. Only when all the above conditions are met will the control module set the average value of the electromyographic integral values ​​of this set of effective breaths as the electromyographic integral baseline value E0. This multi-condition verification mechanism effectively eliminates respiratory variability caused by accidental factors, ensuring that the baseline value comes from a period of stable breathing and constant effort, providing a reliable and personalized comparison benchmark for subsequent treatment stages, and significantly reducing the risk of misadjustment due to inaccurate calibration data.

[0049] In another embodiment, in the respiratory physiotherapy device, before calculating the current electromyographic integral value E1 during the treatment phase, the control module filters the raw signal collected by the diaphragm electromyographic sensor, with a filter passband frequency range of 10-500Hz.

[0050] Current breathing training devices often directly use the raw diaphragmatic electromyography (EMG) signals for integration calculations. EMG signals are highly susceptible to various noise interferences, including power frequency interference from the power supply, crosstalk from other muscle activities, and background noise caused by poor electrode-skin contact. These noise signals mix with the measured diaphragmatic EMG signals. If integration is performed directly without processing, the calculated EMG integral value will be severely distorted, failing to accurately reflect the user's true respiratory effort and thus misleading the generation of resistance adjustment commands.

[0051] Before calculating the current electromyographic integral value, this scheme introduces a crucial signal preprocessing step: filtering the raw signal acquired by the diaphragm electromyography sensor. The filter passband frequency range is set to 10-500Hz. This frequency range aims to effectively retain the effective electromyographic signal components characterizing diaphragmatic contraction while filtering out low-frequency baseline drift and high-frequency environmental electromagnetic noise.

[0052] The purpose of this filtering process is to purify the signal, improve the signal-to-noise ratio, and ensure the accuracy of subsequent integral calculations. Its effect is to suppress out-of-band noise, making the final electromyographic integral value more accurately represent the user's respiratory effort level. Based on this clean signal, the resistance adjustment commands generated are more reliable, reducing the risk of misjudgment and misadjustment caused by signal noise, and improving the precision of the entire device control and the effectiveness of the training process.

[0053] In another embodiment, the control module of the respiratory therapy device is further configured as follows: During the treatment phase, if the current electromyographic integral value E1 remains below 80% of the electromyographic integral benchmark value E0 for several respiratory cycles, an alarm signal is generated and the execution of the first regulation command, the second regulation command, and the resistance reduction command is suspended; the generation and execution of the emergency resistance reduction command are not affected.

[0054] Existing breathing training devices typically adjust resistance dynamically based on real-time monitoring of breathing effort signals, but lack the ability to assess the medium- to long-term trends of the user's condition. When a user's breathing effort remains consistently insufficient due to fatigue or other reasons, these devices may still mechanically execute resistance increase or decrease logic based on instantaneous values. This single, immediate judgment mode may fail to recognize that the user is already in a state of low overall effort, and continuing to execute the original resistance adjustment strategy may actually exacerbate the user's respiratory burden, leading to discomfort or training interruption, thus failing to achieve safe and effective rehabilitation training.

[0055] This approach incorporates a monitoring and response mechanism for persistent changes in respiratory effort. During treatment, the system continuously monitors multiple respiratory cycles. If the current integrated electromyography (EMG) value is consistently below 80% of the baseline value, a warning signal is generated, and routine resistance increase and decrease commands are suspended. This design enables the system to distinguish between occasional, transient insufficiency and persistent fatigue.

[0056] The purpose of this mechanism is to prevent the device from mechanically increasing resistance when the user's overall respiratory muscle strength is low and effort level remains consistently low, thereby avoiding overload and training risks. Its effect is to provide the user with a recovery period by pausing unnecessary resistance adjustments, while simultaneously alerting the user or medical personnel to their current condition through warning signals, thus improving the safety and humanization of the training process. The generation and execution of emergency resistance reduction commands are unaffected, ensuring that emergency situations such as shortness of breath are addressed first in any situation, achieving a balance between safety and personalized training.

[0057] In another embodiment, the control module of the respiratory therapy device is further configured as follows: Perform signal quality monitoring during the treatment phase; Signal quality monitoring includes analyzing the raw signals acquired by the diaphragm electromyography sensor. The analysis methods include at least one of calculating the signal amplitude, assessing the noise level, or detecting abnormal waveforms. If the analysis results indicate that the signal quality is lower than the preset quality threshold for multiple consecutive respiratory cycles, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be paused, and a signal abnormality warning will be generated. If the signal amplitude output by the diaphragm electromyography sensor is lower than the preset minimum threshold for 5 consecutive seconds, an electrode detachment warning is generated, and the system switches to pure pneumatic feedback regulation mode. In pure pneumatic feedback regulation mode, the first regulation command and emergency resistance reduction command are generated only based on the signal output by the differential pressure sensor.

[0058] Existing breathing training devices typically assume that sensor signals are always reliable during operation and adjust resistance accordingly. However, in actual use, the signal quality of diaphragmatic electromyography (EMG) sensors is easily affected by various factors, such as poor electrode-skin contact, impedance changes due to sweat, artifacts caused by limb movement, or partial sensor detachment. These factors can lead to a decrease in the quality of the acquired EMG signals or even complete failure. Traditional devices lack the ability to assess signal quality in real time and continue to use unreliable signals to generate control commands, resulting in misadjustment that may increase the user's respiratory load or pose safety risks.

[0059] This solution introduces a complete signal quality monitoring and fault response mechanism. During the treatment phase, the system continuously performs multi-dimensional analysis on the raw signals acquired by the diaphragmatic electromyography (EMG) sensor, including calculating signal amplitude, assessing noise levels, and detecting abnormal waveforms. When the analysis results indicate that the signal quality is below a preset quality threshold for several consecutive respiratory cycles, the system will pause the generation of routine adjustment commands based on the EMG signal and issue a signal abnormality warning to prevent erroneous judgments based on low-quality signals.

[0060] If the detected signal amplitude is below the preset minimum threshold for 5 consecutive seconds, it is judged as a serious fault such as electrode detachment. The system will generate an electrode detachment warning and automatically switch to pure pneumatic feedback regulation mode. In this mode, the device continues to generate the first regulation command and emergency resistance reduction command solely based on the differential pressure sensor signal, maintaining the most basic safety regulation function. This design ensures that even in the extreme case of complete failure of the electromyographic signal, the device can still provide a certain degree of safety based on the pneumatic signal, avoiding complete system paralysis due to the failure of a single sensor, and significantly improving the robustness and safety of the device.

[0061] In another embodiment, in the respiratory therapy device, before the control module generates the first adjustment command, the second adjustment command, or the resistance reduction command during the treatment phase, it also performs signal mutation collaborative verification: synchronously monitoring the slope K of the differential pressure sensor signal change. p Slope K of the diaphragm electromyography sensor signal change e ; When K p With K e The Pearson correlation coefficient is >0.7, and |K p |>5cmH2O / s, while K e >When the slope of the diaphragm electromyography sensor signal change obtained from the statistics of calm breathing during the calibration phase reaches the upper limit of the normal range, it is determined to be an effective deep breathing effort; When K p With K e The Pearson correlation coefficient is ≤0.3, or K p With K e When the signs of the positive and negative symbols are opposite, it is determined to be motion artifact interference; The dynamic threshold adjustment algorithm is executed as follows: when motion artifact interference is detected, starting from the initial safety threshold Th0 = 5 cmH2O / s, the threshold is adjusted according to the formula Th... current = Th0 + 0.5T to calculate the current safety threshold, where T is the duration of the interference in seconds. When |K is detected continuously for 2 seconds... p When all values ​​are below the current safety threshold, the signal is considered to have stabilized. The judgment process of generating the first adjustment command, the second adjustment command, and the drag reduction command will continue only when the deep breathing effort is determined to be effective. If it is determined to be motion artifact interference, or K p With K e If the Pearson correlation coefficient is ≤0.5, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be suspended, and the signal output by the diaphragm electromyography sensor will be marked as needing recalibration; the generation and execution of the emergency resistance reduction command will not be affected by this. When K p In the inspiratory phase, the instantaneous flow rate exceeds 10 cmH2O / s, and K p With K e When the Pearson correlation coefficient is <0.3, it is determined to be choking, and an emergency choking resistance reduction command is generated: reduce the resistance to 5 cmH2O and maintain it for 60 seconds. The emergency choking resistance reduction command is executed immediately, and its priority is the same as the emergency resistance reduction command, and higher than the first adjustment command, the second adjustment command, and the resistance reduction command. If the emergency choking resistance reduction command, the emergency resistance reduction command, or the first adjustment command is generated after the resistance reduction command is generated but before its execution, the resistance reduction command is abandoned, and one operation is executed in the order of the emergency resistance reduction command or the emergency choking resistance reduction command and the first adjustment command.

[0062] Current breathing training devices typically process air pressure and electromyographic signals independently when monitoring respiratory signals, generating adjustment commands based on simple amplitude thresholds. However, users may experience various signal abrupt changes during training, such as genuine deep breathing efforts, artifacts caused by body movement, or sudden coughing. Traditional devices lack the ability to distinguish between these different types of signal abrupt changes, often treating any signal change exceeding the threshold as a valid breathing effort, thus frequently triggering false resistance increase or decrease commands. This false triggering not only disrupts the training rhythm but may also cause discomfort or even risks to the user due to inappropriate resistance adjustment.

[0063] This solution effectively overcomes the limitation of traditional equipment in being unable to distinguish the nature of signal abrupt changes by introducing a signal mutation collaborative verification mechanism. The core of this system lies in synchronously monitoring the slope K of the differential pressure sensor signal change. p Slope K of the diaphragm electromyography sensor signal change e It also calculates the Pearson correlation coefficient between the two in real time, thereby accurately determining the nature of the signal change.

[0064] When K p With K e When a strong positive correlation is observed (Pearson correlation coefficient greater than 0.7), and both values ​​significantly exceed the upper limit of the normal range obtained from the statistical analysis of calm breathing during the calibration phase, the system determines this to be an effective deep breathing effort actively undertaken by the user. In this case, the system continues to execute the normal instruction generation process to ensure that the generated resistance increase or decrease instructions accurately match the user's actual breathing effort, achieving personalized and adaptive training.

[0065] However, in reality, the physiological states of users and the quality of signals are complex and variable. When K p With K e When the Pearson correlation coefficient is as low as ≤0.3, or when the trends of the two are opposite in sign, the system clearly determines that the interference is caused by motion artifacts due to body movement. Once such severe interference is identified, the system immediately suspends the generation of all routine adjustment commands, fundamentally avoiding malfunctions caused by unreliable signals.

[0066] It is worth noting that signal reliability is a continuous spectrum. There may be intermediate states where the signal, while not reaching the severity of "motion artifacts," has a correlation (e.g., a correlation coefficient between 0.3 and 0.5) that is too weak to reliably reflect the user's breathing effort. If relying solely on "motion artifacts" as a criterion, the system will overlook these unreliable signals in the gray area and continue to generate adjustment commands based on them, which also carries the risk of misjudgment. Therefore, the system adds a broader safety barrier: as long as K... p With K e If the Pearson correlation coefficient is ≤0.5, the system will suspend the generation of the first and second adjustment commands and the impedance reduction command, regardless of whether it has been identified as a motion artifact. This is equivalent to setting a higher threshold for signal reliability. Its design philosophy is: when the signal quality is questionable, inaction is preferred over potentially erroneous actions, thereby greatly improving the overall security and robustness of the system.

[0067] To counter explicit motion artifact interference, the system does not passively wait but actively manages the situation using a dynamic threshold adjustment strategy. Starting with an initial safety threshold of 5 cmH2O / s, this threshold gradually widens according to certain rules as the interference duration increases, forming a dynamic safety boundary. Only when the system detects that the signal has remained stable below the current safety threshold for 2 consecutive seconds does it determine that the signal has stabilized and exit the interference response state. This mechanism enables the system to intelligently adapt to interference environments of varying degrees and durations, maintaining its functional integrity even under non-ideal conditions.

[0068] In addition, the system has been specifically optimized for identifying and responding to sudden choking events. When K is detected... pWhen a transient change occurs during the inspiratory phase, with a value exceeding 10 cmH2O / s and a very low correlation coefficient with Ke (less than 0.3), the system immediately identifies it as a coughing event and generates a highest-priority emergency drag reduction command. This command ensures that drag is rapidly reduced to a safe level within a very short time after coughing occurs, greatly reducing the user's breathing burden and discomfort, and effectively ensuring training safety. This command has the highest priority; if it is generated after a drag reduction command but before its execution, the system will automatically abandon the original drag reduction command and immediately respond to the emergency, ensuring the absolute priority of safety logic.

[0069] Example 2 A 60-year-old male patient with chronic obstructive pulmonary disease (COPD) underwent training using a respiratory physiotherapy device with signal mutation co-verification functionality. The patient was 1.75m tall, weighed 70kg, and had an FEV1 / FVC ratio of 56% on pulmonary function tests. The slope K of the diaphragmatic electromyographic signal change was statistically analyzed during the device calibration phase. e The upper limit of the normal range is 1.2 μV / s. During the training process, at the 5-minute mark, the patient adjusted their sitting posture, and the system detected K... p With K e The Pearson correlation coefficient is 0.2. Logically, the system first determines that the situation meets the condition of "correlation coefficient ≤ 0.3", thus classifying it as motion artifact interference. Simultaneously, the result also meets the condition of "correlation coefficient ≤ 0.5", therefore the system suspends the generation of the first adjustment command, the second adjustment command, and the drag reduction command, effectively avoiding malfunctions caused by signal interference.

[0070] At the 10-minute mark, the patient took a deep breath. p Reaching 6.8 cmH2O / s (>5 cmH2O / s), K e The pressure rose to 1.8 μV / s (exceeding the calibration limit of 1.2 μV / s), and the correlation coefficient between the two was 0.8 (>0.7). The system determined this to be an effective deep breathing effort and continued to execute the judgment process of generating the first adjustment command, the second adjustment command, and the drag reduction command. According to the monitored parameters, d△P / dt (5.3 cmH2O / s) did not exceed the threshold of 5 cmH2O / s, but E1 (50 μVs) was lower than 120% of E0 (48 μVs). Therefore, according to the strategy, the second adjustment command was generated, increasing the drag by 3 cmH2O.

[0071] At the 15-minute mark, the patient suddenly coughed up air. p During the inspiratory phase, the flow rate instantaneously rises to 12 cmH2O / s (>10 cmH2O / s), while K... eNo synchronous increase was observed, and the correlation coefficient between the two was only 0.1 (<0.3). The system immediately identified it as a coughing event and generated an emergency resistance reduction command, controlling the airflow resistance regulating valve to immediately reduce the resistance from the current 12 cmH2O to 5 cmH2O and maintain it for 60 seconds. The patient's coughing symptoms were rapidly relieved. The entire training lasted 20 minutes. After the training, the patient's blood oxygen saturation remained at 96%, the subjective fatigue score was 2 points, and the respiratory rate stabilized at 18 breaths / min.

[0072] Comparative Example 5 The same patient was trained using a single conventional device. This device employed a fixed threshold judgment method based on a single air pressure signal. Specifically, when the real-time air pressure change rate dΔP / dt > 5 cmH2O / s, an increase in resistance command was generated; when the instantaneous pressure inside the mask > 15 cmH2O, a decrease in resistance command was generated. This device could not acquire or process electromyographic signals, therefore it could not calculate the signal slope K. p With K e It is also impossible to calculate the correlation coefficient between the two.

[0073] At the 6-minute mark of training, the patient's movement caused violent chest rise and fall, which unexpectedly triggered a sharp fluctuation in the air pressure signal, with the real-time dΔP / dt reaching 5.8 cmH2O / s. Based on its single air pressure threshold logic, the device misinterpreted this as an increase in respiratory effort and generated a resistance increase command, raising the resistance from 10 cmH2O to 14 cmH2O, thus increasing the patient's respiratory load.

[0074] At the 12-minute mark, the patient voluntarily took a slow, deep breath, and his K... e The slope (1.8 μV / s) increased significantly, but the device did not respond to this at all because it could not process electromyographic signals. At the same time, because its inhalation speed was slow, the rate of change of air pressure d△P / dt was only 4.5 cmH2O / s, which did not reach the device's resistance increase threshold (5 cmH2O / s). Therefore, no adjustment command was generated, and the opportunity to increase resistance according to the actual effort level to enhance training was missed.

[0075] At the 18-minute mark, the patient suddenly coughed, causing a rapid increase in pressure within the mask, with the instantaneous dΔP / dt exceeding 10 cmH2O / s. Based on its simplistic logic, the device misinterpreted this drastic pressure change as a respiratory effort requiring increased drag and issued an increase in drag command. This contradicted the actual need for emergency drag reduction, causing the patient's airway resistance to increase during coughing, exacerbating respiratory distress and prolonging the duration of the cough.

[0076] After training, the patient's subjective fatigue score was 6 points, and the respiratory rate was 22 breaths / min. This comparative study indicates that traditional methods relying solely on a single barometric pressure signal and lacking a co-validation mechanism cannot distinguish the true cause of signal mutations, are highly prone to misjudgment and erroneous operation, and have significant safety and effectiveness deficiencies.

[0077] In another embodiment, the control module of the respiratory therapy device is further configured as follows: During the treatment phase, if two or more first or second regulatory commands are generated within three consecutive respiratory cycles, the execution of subsequent first and second regulatory commands shall be suspended, and the resistance shall be maintained unchanged in the next expiratory phase until the respiratory rate recovers to below 20 breaths / min and E1 is within 120%-130% of E0. If an emergency drag reduction command is detected during the drag maintenance period, the emergency drag reduction command shall be executed immediately.

[0078] Current breathing training devices typically make immediate decisions based on real-time monitoring values ​​for each individual respiratory cycle when automatically adjusting resistance. This cycle-by-cycle response mode lacks an overall assessment of adjustment trends. When a user needs the device to continuously increase resistance for various reasons within a short period, the device mechanically executes each resistance increase command. Continuous resistance increases significantly increase the user's respiratory load, easily leading to problems such as excessively rapid breathing rate, disordered breathing patterns, and hyperventilation, posing certain safety risks. The device cannot recognize this abnormal state caused by its own continuous operation, nor can it proactively pause adjustments to allow the user a recovery and adaptation process.

[0079] The respiratory therapy device provided in this solution effectively solves the above problems by introducing trend judgment and intervention mechanisms into the control logic. During the treatment phase, the device continuously monitors the generation of resistance adjustment commands. If the system detects two or more resistance increase commands generated within three consecutive respiratory cycles, it determines that the current situation is in a rapid resistance increase trend that may worsen the user's respiratory condition. At this time, the system will not continue to execute subsequent resistance increase commands, but will actively pause further resistance increase and maintain the resistance at the current level in the next expiratory phase.

[0080] This state of pausing and maintaining resistance will continue until the system detects that the user's breathing has returned to stability. Specific recovery conditions include two aspects: first, the user's respiratory rate drops below 20 breaths / min; second, their current respiratory effort, i.e., the integrated electromyography (EMG) value E1, falls back to the target range of 120%-130% of the baseline value E0. The purpose of this design is to prevent the resistance from increasing indefinitely, providing the user's respiratory muscles with a brief period of adaptation and recovery, preventing respiratory fatigue and pattern incoordination caused by a sudden increase in load, thereby improving the safety and comfort of the training process.

[0081] Furthermore, this mechanism ensures absolute priority for safety protection. Even during periods when the system actively maintains resistance and suspends routine resistance increases, if an emergency indicating rapid breathing is detected, such as an average respiratory rate exceeding 25 breaths / min, the system will immediately generate and execute a highest-priority emergency resistance reduction command. This design allows the device to maintain an immediate response capability to sudden dangerous situations while implementing interventions, achieving an effective combination of proactive intervention and reactive emergency protection.

[0082] In another embodiment, the control module of the respiratory therapy device is further configured as follows: If the E1 value is below 70% or above 130% of E0 for three consecutive respiratory cycles during the treatment phase, a transition phase lasting at least five respiratory cycles will be initiated immediately. Or, after executing the emergency decompression command or exiting the resistance maintenance period, enter a transition phase lasting at least 5 respiratory cycles; During the transition phase, the generation of the first adjustment command, the second adjustment command, and the drag reduction command will be suspended; the generation and execution of the emergency drag reduction command will not be affected by the transition phase. During the transition phase, the integral value of electromyography (EMG) for each complete respiratory cycle is calculated in real time. At the end of the transition phase, the arithmetic mean of all calculated EMG integral values ​​is taken and set as Em. 0过渡 And this is used to temporarily replace the electromyographic integral baseline value E0; After the transition phase ends, a gradual change process for the electromyographic integral baseline value is initiated; During the gradual change of the electromyographic integral baseline value, at the beginning of each subsequent respiratory cycle, the electromyographic integral baseline value used for comparison is changed from E... 0过渡 Adjust the value to the original electromyography integral baseline value E0 once until the difference between the two is less than the preset threshold or reaches the original electromyography integral baseline value E0. During the gradual change of the electromyographic integral baseline value, the following instruction generation logic is resumed: monitor the rate of increase of differential pressure d△P / dt and calculate the current electromyographic integral value E1, and determine whether to generate the first adjustment instruction, the second adjustment instruction, or the resistance reduction instruction based on this.

[0083] Existing breathing training devices typically use a fixed baseline value to measure the user's breathing effort throughout the training process. This static baseline setting method has significant drawbacks because the user's physiological state, electromyographic signal characteristics, and breathing effort level can drift with changes in training progress, fatigue level, or attention. If the user's electromyographic integral value continuously deviates from the initially calibrated baseline range due to improved condition or fatigue, and the device continues to judge based on the fixed old baseline, it will lead to complete inaccurate resistance adjustment. Either the resistance will be continuously increased, causing overload, or the resistance will be continuously decreased, resulting in undertraining, failing to achieve effective and safe personalized rehabilitation training.

[0084] The respiratory therapy device provided in this solution effectively solves the above problems by introducing a dynamic benchmark management mechanism. When the system detects that the user's current respiratory effort level, i.e., the integrated electromyography (EMG) value E1, deviates significantly from the initial benchmark value E0 for three consecutive respiratory cycles (e.g., below 70% or above 130%), the system determines that the user's state has undergone a substantial change, and the original benchmark value E0 is no longer applicable. At this time, the system will not continue to use the invalid benchmark for misjudgment, but will immediately initiate a transition phase. During the transition phase, the system suspends all routine resistance adjustments based on effort level, providing only a stable breathing environment for the user and collecting new respiratory data. The system will calculate the integrated EMG value for each complete respiratory cycle during this phase in real time, and at the end of the transition phase, set the arithmetic mean of these newly collected values ​​as a temporary new benchmark value that better reflects the user's current state, denoted as E0 transition, and temporarily replace the original E0 with this value.

[0085] After the transition phase, the system does not immediately switch completely to the new baseline. Instead, it initiates a gradual adjustment process called the electromyographic integral baseline gradient process. During this process, at the beginning of each subsequent respiratory cycle, the system gradually transitions the baseline value used for comparison from E0 back to the original E0 by a small, preset step size or proportion. This process is slow and smooth, designed to allow the user a gradual adaptation period, avoiding drastic fluctuations in resistance settings caused by sudden baseline switching. During this gradient process, the system resumes its normal instruction generation logic, making resistance adjustment decisions based on the dynamically adjusted baseline value, thereby ensuring the continuity and effectiveness of training. The entire mechanism aims to enable the baseline value to intelligently and smoothly evolve with changes in the user's state, maintaining the accuracy and adaptability of resistance adjustment at all times, effectively improving the long-term training effect of the equipment and the user experience.

[0086] Example 3 A 58-year-old female patient with respiratory muscle weakness due to interstitial lung disease underwent training using this respiratory physiotherapy device. The initial calibration electromyographic integral (E0) baseline value was 35 μVs. After approximately 10 minutes of training, the patient gradually became fatigued, and her respiratory effort continuously decreased. The system monitored E1 values ​​for three consecutive respiratory cycles of 22 μVs, 24 μVs, and 23 μVs (all below 70% of E0). The system immediately initiated a transition phase, pausing routine adjustment commands and collecting new E1 values ​​(25 μVs, 26 μVs, 24 μVs, 25 μVs, 25 μVs) over the next five respiratory cycles, calculating the E1 value. 0过渡 The initial E0 was 25 μVs. The system then initiated a gradual regression process, returning to the original E0 (35 μVs) in steps of 1 μVs per respiratory cycle. After three cycles, the baseline for comparison was adjusted to 28 μVs. At this point, the patient's E1 value rose to 27 μVs (96.4% of the new baseline), and the system determined that the effort level had returned to the normal range, allowing training to continue safely and effectively. Throughout the process, the patient's blood oxygen saturation remained above 95%, with no discomfort.

[0087] Comparative Example 6 The same patient was trained using a conventional fixed-benchmark respiratory trainer, with an initial calibration baseline E0 of 35 μVs. After approximately 10 minutes of training, the patient experienced fatigue, and the E1 value continuously decreased to around 23 μVs. However, because the device consistently used a fixed baseline of 35 μVs for judgment, the system continuously detected E1 < 120% E0 (i.e., 42 μVs), thus erroneously generating continuous resistance increase commands, causing the resistance to climb from 8 cmH2O to 18 cmH2O within 4 minutes. This excessively high resistance load further exacerbated the patient's dyspnea and fatigue, with the respiratory rate rising to 28 breaths / min and oxygen saturation dropping to 92%, forcing the training to be interrupted at 14 minutes. After training, the patient felt extremely fatigued, and the maximum inspiratory pressure did not improve.

[0088] Results: The device in Example 3, through its dynamic benchmark management mechanism, accurately identified significant changes in the user's condition and smoothly adjusted the evaluation criteria through transition and gradual changes, ensuring that the training load always matched the patient's actual ability, thus guaranteeing the safety and effectiveness of the training. In contrast, the fixed benchmark device in Comparative Example 6 could not adapt to changes in the user's condition; its mechanical adjustment logic ultimately led to a severe disconnect between the resistance setting and the patient's actual ability, resulting not only in ineffective training but also safety issues. This demonstrates that dynamic benchmark management is crucial for achieving long-term, safe, and personalized respiratory rehabilitation.

[0089] In another embodiment, the control module of the respiratory therapy device is further configured to: During the gradual change of the electromyographic integral reference value, the control module is also configured to: While calculating the current electromyographic integral value E1 in each respiratory cycle, the short-term rate of change δ of the current electromyographic integral value E1 relative to the electromyographic integral value of the previous respiratory cycle is also calculated. If the absolute value of the short-term change rate δ of the electromyography integral value for two consecutive respiratory cycles is greater than the preset change rate threshold (e.g., |δ|>20%), the gradual change process of the electromyography integral baseline value is immediately stopped, and the current electromyography integral baseline value used for comparison is locked as the current electromyography integral value E1, which is used as the new electromyography integral baseline value. The generation and execution of emergency drag reduction commands are not affected by the abort of the gradual change process.

[0090] Existing breathing training devices typically use fixed time periods or adjustment steps when adjusting baseline values, lacking sensitivity to changes in the user's real-time state. When the user's breathing effort changes suddenly and drastically during the gradual adjustment of the baseline value due to external stimuli, emotional fluctuations, or physiological changes, this pre-set, linear gradual adjustment mechanism cannot promptly identify such abrupt changes. The system continues to slowly adjust the baseline value according to the original plan, resulting in a severe disconnect between the adjustment process and the user's actual state. This lag causes the system to continue using an intermediate baseline value that is no longer representative for a period of time, potentially generating a series of inappropriate resistance adjustment commands and even exacerbating the user's breathing disturbances.

[0091] The respiratory therapy device provided by this invention effectively solves the above-mentioned problems by introducing a short-term rate of change monitoring and dynamic termination mechanism during the gradual change process. In each respiratory cycle during the gradual change of the electromyography (EMG) integral baseline value, the system not only calculates the current EMG integral value E1, but also simultaneously calculates the short-term rate of change δ of this value relative to the EMG integral value of the previous respiratory cycle. This rate of change δ is a key indicator, reflecting in real time the instantaneous fluctuation speed and direction of the user's respiratory effort.

[0092] The system sets a preset threshold for the short-term rate of change δ. When the system detects that the absolute value of the short-term rate of change δ exceeds this preset threshold for two consecutive respiratory cycles, it immediately determines that the user's breathing effort level is undergoing a drastic and continuous change. At this point, the system will not mechanically continue to execute the original, slow gradual change plan, but will immediately terminate the entire gradual change process. The purpose is to quickly terminate an adjustment strategy that is no longer consistent with the actual situation, avoiding the continuous generation of erroneous control decisions due to adjustment lag.

[0093] After the gradual change process is aborted, the system performs a crucial operation: it locks the current electromyographic integral (EMG) baseline value used for comparison to the newly calculated EMG integral value E1, and directly sets this value as the new, official EMG baseline value. This step instantly synchronizes the system's evaluation baseline with the user's latest, drastic state change, achieving a "hard reset" of the baseline value. This ensures that the system can immediately assess the subsequent respiratory effort using the baseline value best suited to the current state, thereby ensuring the accuracy and timeliness of any subsequent adjustment commands. This significantly improves the system's rapid adaptability and overall control precision when the user's state is unstable. The entire mechanism ensures absolute priority for safety protection; the generation and execution of emergency resistance reduction commands are completely unaffected by this abort process.

[0094] Although the embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details.

Claims

1. A respiratory physiotherapy device, characterized in that, include: The mouth and nose mask is equipped with a one-way air inlet and a one-way air outlet; An airflow resistance regulating valve is located at the front end of the one-way air inlet; The diaphragmatic electromyography sensor contains two surface electrodes and is attached 3-5 cm below the xiphoid process of the user. A differential pressure sensor is installed on the mouth and nose mask, with its first and second pressure taps connected to the inside of the mask and the outside atmosphere, respectively. The control module is configured as follows: During the calibration phase, when the user takes multiple calm breaths, the average value of the diaphragm electromyography integral value in each inspiratory phase is calculated based on the signal output by the diaphragm electromyography sensor and set as the electromyography integral reference value E0. During each inspiratory phase of the treatment, the rate of change of pressure difference d△P / dt is monitored by a differential pressure sensor, and the current electromyographic integral value E1 is calculated. If d△P / dt>5cmH2O / s, the first adjustment command is generated: control the airflow resistance regulating valve to increase the resistance by 3-5cmH2O; If E1 < 120%E0, a second adjustment command is generated: increase the resistance by 2-4 cmH2O; If E1 > 150%E0, a resistance reduction instruction is generated: after reducing the resistance by 4-6 cmH2O in the subsequent expiratory phase, the resistance reduction instruction is cleared. The respiratory rate is calculated based on the signal output by the differential pressure sensor. If the average respiratory rate of multiple consecutive respiratory cycles is >25 breaths / min, an emergency resistance reduction command is generated: reduce the resistance to 5 cmH2O and maintain it for 60 seconds. If the first adjustment command, the second adjustment command, and the emergency drag reduction command are generated simultaneously, then one operation is executed in the order of the emergency drag reduction command, the first adjustment command, and the second adjustment command. If a drag reduction instruction is generated during the execution of any instruction, the drag reduction instruction is ignored; if an emergency drag reduction instruction or a first adjustment instruction is generated after the drag reduction instruction is generated but before its execution, the drag reduction instruction is abandoned, and one operation is executed in the order of the emergency drag reduction instruction and the first adjustment instruction.

2. The respiratory therapy device as described in claim 1, characterized in that, The control module is also configured as follows: After the respiratory rate is calculated based on the signal output by the differential pressure sensor during the calibration phase, the fluctuation of the respiratory rate is judged by calculating the range or standard deviation of the frequency over 3-5 consecutive respiratory cycles. The control module calculates the peak flow rate of the intake phase based on data monitored by the differential pressure sensor; The control module records a breath as a valid breath if it simultaneously meets the conditions that the respiratory rate fluctuation is less than the preset frequency fluctuation threshold and the peak inspiratory flow is within the preset flow range, and counts the number of valid breaths. The control module completes the calculation of the electromyographic integral baseline value during the calibration phase, and the following conditions must be met simultaneously: The respiratory rate fluctuation is less than the preset frequency fluctuation threshold for 3-5 consecutive respiratory cycles; The peak inspiratory flow rate during each inspiratory phase is within the preset flow rate range for 3-5 consecutive respiratory cycles; And the number of effective breaths collected reached the preset minimum number; Only when all of the above conditions are met will the control module set the average value of the electromyographic integral of this group of effective breathing as the electromyographic integral reference value E0.

3. The respiratory therapy device as described in claim 1, characterized in that, Before calculating the current electromyographic integral value E1 during the treatment phase, the control module filters the raw signal collected by the diaphragm electromyographic sensor. The filter passband frequency range is 10-500Hz.

4. The respiratory therapy device as described in claim 1, characterized in that, The control module is also configured as follows: During the treatment phase, if the current electromyographic integral value E1 remains below 80% of the electromyographic integral benchmark value E0 for several respiratory cycles, an alarm signal is generated and the execution of the first regulation command, the second regulation command, and the resistance reduction command is suspended; the generation and execution of the emergency resistance reduction command are not affected.

5. The respiratory therapy device as described in claim 1, characterized in that, The control module is also configured as follows: Perform signal quality monitoring during the treatment phase; Signal quality monitoring includes analyzing the raw signals acquired by the diaphragm electromyography sensor. The analysis methods include at least one of calculating the signal amplitude, assessing the noise level, or detecting abnormal waveforms. If the analysis results indicate that the signal quality is lower than the preset quality threshold for multiple consecutive respiratory cycles, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be paused, and a signal abnormality warning will be generated. If the signal amplitude output by the diaphragm electromyography sensor is lower than the preset minimum threshold for 5 consecutive seconds, an electrode detachment warning is generated, and the system switches to pure pneumatic feedback regulation mode. In pure pneumatic feedback regulation mode, the first regulation command and emergency resistance reduction command are generated only based on the signal output by the differential pressure sensor.

6. The respiratory therapy device as described in claim 1, characterized in that, Before generating the first adjustment command, second adjustment command, or resistance reduction command during the treatment phase, the control module also performs signal mutation co-verification: synchronously monitoring the slope K of the differential pressure sensor signal change. p Slope K of the diaphragm electromyography sensor signal change e ; When K p With K e The Pearson correlation coefficient is >0.7, and |K p |>5cmH2O / s, while K e >When the slope of the diaphragm electromyography sensor signal change obtained from the statistics of calm breathing during the calibration phase reaches the upper limit of the normal range, it is determined to be an effective deep breathing effort; When K p With K e The Pearson correlation coefficient is ≤0.3, or K p With K e When the signs of the positive and negative symbols are opposite, it is determined to be motion artifact interference; The dynamic threshold adjustment algorithm is executed as follows: when motion artifact interference is detected, starting from the initial safety threshold Th0 = 5cmH2O / s, the threshold is adjusted according to the formula Th... current = Th0 + 0.5T to calculate the current safety threshold, where T is the duration of the interference in seconds. When |K is detected continuously for 2 seconds... p When all values ​​are below the current safety threshold, the signal is considered to have stabilized. The judgment process of generating the first adjustment command, the second adjustment command, and the drag reduction command will continue only when the deep breathing effort is determined to be effective. If it is determined to be motion artifact interference, or K p With K e If the Pearson correlation coefficient is ≤0.5, the generation of the first adjustment command, the second adjustment command, and the resistance reduction command will be suspended, and the signal output by the diaphragm electromyography sensor will be marked as needing recalibration; the generation and execution of the emergency resistance reduction command will not be affected by this. When K p In the inspiratory phase, the instantaneous flow rate exceeds 10 cmH2O / s, and K p With K e When the Pearson correlation coefficient is <0.3, it is determined to be choking, and an emergency choking resistance reduction command is generated: reduce the resistance to 5cmH2O and maintain it for 60s. The emergency choking resistance reduction command is executed immediately and has the same priority as the emergency resistance reduction command.

7. The respiratory therapy device as described in claim 1, characterized in that, The control module is also configured as follows: During the treatment phase, if two or more first or second regulatory commands are generated within three consecutive respiratory cycles, the execution of subsequent first and second regulatory commands shall be suspended, and the resistance shall be maintained unchanged in the next expiratory phase until the respiratory rate recovers to below 20 breaths / min and E1 is within 120%-130% of E0. If an emergency drag reduction command is detected during the drag maintenance period, the emergency drag reduction command shall be executed immediately.

8. The respiratory therapy device as described in claim 1, characterized in that, The control module is also configured as follows: If the E1 value is below 70% or above 130% of E0 for three consecutive respiratory cycles during the treatment phase, a transition phase lasting at least five respiratory cycles will be initiated immediately. Or, after executing the emergency decompression command or exiting the resistance maintenance period, enter a transition phase lasting at least 5 respiratory cycles; During the transition phase, the generation of the first adjustment command, the second adjustment command, and the drag reduction command will be suspended; the generation and execution of the emergency drag reduction command will not be affected by the transition phase. During the transition phase, the integral value of electromyography (EMG) for each complete respiratory cycle is calculated in real time. At the end of the transition phase, the arithmetic mean of all calculated EMG integral values ​​is taken and set as Em. 0过渡 And this is used to temporarily replace the electromyographic integral baseline value E0; After the transition phase ends, a gradual change process for the electromyographic integral baseline value is initiated; During the gradual change of the electromyographic integral baseline value, at the beginning of each subsequent respiratory cycle, the electromyographic integral baseline value used for comparison is changed from E... 0过渡 Adjust the value to the original electromyography integral baseline value E0 once until the difference between the two is less than the preset threshold or reaches the original electromyography integral baseline value E0. During the gradual change of the electromyographic integral baseline value, the following instruction generation logic is resumed: monitor the rate of increase of differential pressure d△P / dt and calculate the current electromyographic integral value E1, and determine whether to generate the first adjustment instruction, the second adjustment instruction, or the resistance reduction instruction based on this.

9. The respiratory therapy device as described in claim 8, characterized in that, The control module is also configured to: While calculating the current electromyographic integral value E1 in each respiratory cycle, the short-term rate of change δ of the current electromyographic integral value E1 relative to the electromyographic integral value of the previous respiratory cycle is also calculated. If the absolute value of the short-term change rate δ of the electromyography integral value for two consecutive respiratory cycles is greater than the preset change rate threshold, the gradual change process of the electromyography integral benchmark value is immediately stopped, and the current electromyography integral benchmark value used for comparison is locked as the current electromyography integral value E1, which is used as the new electromyography integral benchmark value. The generation and execution of emergency drag reduction commands are not affected by the abort of the gradual change process.

Images