Wireless respiration monitoring system and method for clinical treatment

Abstract

The invention provides a wireless respiration monitoring system and method for clinical treatment, and relates to the technical field of respiration test medical treatment, and the method comprises the following steps: collecting respiratory muscle function physiological parameters of a clinical patient; establishing a mathematical evaluation model and a large sample healthy population database, and mapping and outputting a recovery evaluation initial value of the respiratory muscle function; interference item parameters of a clinical patient are collected, and an evaluation correction coefficient is obtained; and adjusting a recovery evaluation initial value of the respiratory muscle function output by the mathematical evaluation model to generate a recovery evaluation correction value. The breathing state quality of the patient after the clinical treatment stage can be predicted in advance through the mathematical evaluation model. In addition, through the collected interference item parameters, the initial value of the recovery evaluation can be calibrated, so that the finally output correction value of the recovery evaluation can more accurately predict the breathing state quality of the patient after the clinical treatment stage. And the accuracy is higher.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of respiratory test medical technology, in particular to a wireless respiratory monitoring system and method for clinical treatment. BACKGROUND

[0002] In the process of respiratory monitoring in clinical treatment, monitoring the respiratory muscle function is of vital significance, which directly reflects the state of the power source driving ventilation, can early identify potential risks such as respiratory muscle fatigue, weakness or coordination disorders, and provides key basis for evaluating ventilation function reserve, predicting offline success rate, guiding respiratory rehabilitation treatment, and avoiding acute respiratory failure, which is the core link to realize precise respiratory support and improve patient clinical outcomes. The respiratory muscle function can be reflected in chest and abdominal movement amplitude, blood oxygen saturation, respiratory flow rate change amplitude, and respiratory depth index, and for different treatment schemes, the recovery of respiratory muscle function will also be different.

[0003] In the prior art, the publication number CN114533034A discloses a respiratory monitoring method, a respiratory monitoring system, a monitoring device and a medical central station system, which comprises: acquiring clinical data, and respiratory data and monitoring data collected by the respiratory support equipment and the monitoring equipment; and displaying the clinical data, the respiratory data and the monitoring data on the respiratory management interface at the same time. In this way, medical staff can quickly view the clinical data, the respiratory data and the monitoring data on the same respiratory management interface, thereby reducing the work intensity and difficulty of medical staff, and improving the safety and effectiveness of respiratory monitoring, and reducing the risk of patient condition delay.

[0004] However, the conventional respiratory monitoring method in the above prior art can only provide real-time feedback of basic parameters such as respiratory rate and blood oxygen saturation to judge the current treatment state, and cannot judge the recovery state after treatment in combination with the patient's own condition and the specific treatment scheme, lacks deep fusion and quality control of multi-modal physiological signals, and cannot effectively eliminate motion artifacts and individual difference interference; at the same time, since the neural network evaluation model based on health big data and the correction mechanism of clinical interference factors such as metabolism, nutrition and medication are not introduced, the evaluation result is often limited to the current state description, it is difficult to realize accurate prediction of respiratory function recovery trend, and it is also difficult to provide layered and personalized treatment decision support, thereby there are significant limitations in clinical prognosis judgment and intervention strategy optimization.

[0005] The above information disclosed in the BACKGROUND section is only for the purpose of enhancing the understanding of the background of the present disclosure, and therefore it can include information that does not constitute the prior art known to those of ordinary skill in the art. SUMMARY

[0006] The purpose of the present application is to provide a wireless respiratory monitoring system and method for clinical treatment to solve the problems raised in the background art. The present application can predict the quality of the respiratory state of the patient after completing the clinical treatment stage in advance through a mathematical evaluation model. In addition, by collecting the interference term parameters, the initial value of the recovery evaluation can be calibrated, so that the final output of the recovery evaluation correction value can more accurately predict the quality of the respiratory state of the patient after completing the clinical treatment stage. Higher accuracy.

[0007] To achieve the above purpose, the present application provides the following technical solutions:

[0008] The wireless respiratory monitoring method for clinical treatment comprises the following steps:

[0009] S1: Collect the respiratory muscle function physiological parameters Y of the clinical patient, the respiratory muscle function physiological parameters Y including chest and abdominal movement amplitude, blood oxygen saturation, respiratory flow rate change amplitude and respiratory depth index, the sampling frequency being in the range of 50-100Hz for capturing details, providing high-precision and low-noise original input signals through Bluetooth wireless transmission;

[0010] S2: Establish a mathematical evaluation model and a large sample healthy population database, the model using a neural network learning framework and being trained and calibrated based on the large sample healthy population database, defining the reference range of patient age and gender stratification, extracting features and normalizing the respiratory muscle function physiological parameters collected in step S1 to optimize the input quality, and then inputting into the mathematical evaluation model to map the output of the recovery evaluation initial value of the respiratory muscle function, the recovery evaluation initial value being used to preliminarily predict the quality of the respiratory state of the patient after completing the clinical treatment stage;

[0011] S3: Collect the interference term parameters of the clinical patient, the interference term parameters including body mass index BMI standard deviation value, serum albumin level value and recent medication record interval, establish an evaluation value correction grading model, and obtain the evaluation correction coefficient through a correction formula after integrating the interference term parameters;

[0012] S4: Based on the evaluation correction coefficient obtained in step S3, adjust the recovery evaluation initial value of the respiratory muscle function output by the mathematical evaluation model to generate a recovery evaluation correction value, the recovery evaluation correction value being used to finally predict the quality of the respiratory state of the patient after completing the clinical treatment stage, and finally output the respiratory muscle function determination result of the measured patient through the recovery evaluation correction value.

[0013] Furthermore, in step S1, the amplitude of chest and abdominal movement is used to reflect the displacement change of respiratory muscle contraction and relaxation; the blood oxygen saturation is obtained by real-time monitoring of the percentage value of blood oxygenation status and used to evaluate gas exchange efficiency; the amplitude of respiratory flow change is generated after continuously recording the time sequence signal of airflow rate to reflect the flow rate during inhalation and exhalation; the respiratory depth index is used to quantify the volume change of each breath.

[0014] Normalize the functional physiological parameter Y:

[0015]

[0016] in:

[0017] The collected physiological parameters of the patient's respiratory muscle function;

[0018] These are the normalized physiological parameters of respiratory muscle function;

[0019] The minimum values ​​of physiological parameters of healthy respiratory muscle function in various population groups; The maximum values ​​of physiological parameters of healthy respiratory muscle function in various population groups;

[0020] The original input signal is the received original multi-channel physiological data stream, and after receiving the data stream, real-time quality control is implemented: dynamic range verification is performed on the chest and abdominal movement amplitude signal, and abnormal peaks exceeding the physiological limits of the human body are automatically eliminated; a bandpass filter of 0.1-20Hz is simultaneously activated to eliminate high-frequency electromyographic interference and low-frequency body movement noise.

[0021] Furthermore, the high-frequency electromyography interference is a signal greater than 20Hz, and the low-frequency body movement noise is a signal less than 0.1Hz. Then, baseline drift correction is performed on the blood oxygen saturation data: the trend term is separated by sliding window mid-range filtering to eliminate the slow drift caused by limb micro-movements; the change amplitude of respiratory flow is compensated by phase alignment, and the millisecond-level transmission delay between chest and abdominal movement and airflow signal is eliminated by cross-correlation algorithm.

[0022] In the cross-correlation algorithm, a regression model is constructed by extracting acceleration features to separate non-respiratory displacement components from chest and abdominal motion signals. Finally, invalid data segments are marked by adaptive threshold segmentation, including coughing or body position reversal periods. Timestamp synchronization calibration is performed on the data of each channel to control the alignment of multimodal data within ±10ms error and output a standardized signal matrix that meets the analysis requirements.

[0023] Furthermore, the initial value for the recovery assessment is calculated using the following formula:

[0024]

[0025] in:

[0026] To restore the initial assessment value, the range is between [0,1], representing the preliminary predictive quality of respiratory muscle function after the patient's recovery;

[0027] The normalized blood oxygen saturation was obtained by feature extraction using the average value of the time series data.

[0028] The normalized amplitude of chest and abdominal movements is obtained by calculating the feature extraction method of the average amplitude of collected chest and abdominal movements.

[0029] The normalized variation in respiratory flow rate is obtained by calculating the feature extraction method over a fixed time period.

[0030] The normalized respiratory depth index is obtained by collecting tidal volume.

[0031] The weighting coefficients for blood oxygen saturation, chest and abdominal movement amplitude, respiratory flow rate change amplitude, and respiratory depth are dynamically optimized through neural network training. All of these are used to reflect the degree of influence of each respiratory muscle function physiological parameter on the recovery of respiratory muscle function.

[0032] and The following relationship must be satisfied: During the training process of the neural network, the neural network learns weights using the health database, so that the calculation formula for the initial value of the recovery assessment statistically minimizes the prediction error, and avoids overfitting through cross-validation calibration.

[0033] Furthermore, the evaluation correction coefficient in step S3 is calculated using the following formula:

[0034]

[0035] in:

[0036] To evaluate the correction coefficient, it is used to compare the initial value with the restoration assessment. To perform correction, when Time indicates positive correction, and the initial value is used to assess recovery. Enlarge it when Time indicates negative correction, which is used to assess the initial value for recovery. Shrink, when Time indicates restoration of the initial assessment value No correction is needed;

[0037] BMI standard deviation;

[0038] This refers to serum albumin levels.

[0039] The interval for recent medication records, and through Make The smaller, The smaller the value, the better for the initial value of the recovery assessment. Perform negative correction. When it is larger, Logarithmic growth is slowing, affecting the initial value of the recovery assessment. The degree of negative correction decreases;

[0040] The weighting coefficients are the standard deviation of body mass index (BMI), serum albumin level, and the interval between recent medication records. All are positive real numbers, and satisfy the following conditions: .

[0041] Furthermore, the standard deviation of body mass index (BMI) is calculated by z-score deviation between the measured BMI and the standard value of healthy individuals of the same age and sex, reflecting the degree of metabolic load abnormality; the serum albumin level is uniformly calibrated to the same detection standard to eliminate systematic errors; during the collection of recent medication records, the average time interval of the patient's three most recent respiratory-related drug uses is extracted, the unit is converted to hours, and the natural logarithm is taken to weaken the marginal effect of long intervals.

[0042] Furthermore, the interference parameters are standardized and mapped. The standardization mapping includes mapping the absolute value of the standard deviation of body mass index (BMI) to the interval [0,3], with a cutoff when |B|≥3; converting the serum albumin level value into a linear decay coefficient based on the clinical threshold of 35g / L; and processing the recent medication record interval through a piecewise function. Finally, the three interference parameters are integrated into the evaluation value correction grading model, and the evaluation correction coefficient is generated through a dynamic weighted formula. The serum albumin level value directly contributes negatively, while the standard deviation of body mass index (BMI) and the recent medication record interval provide positive or negative contributions. The evaluation value correction grading model is validated by a clinical expert rule base before output to ensure that the correction logic conforms to the pathophysiological mechanism of respiratory function compensation.

[0043] Furthermore, the recovery assessment correction value is calculated using the following formula:

[0044]

[0045] in To restore the assessment correction value, which serves as a quantitative indicator for ultimately predicting the respiratory status quality of patients after the completion of clinical treatment, a preset risk threshold rule was then applied.

[0046] Furthermore, in step S4, a healthy threshold for assessing the recovery of respiratory muscle function is set as follows: The patient's respiratory muscle function assessment results are generated using the following grading system:

[0047] when At that time, the respiratory muscle function was determined to be well compensated, and the respiratory status quality was predicted to reach that of healthy people after treatment. Under this result, the current treatment plan was maintained and the test was repeated weekly.

[0048] when At that time, it was determined to be mild functional inhibition. Under this result, nutritional support was optimized and respiratory muscle strengthening training and medication were initiated, and the monitoring period was shortened to 72 hours.

[0049] when At that time, the patient was determined to be at high risk of respiratory failure. Based on this result, the patient was transferred to the intensive care unit and put into mechanical ventilation. At the same time, the model input parameters were recalibrated.

[0050] A wireless respiratory monitoring system for clinical treatment, the respiratory monitoring system being used to perform the above-described respiratory monitoring method, the respiratory monitoring system comprising:

[0051] The data acquisition module is used to collect respiratory muscle function physiological parameters Y of clinical patients. The respiratory muscle function physiological parameters Y include the amplitude of chest and abdominal movement, blood oxygen saturation, the amplitude of respiratory flow change, and respiratory depth index. The sampling frequency is in the range of 50-100Hz to capture details. It provides high-precision and low-noise raw input signals through Bluetooth wireless transmission.

[0052] The model training and evaluation module is used to establish a mathematical evaluation model and a large sample database of healthy people. The model adopts a neural network learning framework and is trained and calibrated based on the large sample database of healthy people. The reference range for patient age and gender stratification is defined. The physiological parameters of respiratory muscle function collected in step S1 are subjected to feature extraction and normalization to optimize the input quality. Then, the input is fed into the mathematical evaluation model to map and output the initial value of respiratory muscle function recovery evaluation. The initial value of recovery evaluation is used to preliminarily predict the respiratory status quality of patients after completing the clinical treatment stage.

[0053] The correction generation module is used to collect interference parameters from clinical patients. The interference parameters include the standard deviation of body mass index (BMI), serum albumin level, and recent medication record interval. An assessment value correction grading model is established, and the assessment correction coefficient is obtained by integrating the interference parameters through the correction formula.

[0054] The hierarchical decision output module and the evaluation correction coefficient obtained from the correction generation module adjust the initial value of the recovery assessment of respiratory muscle function output by the mathematical evaluation model to generate a recovery assessment correction value. The recovery assessment correction value is used to ultimately predict the respiratory status quality of the patient after completing the clinical treatment stage. Finally, the respiratory muscle function judgment result of the tested patient is output through the recovery assessment correction value.

[0055] Compared with the prior art, the beneficial effects of the present invention are:

[0056] By constructing a mathematical assessment model based on neural network learning, early prediction of the respiratory status quality of patients after completing clinical treatment can be achieved based on real-time collected multi-dimensional respiratory physiological parameters, breaking through the limitations of traditional respiratory monitoring which is limited to real-time status feedback. Furthermore, by introducing multi-source interference parameters including metabolism, nutrition, and medication adherence, a graded correction model for assessment values ​​is established to individually correct initial predicted values, significantly improving the accuracy and clinical applicability of respiratory function recovery assessment. Attached Figure Description

[0057] Fig. 1 This is a schematic flowchart of the wireless respiratory monitoring method for clinical treatment according to the present invention;

[0058] Fig. 2 This is a block diagram illustrating the principle of a wireless respiratory monitoring system used in clinical treatment. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0060] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0061] Example:

[0062] Please see Figs. 1-2 The present invention provides the following technical solutions:

[0063] A wireless respiratory monitoring method for clinical treatment includes the following steps:

[0064] S1: During clinical treatment, the system uses advanced sensor technology to collect the patient's respiratory muscle function physiological parameters Y. These parameters Y specifically include the amplitude of chest and abdominal movements, blood oxygen saturation, changes in respiratory flow rate, and respiratory depth. To ensure the capture of subtle dynamic details during respiration, the sampling frequency is precisely set within the high-frequency range of 50 to 100 Hz. All collected data is transmitted in real-time via Bluetooth wireless transmission technology. This transmission method not only improves the convenience of clinical operation but also ensures the high precision of the original input signal while effectively maintaining a low noise level, thus providing a stable and reliable signal foundation for subsequent analysis.

[0065] The amplitude of chest and abdominal movements meticulously captures the displacement changes of respiratory muscles during contraction and relaxation, reflecting the dynamic physical characteristics of muscle activity; the blood oxygen saturation is obtained by continuously monitoring the precise percentage of oxygenation in the blood in real time, effectively assessing the overall efficiency of gas exchange in the patient's body; after continuously recording the time-series signal of airflow rate formation, the system generates the amplitude of respiratory flow rate change, which accurately reflects the actual flow dynamics of airflow during inspiration and expiration; the respiratory depth index is specifically used to quantify the specific volume changes that occur during each breath, providing an objective measurement basis for respiratory depth;

[0066] Normalize the functional physiological parameter Y:

[0067]

[0068] in:

[0069] The collected physiological parameters of the patient's respiratory muscle function; These are the normalized physiological parameters of respiratory muscle function;

[0070] Set truncation rules: If ,but ,like ,but .

[0071] The minimum values ​​of physiological parameters of healthy respiratory muscle function in various population groups; These represent the maximum values ​​of physiological parameters for healthy respiratory muscle function in various population groups. and These values ​​are derived from statistics based on a large sample database of healthy people, defining the "health reference range" for the parameters;

[0072] The raw input signal is essentially the received, unprocessed, raw multi-channel physiological data stream. Once these data streams are successfully received, the system immediately implements a comprehensive real-time quality control procedure: specifically, it performs a precise dynamic range verification process on the chest and abdominal movement amplitude signal, automatically identifying and eliminating abnormal peak signals that exceed the normal physiological limits of the human body; at the same time, the system synchronously activates a bandpass filter in the frequency range of 0.1 to 20 Hz to efficiently eliminate high-frequency electromyographic interference and low-frequency body movement noise, ensuring that the signal is pure and reliable.

[0073] The high-frequency electromyographic interference is essentially a signal component with a frequency higher than 20 Hz, mainly originating from background noise generated by muscle electrical activity; the low-frequency body motion noise refers to signal interference with a frequency lower than 0.1 Hz, usually caused by slow fluctuations due to minor changes in the patient's body position. Subsequently, the system performs baseline drift correction on the blood oxygen saturation data: specifically, through a sliding window median filtering technique, the median is calculated within a continuous time window to intelligently separate the slow trend term of the signal, thereby accurately eliminating the slow baseline drift caused by subtle limb movements and ensuring data stability and reliability. Simultaneously, meticulous phase alignment compensation is implemented for the amplitude of respiratory flow changes: a cross-correlation algorithm is used to accurately detect the time offset relationship between chest and abdominal motion signals and airflow signals, and a dynamic compensation mechanism eliminates millisecond-level transmission delays, ensuring signal synchronization and improving the accuracy of subsequent analysis.

[0074] During the execution of the cross-correlation algorithm, the system first extracts the dynamic acceleration features from the chest and abdominal motion signals. Then, based on these features, a regression model is constructed. This model can intelligently identify and effectively separate displacement component interference introduced by non-respiratory activities (such as daily walking or slight body swaying). Next, an adaptive threshold segmentation technique is used to automatically mark invalid data segments. These invalid segments typically include periods of sudden coughing or body reversal. This segmentation method dynamically adjusts the threshold to sensitively capture abnormal events. Finally, the system performs a strict timestamp synchronization calibration process on all channel data, precisely controlling the synchronization and alignment of multimodal physiological data such as chest and abdominal motion and airflow rate within a small error range of ±10 milliseconds. This results in the output of a highly consistent and optimized standardized signal matrix, which fully meets the accuracy and reliability requirements of subsequent respiratory function analysis.

[0075] S2: During the clinical assessment process, the system first constructs a specialized mathematical assessment model and a database encompassing a large-scale healthy population. This database collects a massive amount of standardized physiological indicators to reflect the baseline of normal respiratory function. The model employs an advanced neural network learning framework, a deep learning method capable of automatically capturing complex data patterns. Based on this large-sample healthy population database, the system performs a rigorous model training and calibration process to ensure the model's generalization ability and avoid overfitting. Simultaneously, the system precisely defines reference ranges for patient age and gender stratification, setting personalized health standard intervals for different population groups.

[0076] The respiratory muscle function physiological parameters collected in step S1 undergo feature extraction processing, including calculating key statistical indicators to extract core information and performing normalization to eliminate dimensional differences and optimize input data quality, thereby improving the accuracy of the analysis. This optimized data is then input into a mathematical evaluation model. The model dynamically predicts and outputs an indicator called "Restore Initial Assessment Value" through an internal mapping mechanism. This value quantifies the initial state of the patient's respiratory muscle function in numerical form between 0 and 1, and is directly used to predict the possible quality level of the patient's respiratory status after completing the entire clinical treatment phase, providing a reliable basis for subsequent decision-making.

[0077] The initial values ​​for the recovery assessment are calculated using the following formula:

[0078]

[0079] in:

[0080] To restore the initial assessment value, the range is between [0,1], representing the preliminary predictive quality of respiratory muscle function after the patient's recovery;

[0081] Normalized blood oxygen saturation, obtained through feature extraction by averaging time-series data, represents the output outcome, i.e., the ultimate efficiency of gas exchange. Blood oxygen saturation is one of the core objectives of the respiratory system and the ultimate indicator of respiratory function effectiveness.

[0082] The normalized amplitude of chest and abdominal movements, obtained through feature extraction of the average amplitude of collected chest and abdominal movements, directly reflects the mechanical activity of the respiratory muscles and is a direct manifestation of respiratory drive. This parameter is particularly important in patients with neuromuscular diseases.

[0083] The normalized variation in respiratory flow rate is obtained through feature extraction over a fixed time period, referring to the calculation of peak flow rate, flow rate standard deviation, or average inspiratory flow rate. It reflects the patency of the airway and the speed at which the respiratory muscles generate airflow.

[0084] This is a normalized indicator of respiratory depth, obtained by collecting tidal volume data. Tidal volume is the amount of gas inhaled or exhaled with each quiet breath; it is the gold standard for assessing respiratory depth and is directly related to ventilation efficiency.

[0085] The weighting coefficients for blood oxygen saturation, chest and abdominal movement amplitude, respiratory flow rate change amplitude, and respiratory depth are dynamically optimized through neural network training. All of these are used to reflect the degree of influence of each respiratory muscle function physiological parameter on the recovery of respiratory muscle function.

[0086] and The following relationship must be satisfied: During the training process of the neural network, the neural network learns weights using the health database, so that the calculation formula for the initial value of the recovery assessment statistically minimizes the prediction error, and avoids overfitting through cross-validation calibration.

[0087] S3: In the clinical monitoring process, the system actively collects the patient's interfering parameters. These parameters aim to capture external factors that may affect the recovery of respiratory function. Specifically, these include the standard deviation of body mass index (BMI) (obtained by calculating the z-score deviation between the measured BMI and the standard value of healthy individuals of the same age and sex, reflecting the degree of metabolic load abnormality), serum albumin levels (uniformly calibrated to the same testing standard, indicating nutritional status), and recent medication record intervals (extracting the average time interval between the patient's three most recent respiratory-related drug uses, converted to hours). Subsequently, the system specifically establishes an assessment value correction grading model. This model intelligently integrates the above interfering parameters using a dynamic weighting mechanism and efficiently calculates the assessment correction coefficient through a preset correction formula. This coefficient is directly used to adjust the accuracy and reliability of subsequent respiratory muscle function assessment values.

[0088] The assessment correction coefficient is calculated using the following formula:

[0089]

[0090] in:

[0091] To evaluate the correction coefficient, it is used to compare the initial value with the restoration assessment. To perform correction, when Time indicates positive correction, and the initial value is used to assess recovery. Enlarge it when Time indicates negative correction, which is used to assess the initial value for recovery. Shrink, when Time indicates restoration of the initial assessment value No correction is needed;

[0092] The standard deviation of BMI is given by the following formula: Time increases metabolic load and respiratory work, while the accumulation of pleural fat restricts lung expansion;

[0093] when Time represents muscle depletion, including the atrophy of respiratory muscles, leading to decreased muscle strength;

[0094] Serum albumin level is a core indicator used to measure nutritional status and physiological reserves. Low albumin is directly related to insufficient protein synthesis in respiratory muscles and weakened muscle strength.

[0095] The interval for recent medication records. It represents a proxy indicator of treatment adherence and continuity of care, and is measured through... Make The smaller, The smaller the value, the better for the initial value of the recovery assessment. Perform negative correction. When it is larger, Logarithmic growth is slowing, affecting the initial value of the recovery assessment. The degree of negative correction decreases, and the smaller the interval between medication records, the greater the favorable directional interference to the initial value of the above recovery assessment.

[0096] The weighting coefficients are the standard deviation of body mass index (BMI), serum albumin level, and the interval between recent medication records. All are positive real numbers, and satisfy the following conditions: .

[0097] In the data processing stage, the standard deviation of body mass index (BMI) is accurately obtained by calculating the z-score statistical deviation between the patient's measured BMI value and the standard reference value of healthy people of the same age and sex. This deviation quantitatively reflects the degree of abnormality of the actual metabolic load relative to the normal level. The serum albumin level value is uniformly calibrated to the same laboratory testing standard. This method is specifically used to eliminate systematic errors caused by differences in instruments or reagents, and to ensure the comparability and accuracy of the data.

[0098] During the collection of recent medication record intervals, the system efficiently extracts the average time interval between the patient's three most recent respiratory-related drug use events as a key input. The units are then uniformly converted to hours to unify the dimensions, and the data is smoothed by taking the natural logarithm to significantly weaken the marginal impact of longer intervals on the overall data distribution.

[0099] The interference parameters are standardized and mapped, including mapping the absolute value of the standard deviation of body mass index (BMI) to the interval [0,3], with a cutoff at |B|≥3; converting serum albumin level values ​​into a linear decay coefficient based on the clinical threshold of 35g / L; and processing the recent medication record interval using a piecewise function. Finally, the three interference parameters are integrated into the evaluation value correction grading model, and the evaluation correction coefficient is generated through a dynamic weighted formula. The serum albumin level value directly contributes negatively, while the standard deviation of body mass index (BMI) and the recent medication record interval provide positive or negative contributions. The evaluation value correction grading model is validated by a clinical expert rule base before output to ensure that the correction logic conforms to the pathophysiological mechanism of respiratory function compensation.

[0100] S4: Based on the assessment correction coefficient obtained in step S3, the initial value of the recovery assessment of respiratory muscle function output by the mathematical assessment model is adjusted to generate a recovery assessment correction value. The recovery assessment correction value is used to ultimately predict the respiratory status quality of the patient after completing the clinical treatment stage. Finally, the respiratory muscle function judgment result of the tested patient is output through the recovery assessment correction value.

[0101] The corrective value for restoration assessment is calculated using the following formula:

[0102]

[0103] in To restore the assessment correction value, which serves as a quantitative indicator for ultimately predicting the respiratory status quality of patients after the completion of clinical treatment, a preset risk threshold rule was then applied.

[0104] Set the health threshold for assessing the recovery of respiratory muscle function as: The patient's respiratory muscle function assessment results are generated using the following grading system:

[0105] when At that time, the respiratory muscle function was determined to be well compensated, and the respiratory status quality was predicted to reach that of healthy people after treatment. Under this result, the current treatment plan was maintained and the test was repeated weekly.

[0106] when At that time, it was determined to be mild functional inhibition. Under this result, nutritional support was optimized and respiratory muscle strengthening training and medication were initiated, and the monitoring period was shortened to 72 hours.

[0107] when At that time, the patient was determined to be at high risk of respiratory failure. Based on this result, the patient was transferred to the intensive care unit and put into mechanical ventilation. At the same time, the model input parameters were recalibrated.

[0108] This embodiment also provides a wireless respiratory monitoring system for clinical treatment. The respiratory monitoring system is used to perform the above-described respiratory monitoring method. The respiratory monitoring system includes:

[0109] The data acquisition module is used to collect respiratory muscle function physiological parameters Y of clinical patients. The respiratory muscle function physiological parameters Y include the amplitude of chest and abdominal movement, blood oxygen saturation, the amplitude of respiratory flow change, and respiratory depth index. The sampling frequency is in the range of 50-100Hz to capture details. It provides high-precision and low-noise raw input signals through Bluetooth wireless transmission.

[0110] The model training and evaluation module is used to establish a mathematical evaluation model and a large sample database of healthy people. The model adopts a neural network learning framework and is trained and calibrated based on the large sample database of healthy people. The reference range for patient age and gender stratification is defined. The physiological parameters of respiratory muscle function collected in step S1 are subjected to feature extraction and normalization to optimize the input quality. Then, the input is fed into the mathematical evaluation model to map and output the initial value of respiratory muscle function recovery evaluation. The initial value of recovery evaluation is used to preliminarily predict the respiratory status quality of patients after completing the clinical treatment stage.

[0111] The correction generation module is used to collect interference parameters from clinical patients. The interference parameters include the standard deviation of body mass index (BMI), serum albumin level, and recent medication record interval. An assessment value correction grading model is established, and the assessment correction coefficient is obtained by integrating the interference parameters through the correction formula.

[0112] The hierarchical decision output module and the evaluation correction coefficient obtained from the correction generation module adjust the initial value of the recovery assessment of respiratory muscle function output by the mathematical evaluation model to generate a recovery assessment correction value. The recovery assessment correction value is used to ultimately predict the respiratory status quality of the patient after completing the clinical treatment stage. Finally, the respiratory muscle function judgment result of the tested patient is output through the recovery assessment correction value.

[0113] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0114] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0115] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0116] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A wireless respiratory monitoring method for clinical treatment, characterized in that, Includes the following steps: S1: Collect respiratory muscle function physiological parameters Y from clinical patients. The respiratory muscle function physiological parameters Y include chest and abdominal movement amplitude, blood oxygen saturation, respiratory flow rate change amplitude, and respiratory depth index. The sampling frequency is in the range of 50-100Hz to capture details. The data is transmitted wirelessly via Bluetooth to provide high-precision and low-noise raw input signals. S2: Establish a mathematical assessment model and a large-sample healthy population database. The model adopts a neural network learning framework and is trained and calibrated based on the large-sample healthy population database. Define reference ranges for patient age and gender stratification. Optimize the input quality by extracting and normalizing the physiological parameters of respiratory muscle function collected in step S1. Then input the parameters into the mathematical assessment model and map the output of the initial value of respiratory muscle function recovery assessment. The initial value of recovery assessment is used to preliminarily predict the respiratory status quality of the patient after completing the clinical treatment stage. S3: Collect interference parameters from clinical patients, including the standard deviation of body mass index (BMI), serum albumin level, and recent medication record interval. Establish an assessment value correction grading model, and obtain the assessment correction coefficient through the correction formula after integrating the interference parameters. S4: Based on the assessment correction coefficient obtained in step S3, the initial value of the recovery assessment of respiratory muscle function output by the mathematical assessment model is adjusted to generate a recovery assessment correction value. The recovery assessment correction value is used to ultimately predict the respiratory status quality of the patient after completing the clinical treatment stage. Finally, the respiratory muscle function judgment result of the tested patient is output through the recovery assessment correction value.

2. The wireless respiratory monitoring method for clinical treatment according to claim 1, characterized in that: In step S1, the amplitude of chest and abdominal movement is used to reflect the displacement change of respiratory muscle contraction and relaxation. The blood oxygen saturation is obtained by real-time monitoring of the percentage value of blood oxygenation and used to evaluate gas exchange efficiency. After continuously recording the time-series signal of airflow rate, the amplitude of respiratory flow rate change is generated to reflect the flow rate during inhalation and exhalation. The respiratory depth index is used to quantify the volume change of each breath. Normalize the functional physiological parameter Y: ， in: The collected physiological parameters of the patient's respiratory muscle function; These are the normalized physiological parameters of respiratory muscle function; The minimum values ​​of physiological parameters of healthy respiratory muscle function in various population groups; The maximum values ​​of physiological parameters of healthy respiratory muscle function in various population groups; The original input signal is the received original multi-channel physiological data stream, and after receiving the data stream, real-time quality control is implemented: dynamic range verification is performed on the chest and abdominal movement amplitude signal, and abnormal peaks exceeding the physiological limits of the human body are automatically eliminated; a bandpass filter of 0.1-20Hz is simultaneously activated to eliminate high-frequency electromyographic interference and low-frequency body movement noise.

3. The wireless respiratory monitoring method for clinical treatment according to claim 2, characterized in that: The high-frequency electromyography interference is a signal greater than 20Hz, and the low-frequency body movement noise is a signal less than 0.1Hz. Then, baseline drift correction is performed on the blood oxygen saturation data: the trend term is separated by sliding window mid-range filtering to eliminate the slow drift caused by limb micro-movements; the change amplitude of respiratory flow is compensated by phase alignment, and the cross-correlation algorithm is used to eliminate the millisecond-level transmission delay between chest and abdominal movements and airflow signals. In the cross-correlation algorithm, a regression model is constructed by extracting acceleration features to separate non-respiratory displacement components from chest and abdominal motion signals. Finally, invalid data segments are marked by adaptive threshold segmentation, including coughing or body position reversal periods. Timestamp synchronization calibration is performed on the data of each channel to control the alignment of multimodal data within ±10ms error and output a standardized signal matrix that meets the analysis requirements.

4. The wireless respiratory monitoring method for clinical treatment according to claim 1, characterized in that: The initial value for the recovery assessment is calculated using the following formula: ， in: To restore the initial assessment value, the range is between [0,1], representing the preliminary predictive quality of respiratory muscle function after the patient's recovery; The normalized blood oxygen saturation was obtained by feature extraction using the average value of the time series data. The normalized amplitude of chest and abdominal movements is obtained by calculating the feature extraction method of the average amplitude of collected chest and abdominal movements. The normalized variation in respiratory flow rate is obtained by calculating the feature extraction method over a fixed time period. The normalized respiratory depth index is obtained by collecting tidal volume. The weighting coefficients for blood oxygen saturation, chest and abdominal movement amplitude, respiratory flow rate change amplitude, and respiratory depth are dynamically optimized through neural network training. All of these are used to reflect the degree of influence of each respiratory muscle function physiological parameter on the recovery of respiratory muscle function. and The following relationship must be satisfied: During the training process of the neural network, the neural network learns weights using the health database, so that the calculation formula for the initial value of the recovery assessment statistically minimizes the prediction error, and avoids overfitting through cross-validation calibration.

5. The wireless respiratory monitoring method for clinical treatment according to claim 4, characterized in that: The evaluation correction coefficient in step S3 is calculated using the following formula: ， in: To evaluate the correction coefficient, it is used to compare the initial value with the restoration assessment. To perform correction, when Time indicates positive correction, and the initial value is used to assess recovery. Enlarge it when Time indicates negative correction, which is used to assess the initial value for recovery. Shrink, when Time indicates restoration of the initial assessment value No correction is needed; BMI standard deviation; This refers to serum albumin levels. The interval for recent medication records, and through Make The smaller, The smaller the value, the better for the initial value of the recovery assessment. When negative correction is applied, the larger the value, Logarithmic growth is slowing, affecting the initial value of the recovery assessment. The degree of negative correction decreases; The weighting coefficients are the standard deviation of body mass index (BMI), serum albumin level, and the interval between recent medication records. All are positive real numbers, and satisfy the following conditions: .

6. The wireless respiratory monitoring method for clinical treatment according to claim 5, characterized in that: The body mass index (BMI) standard deviation is calculated by comparing the measured BMI with the standard value of healthy individuals of the same age and sex using the z-score, reflecting the degree of metabolic load abnormality. The serum albumin level is uniformly calibrated to the same detection standard to eliminate systematic errors. During the collection of recent medication records, the average time interval between the patient's three most recent respiratory-related drug uses is extracted, converted to hours, and the natural logarithm is used to weaken the marginal effect of long intervals.

7. The wireless respiratory monitoring method for clinical treatment according to claim 1, characterized in that: The interference parameters are standardized and mapped, including mapping the absolute value of the standard deviation of body mass index (BMI) to the interval [0,3], with a cutoff at |B|≥3; converting serum albumin level values ​​into a linear decay coefficient based on the clinical threshold of 35g / L; and processing the recent medication record interval using a piecewise function. Finally, the three interference parameters are integrated into the evaluation value correction grading model, and the evaluation correction coefficient is generated through a dynamic weighted formula. The serum albumin level value directly contributes negatively, while the standard deviation of body mass index (BMI) and the recent medication record interval provide positive or negative contributions. The evaluation value correction grading model is validated by a clinical expert rule base before output to ensure that the correction logic conforms to the pathophysiological mechanism of respiratory function compensation.

8. The wireless respiratory monitoring method for clinical treatment according to claim 1, characterized in that: The recovery assessment correction value is calculated using the following formula: ， in To restore the assessment correction value, which serves as a quantitative indicator for ultimately predicting the respiratory status quality of patients after the completion of clinical treatment, a preset risk threshold rule was then applied.

9. The wireless respiratory monitoring method for clinical treatment according to claim 8, characterized in that: In step S4, the healthy threshold for assessing the recovery of respiratory muscle function is set as follows: The patient's respiratory muscle function assessment results are generated using the following grading system: when At that time, the respiratory muscle function was determined to be well compensated, and the respiratory status quality was predicted to reach that of healthy people after treatment. Under this result, the current treatment plan was maintained and the test was repeated weekly. when At that time, it was determined to be mild functional inhibition. Under this result, nutritional support was optimized and respiratory muscle strengthening training and medication were initiated, and the monitoring period was shortened to 72 hours. when At that time, the patient was determined to be at high risk of respiratory failure. Based on this result, the patient was transferred to the intensive care unit and put into mechanical ventilation. At the same time, the model input parameters were recalibrated.

10. A wireless respiratory monitoring system for clinical treatment, characterized in that, The respiratory monitoring system is used to perform the respiratory monitoring method as described in any one of claims 1-9, the respiratory monitoring system comprising: The data acquisition module is used to collect respiratory muscle function physiological parameters Y of clinical patients. The respiratory muscle function physiological parameters Y include the amplitude of chest and abdominal movement, blood oxygen saturation, the amplitude of respiratory flow change, and respiratory depth index. The sampling frequency is in the range of 50-100Hz to capture details. It provides high-precision and low-noise raw input signals through Bluetooth wireless transmission. The model training and evaluation module is used to establish a mathematical evaluation model and a large sample database of healthy people. The model adopts a neural network learning framework and is trained and calibrated based on the large sample database of healthy people. The reference range for patient age and gender stratification is defined. The physiological parameters of respiratory muscle function collected in step S1 are subjected to feature extraction and normalization to optimize the input quality. Then, the input is fed into the mathematical evaluation model to map and output the initial value of respiratory muscle function recovery evaluation. The initial value of recovery evaluation is used to preliminarily predict the respiratory status quality of patients after completing the clinical treatment stage. The correction generation module is used to collect interference parameters from clinical patients. The interference parameters include the standard deviation of body mass index (BMI), serum albumin level, and recent medication record interval. An assessment value correction grading model is established, and the assessment correction coefficient is obtained by integrating the interference parameters through the correction formula. The hierarchical decision output module and the evaluation correction coefficient obtained from the correction generation module adjust the initial value of the recovery assessment of respiratory muscle function output by the mathematical evaluation model to generate a recovery assessment correction value. The recovery assessment correction value is used to ultimately predict the respiratory status quality of the patient after completing the clinical treatment stage. Finally, the respiratory muscle function judgment result of the tested patient is output through the recovery assessment correction value.

Images