Lung function evaluation method

Abstract

The invention discloses a lung function evaluation method, and belongs to the technical field of rehabilitation monitoring, and the method comprises the steps: obtaining a prostrating ventilation coefficient according to the number of prostrating ventilation times and the total duration of prostrating ventilation; a blood oxygen evaluation coefficient is obtained according to the mean value of the blood oxygenation index difference values before and after lung treatment in the treatment cycle, the blood oxygenation index difference values before and after the treatment cycle and the blood oxygenation index after treatment; obtaining a lung ultrasound evaluation coefficient according to the mean value of the difference values of the lung ultrasound scoring indexes before and after the lung treatment in the treatment cycle, the difference values of the lung ultrasound scoring indexes before and after the treatment cycle and the lung ultrasound scoring indexes after the treatment cycle; a lung function evaluation coefficient is obtained based on the prone ventilation coefficient, the APACHEII score of the patient, the blood oxygen evaluation coefficient and the lung ultrasound evaluation coefficient; constructing a follow-up frequency optimization model based on the lung function evaluation coefficient and the reference follow-up frequency, and outputting and determining a target follow-up frequency; according to the method, individual treatment response is quantified through multi-model cooperation, and the outlier patient management precision is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of rehabilitation monitoring technology, and in particular relates to a method for evaluating lung function. Background Technology

[0002] Traditional methods for assessing pulmonary function often rely on single indicators (such as blood gas analysis or imaging scores), making it difficult to dynamically reflect individualized treatment responses. Current technologies lack a systematic integration of multi-dimensional data (treatment process, physiological indicators, structural changes), leading to delays in identifying high-risk patients and rigid follow-up strategies. Especially for outliers with complex conditions (such as those with abnormal treatment responses or severe underlying diseases), routine management is prone to assessment biases, delaying intervention. Therefore, there is an urgent need for a multi-model collaborative evaluation system that integrates dynamic treatment data, individual physiological characteristics, and structural changes to achieve precise patient management. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a method for evaluating lung function, which solves the aforementioned problems.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for evaluating lung function, comprising the following steps:

[0005] A prone ventilation model was constructed based on the number of prone ventilations and the total duration of prone ventilation, and the prone ventilation coefficient was output.

[0006] Based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment, a blood oxygenation evaluation model is constructed and a blood oxygenation evaluation coefficient is output.

[0007] Based on the mean difference of lung ultrasound score index before and after lung treatment within the treatment cycle, the difference of lung ultrasound score index before and after the treatment cycle, and the lung ultrasound score index after the treatment cycle, a lung ultrasound evaluation model is constructed and lung ultrasound evaluation coefficients are output.

[0008] A lung function evaluation model was constructed based on the prone ventilation coefficient, the blood oxygenation evaluation coefficient under the patient's APACHE II score (Acute Physiology and Chronic Health Assessment II), and the lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output.

[0009] A follow-up frequency optimization model is constructed based on lung function evaluation coefficients and baseline follow-up frequency to determine the target follow-up frequency.

[0010] Based on the above technical solutions, the present invention also provides the following optional technical solutions:

[0011] A further technical solution: The follow-up frequency optimization model is expressed as follows:

[0012] Ftar =F0+δ(μ-P cur )

[0013] Among them, F tar The target follow-up frequency is represented by F0, the baseline follow-up frequency is represented by δ, the follow-up frequency adjustment scaling factor is represented by μ, and the reference value of the lung function evaluation coefficient is represented by P. cur This represents the current lung function evaluation coefficient.

[0014] Further technical solution: The steps for constructing a prone ventilation model and outputting the prone ventilation coefficient based on the number of prone ventilations and the total duration of prone ventilation are as follows:

[0015] The number of prone ventilations was obtained by performing a maximum-min normalization process;

[0016] The duration factor is obtained by comparing the total duration of prone ventilation with the maximum permissible total duration.

[0017] A prone ventilation model is constructed based on the frequency factor and duration factor to output the prone ventilation coefficient. The prone ventilation model is expressed as follows:

[0018]

[0019] Where α represents the prone ventilation coefficient, N ind T represents the frequency factor. ind The duration factor is indicated by the prone ventilation coefficient, which is directly proportional to the ventilation effect.

[0020] Further technical solution: The steps for constructing a blood oxygenation evaluation model and outputting blood oxygenation evaluation coefficients based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the post-treatment blood oxygenation index are as follows:

[0021] The mean difference of blood oxygenation index before and after lung treatment within the treatment cycle, the difference of blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment are processed by maximum-min normalization to obtain the mean blood oxygenation factor, the difference of blood oxygenation factor, and the blood oxygenation factor.

[0022] A blood oxygenation evaluation model is constructed based on the mean blood oxygenation factor, the difference blood oxygenation factor, and the blood oxygenation factor, and outputs the blood oxygenation evaluation coefficient. The blood oxygenation evaluation model is expressed as follows:

[0023] β=a1ΔB mean,ind +a2ΔB total,ind +a3B end,ind

[0024] Where β represents the blood oxygenation evaluation coefficient, ΔB mean,ind ΔB represents the mean blood oxygenation factor. total,indB represents the oxygenation difference factor. end,ind Indicates blood oxygenation factor, a i Represents the weight coefficients and ∑a i =1, the blood oxygen evaluation coefficient is directly proportional to the blood oxygen improvement effect.

[0025] Further technical solution: The steps for constructing a lung ultrasound evaluation model and outputting lung ultrasound evaluation coefficients based on the mean difference in lung ultrasound scoring index before and after treatment within the treatment cycle, the difference in lung ultrasound scoring index before and after treatment, and the lung ultrasound scoring index after treatment are as follows:

[0026] The mean difference of lung ultrasound score index before and after treatment within the treatment cycle, the difference of lung ultrasound score index before and after treatment cycle, and the lung ultrasound score index after treatment cycle are processed by maximum-min normalization to obtain the ultrasound mean factor, ultrasound difference factor, and ultrasound factor.

[0027] A lung ultrasound evaluation model is constructed based on the ultrasound mean factor, ultrasound difference factor, and ultrasound factor to output lung ultrasound evaluation coefficients. The lung ultrasound evaluation model is expressed as follows:

[0028] γ=b1ΔL mean,ind +b2ΔL total,ind +b3L end,ind

[0029] Where γ represents the lung ultrasound evaluation coefficient, ΔL mean,ind ΔL represents the ultrasound mean factor. total,ind L represents the ultrasound difference factor. end,ind Indicates the ultrasound factor, b i Represents the weight coefficients and ∑b i =1, the lung ultrasound evaluation coefficient is directly proportional to the improvement effect of blood oxygenation.

[0030] Further technical solution: The steps for constructing a pulmonary function evaluation model and outputting pulmonary function evaluation coefficients based on prone ventilation coefficient, patient APACHE II score (Acute Physiology and Chronic Health Evaluation II), and lung ultrasound evaluation coefficient are as follows:

[0031] The APACHE II score of the patient was processed by maximum-min normalization to obtain the APACHE II score index;

[0032] A lung function evaluation model was constructed based on the APACHE II score index, prone ventilation coefficient, blood oxygenation evaluation coefficient, and lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output.

[0033] The lung function assessment model is represented as follows:

[0034] P = c1α + c2βexp(-kA A)+c3γexp(-k A A)

[0035] Wherein, P represents the lung function evaluation coefficient, α represents the prone ventilation coefficient, β represents the blood oxygenation evaluation coefficient, γ represents the lung ultrasound evaluation coefficient, A represents the APACHE II score index, and k A c represents the APACHE II score index decay factor. i Denotes the weight coefficients and ∑c i =1, meaning the improvement in blood oxygenation is directly proportional to the lung function evaluation coefficient.

[0036] Further technical solution: The method for obtaining the blood oxygenation index is as follows:

[0037] Obtain arterial blood oxygen partial pressure, arterial blood carbon dioxide partial pressure, and arterial blood pH value;

[0038] Import arterial blood oxygen partial pressure into the formula Obtain the arterial oxygen partial pressure index, O cur Indicates the partial pressure of oxygen in arterial blood, O opt This represents the ideal value of arterial blood oxygen partial pressure;

[0039] Arterial blood carbon dioxide partial pressure is incorporated into the formula. Obtain the partial pressure index of carbon dioxide in arterial blood, CO cur Indicates the partial pressure of carbon dioxide in arterial blood, CO opt This represents the ideal value of arterial blood carbon dioxide partial pressure;

[0040] Import arterial blood pH value into the formula The pH index of arterial blood is obtained from the sample. pH represents the pH value of arterial blood. opt ΔpH represents the standard arterial blood pH value. max This indicates the maximum permissible deviation of arterial blood pH.

[0041] A blood oxygenation index is constructed based on the arterial blood oxygen partial pressure index, arterial blood carbon dioxide partial pressure index, and arterial blood pH index. The blood oxygenation index is then output as follows:

[0042] B = d1O + d2(1-CO) + d3(1-pH) ind )

[0043] Where B represents the blood oxygenation index, O represents the arterial blood oxygen partial pressure index, CO represents the arterial blood carbon dioxide partial pressure index, and pH... ind Indicates the pH index of arterial blood, d i Denotes the weight coefficients and ∑d i =1, the higher the blood oxygenation index, the better the oxygenation status.

[0044] Further technical solution: The lung ultrasound scoring index is obtained as follows:

[0045] The ratio of the diaphragm thickness at the end of inspiration to the diaphragm thickness at the end of inspiration and the diaphragm thickness at the end of inspiration is calculated to obtain the diaphragm thickening rate.

[0046] Import the diaphragm thickening rate into the formula The diaphragm thickening rate index is obtained from the data. DTF represents the diaphragm thickening rate. opt This indicates the reference value for the diaphragm thickening rate;

[0047] Import the diaphragmatic movement (the distance the diaphragm moves between deep expiration and inspiration, in millimeters) into the formula. The diaphragm movement index (DE) is obtained from the data. DE represents the degree of diaphragm movement. opt This indicates the reference value for diaphragmatic mobility;

[0048] Summarize the LUS scores under the 12-zone method The LUS score index is obtained by comparing it with the maximum sum of scores.

[0049] A lung ultrasound scoring model is constructed based on the LUS score index, diaphragm movement index, and diaphragm thickening rate index to output a lung ultrasound scoring index. The lung ultrasound scoring model is expressed as follows:

[0050] L=e1(1-LUS ind )+e2DE ind +e3DTF ind

[0051] Where L represents the Lung Ultrasound Score Index, LUS ind Denotes the LUS rating index, DE ind The Diaphragmatic Movement Index (DTF) represents the diaphragmatic movement index. ind The diaphragm thickening rate index, e i Denotes the weight coefficients and ∑e i =1, where a higher lung ultrasound score index indicates better oxygenation.

[0052] This invention provides a method for evaluating lung function, which has the following advantages compared with the prior art:

[0053] 1. This invention dynamically quantifies treatment response using a prone ventilation model, a blood oxygenation evaluation model, and a lung ultrasound evaluation model. Combined with the attenuation effect of the APACHE II score index, it accurately identifies outlier patients (such as P...). cur Those that deviate significantly from μ);

[0054] 2. This invention can output the coefficient P through a lung function evaluation model. curThe follow-up frequency optimization model automatically adjusts the follow-up frequency, allowing high-risk patients to receive higher follow-up density and avoiding management omissions. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0057] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0058] Please see Figure 1 A lung function evaluation method provided in one embodiment of the present invention includes the following steps:

[0059] A prone ventilation model was constructed based on the number of prone ventilations and the total duration of prone ventilation, and the prone ventilation coefficient was output.

[0060] Based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment, a blood oxygenation evaluation model is constructed and a blood oxygenation evaluation coefficient is output.

[0061] Based on the mean difference of lung ultrasound score index before and after lung treatment within the treatment cycle, the difference of lung ultrasound score index before and after the treatment cycle, and the lung ultrasound score index after the treatment cycle, a lung ultrasound evaluation model is constructed and lung ultrasound evaluation coefficients are output.

[0062] A lung function evaluation model was constructed based on the prone ventilation coefficient, the blood oxygenation evaluation coefficient under the patient's APACHE II score (Acute Physiology and Chronic Health Assessment II), and the lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output.

[0063] A follow-up frequency optimization model is constructed based on lung function evaluation coefficients and baseline follow-up frequency to determine the target follow-up frequency.

[0064] The prone ventilation frequency refers to the number of times the patient performs prone ventilation per unit of time, which can be obtained from ventilator recording data or nursing operation logs, and is used to reflect treatment compliance and intervention intensity. The total prone ventilation duration refers to the cumulative prone ventilation time, which can be timed using bedside monitoring equipment, and is used to assess the continuity of ventilation intervention. The mean difference in blood oxygenation index refers to the average improvement in blood oxygenation index measured at various times during the treatment cycle, which can be calculated from arterial blood gas analysis data, and is used to capture the stability of blood oxygenation changes. The difference in lung ultrasound score index refers to the quantitative value of changes in ultrasound imaging characteristics before and after treatment, which can be obtained using a 12-region scoring system, and is used to characterize the degree of lung structural recovery. The APACHE II score refers to the standardized scoring result of the Acute Physiology and Chronic Health Evaluation System, Version 2, which can be obtained through a clinical information system, and is used to correct for the influence of the patient's baseline health status on efficacy assessment.

[0065] Specifically, the prone ventilation model transforms ventilation frequency and duration into standardized coefficients through normalization, for example, limiting the ratio of actual ventilation frequency to a preset target value to the range of 0-1 to avoid interference from extreme values. The blood oxygenation assessment model simultaneously considers the average level of blood oxygenation improvement during the treatment cycle, the overall magnitude of change, and the final state after treatment, for example, using a weighted summation method to balance short-term fluctuations and long-term trends. The lung ultrasound assessment model integrates the dynamic changes in ultrasound scores during treatment, for example, combining the mean difference in scores with the final score to reflect both recovery speed and final effect. The pulmonary function assessment model introduces the APACHE II score as an exponential decay factor, for example, using an exponential function to reduce the weight of blood oxygenation and ultrasound scores in patients with poor baseline health, making the assessment results more consistent with individual circumstances. The follow-up frequency optimization model dynamically adjusts the follow-up interval based on the difference between the pulmonary function assessment coefficient and the reference value, for example, automatically increasing the follow-up density when the assessment coefficient is below a threshold.

[0066] Compared to existing technologies, traditional methods typically analyze blood gas parameters or imaging scores separately. This new approach, for the first time, integrates prone ventilation parameters, dynamic blood oxygenation trends, lung structural recovery progress, and baseline health status into a multi-dimensional calculation. For example, in assessing patients with acute exacerbations of chronic obstructive pulmonary disease (COPD), existing technologies may only focus on a single measurement of the PaO2 / FiO2 ratio. This new approach, however, simultaneously calculates the intensity of prone ventilation, the characteristics of blood oxygenation fluctuations during treatment, and the degree of improvement in diaphragmatic mobility. Furthermore, it uses the APACHE II score to correct for the impact of comorbidities on efficacy, significantly enhancing the clinical guidance value of the assessment results.

[0067] Through the above technical solutions, this application can solve the problem of assessment lag caused by the single data dimension of traditional methods, and realize the collaborative analysis of treatment process parameters, dynamic changes in physiological indicators, and structural recovery progress. For example, in the management of ARDS patients, prone ventilation coefficient can be used to identify cases with insufficient intervention, combined with blood oxygenation evaluation coefficient to detect occult oxygenation deterioration, and lung ultrasound evaluation coefficient to detect early progression of atelectasis. Finally, respiratory support strategies can be adjusted in a timely manner based on multi-dimensional evaluation results. At the same time, by dynamically adjusting the weights of the APACHE II score, patients with different baseline conditions can obtain personalized evaluation standards, avoiding a simple comparison between patients with severe pneumonia and patients with chronic lung disease. The follow-up frequency optimization model transforms static fixed follow-up into a dynamic adjustment mode. For example, for high-risk patients with continuously declining evaluation coefficients, the follow-up interval is automatically shortened, and imaging re-examination and blood gas analysis are prioritized.

[0068] Preferably, the steps for constructing a prone ventilation model and outputting the prone ventilation coefficient based on the number of prone ventilations and the total duration of prone ventilation are as follows:

[0069] The number of prone ventilations was obtained by performing a maximum-min normalization process;

[0070] The duration factor is obtained by comparing the total duration of prone ventilation with the maximum permissible total duration.

[0071] A prone ventilation model is constructed based on the frequency factor and duration factor to output the prone ventilation coefficient. The prone ventilation model is expressed as follows:

[0072]

[0073] Where α represents the prone ventilation coefficient, N ind T represents the frequency factor. ind The duration factor is indicated by the prone ventilation coefficient, which is directly proportional to the ventilation effect.

[0074] The maximum-minimum normalization process refers to a linear transformation method that maps prone ventilation frequency data to a preset numerical range. Specifically, it can be achieved by subtracting the minimum value from the original data and then dividing by the difference between the maximum and minimum values, thus eliminating absolute differences in ventilation frequency among different patients. The duration factor is the ratio of the total prone ventilation duration to the preset maximum allowable total duration. This can be achieved by dividing the actual total duration by the maximum allowable total duration, representing the degree to which ventilation duration is achieved relative to the clinically permissible upper limit. The prone ventilation model refers to a calculation rule that mathematically integrates the frequency factor and duration factor into a single coefficient. This can be achieved using a square root function combined with minimum value constraint operations, balancing the contributions of ventilation frequency and duration to the final evaluation results.

[0075] Specifically, the prone ventilation count is first input into the maximum-minimum normalization processing module. A linear transformation converts the original count into a frequency factor within the range of 0 to 1. For example, if a patient's ventilation count is 10 and the preset minimum is 5 and maximum is 20, the frequency factor can be calculated as (10-5) / (20-5) = 0.333. Then, the total prone ventilation duration is compared to the preset maximum allowable total duration. For example, if the total duration is 120 minutes and the maximum allowable total duration is 180 minutes, the duration factor is 120 / 180 = 0.667. Finally, the frequency factor and duration factor are input into the prone ventilation model. The minimum of the two factors is taken and the square root is calculated. For example, when the frequency factor is 0.8 and the duration factor is 0.9, the prone ventilation coefficient is calculated as 0.848. A higher coefficient indicates better prone ventilation performance.

[0076] Compared to existing technologies, traditional methods typically only count the number of prone ventilations or the total duration as independent assessment indicators, failing to consider the impact of individual patient tolerance differences on the assessment results, and lacking quantitative analysis of the synergistic effect of frequency and duration. This protocol eliminates absolute numerical differences between individuals through normalization, reflects the degree of conformity between treatment intensity and clinical standards through ratio calculations, and achieves a comprehensive dynamic assessment of the effectiveness of prone ventilation by fusing multi-dimensional data through a nonlinear model.

[0077] Preferably, the step of constructing a blood oxygenation evaluation model and outputting blood oxygenation evaluation coefficients based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the post-treatment blood oxygenation index is as follows:

[0078] The mean difference of blood oxygenation index before and after lung treatment within the treatment cycle, the difference of blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment are processed by maximum-min normalization to obtain the mean blood oxygenation factor, the difference of blood oxygenation factor, and the blood oxygenation factor.

[0079] A blood oxygenation evaluation model is constructed based on the mean blood oxygenation factor, the difference blood oxygenation factor, and the blood oxygenation factor, and outputs the blood oxygenation evaluation coefficient. The blood oxygenation evaluation model is expressed as follows:

[0080] β=a1ΔB mean,ind +a2ΔB total,ind +a3B end,ind

[0081] Where β represents the blood oxygenation evaluation coefficient, ΔB mean,ind ΔB represents the mean blood oxygenation factor. total,ind B represents the oxygenation difference factor. end,ind Indicates blood oxygenation factor, a i Represents the weight coefficients and ∑a i=1, the blood oxygen evaluation coefficient is directly proportional to the blood oxygen improvement effect.

[0082] Among them, the mean blood oxygenation factor refers to the normalized quantitative index of the average difference of blood oxygenation index before and after each treatment within the treatment cycle. Specifically, it can be achieved by mapping the original mean data to the 0-1 interval, which is used to reflect the average improvement of blood oxygenation level during the treatment process.

[0083] Among them, the blood oxygenation difference factor refers to the quantitative index after normalization of the total difference in blood oxygenation index before and after the entire treatment cycle. Specifically, it can be achieved by calculating the ratio of the total difference to the preset maximum allowable difference, which is used to characterize the overall trend of blood oxygenation level during the treatment cycle.

[0084] Among them, the blood oxygenation factor refers to the quantitative index after normalization of the real-time measurement value of the blood oxygenation index after treatment. Specifically, it can be calculated by comparing the current measurement value with the ideal blood oxygenation threshold, and is used to reflect the patient's current blood oxygenation status.

[0085] The weighting coefficients refer to the contribution ratio parameters of the mean blood oxygenation factor, the difference blood oxygenation factor, and the blood oxygenation factor in the blood oxygenation evaluation model. Specifically, they can be determined by using regression analysis based on clinical data to determine the relative importance of each factor to the blood oxygenation improvement effect. The balance of the model output is ensured by constraining the sum of the weighting coefficients to 1.

[0086] Specifically, the method first calculates the mean difference in blood oxygenation index before and after each treatment within the treatment cycle, and then normalizes this mean into a mean blood oxygenation factor to eliminate differences in data dimensions among different patients. Next, the total difference in blood oxygenation index before and after the entire treatment cycle is calculated, normalized, and used to generate a mean blood oxygenation factor to reflect long-term trends. Simultaneously, the post-treatment blood oxygenation index is normalized to generate a blood oxygenation factor, capturing real-time status data. Then, the three factors are substituted into a linear combination model, and the contribution ratio of each factor to the final evaluation result is adjusted through preset weight coefficients. For example, when the weight of the mean blood oxygenation factor is set to 0.5, it indicates that the average improvement during the treatment process has a dominant influence on the evaluation result. This model preserves the blood oxygenation change characteristics of a single treatment, integrates the dynamic trend within the overall treatment cycle, and combines current blood oxygenation status data to form a multi-dimensional comprehensive evaluation result.

[0087] Compared to existing technologies, traditional methods typically evaluate blood oxygenation using only a single time point after treatment, failing to distinguish between dynamic changes and overall trends during treatment. This approach, however, introduces mean difference and total difference factors within the treatment cycle to quantify the improvement in blood oxygenation at different stages. For example, for patients with significant fluctuations in blood oxygenation at the beginning of treatment but stable later, the mean blood oxygenation factor effectively smooths out short-term fluctuations, while the difference factor accurately reflects the overall improvement. Furthermore, existing technologies lack a weighting mechanism for multi-dimensional indicators, while this approach, through normalization and weighted combination, ensures that the evaluation results reflect the independent contributions of different dimensions and allow for adjustment of indicator priorities based on clinical needs.

[0088] Through the above technical solution, this application can solve the problem of one-sided evaluation caused by traditional methods relying on a single time point or a single indicator. By integrating dynamic change data, overall efficacy data, and current status data within the treatment cycle, it achieves a comprehensive quantitative evaluation of the improvement in blood oxygenation. This solution can accurately identify the fluctuation characteristics of blood oxygenation levels during treatment. For example, for patients whose blood oxygenation index temporarily increases after treatment but whose overall trend is downward, the blood oxygenation difference factor can promptly capture abnormal trends. At the same time, through the flexible configuration of weight coefficients, the evaluation focus can be adjusted according to different disease characteristics. For example, for patients with extremely low baseline blood oxygenation levels, the weight of the blood oxygenation factor can be increased to strengthen the evaluation weight of the current status, thereby providing a reliable basis for individualized treatment.

[0089] Preferably, the step of constructing a lung ultrasound evaluation model and outputting lung ultrasound evaluation coefficients based on the mean difference in lung ultrasound scoring index before and after treatment within the treatment cycle, the difference in lung ultrasound scoring index before and after treatment, and the lung ultrasound scoring index after treatment is as follows:

[0090] The mean difference of lung ultrasound score index before and after treatment within the treatment cycle, the difference of lung ultrasound score index before and after treatment cycle, and the lung ultrasound score index after treatment cycle are processed by maximum-min normalization to obtain the ultrasound mean factor, ultrasound difference factor, and ultrasound factor.

[0091] A lung ultrasound evaluation model is constructed based on the ultrasound mean factor, ultrasound difference factor, and ultrasound factor to output lung ultrasound evaluation coefficients. The lung ultrasound evaluation model is expressed as follows:

[0092] γ=b1ΔL mean,ind +b2ΔL total,ind +b3L end,ind

[0093] Where γ represents the lung ultrasound evaluation coefficient, ΔL mean,ind ΔL represents the ultrasound mean factor. total,ind L represents the ultrasound difference factor. end,indIndicates the ultrasound factor, b i Represents the weight coefficients and ∑b i =1, the lung ultrasound evaluation coefficient is directly proportional to the improvement effect of blood oxygenation.

[0094] The ultrasound mean factor refers to the result of maximum-minimum normalization of the mean difference in lung ultrasound score index before and after each treatment within the treatment cycle. Specifically, this is achieved by comparing the mean with a preset maximum possible difference range, reflecting the average degree of improvement in lung structure during treatment. The ultrasound difference factor refers to the result of maximum-minimum normalization of the overall difference in lung ultrasound score index before and after the treatment cycle. Specifically, this is achieved by calculating the ratio of the overall difference to a preset maximum allowable difference, quantifying the cumulative improvement in lung structure throughout the treatment process. The ultrasound factor refers to the result of maximum-minimum normalization of the lung ultrasound score index after the treatment cycle. Specifically, this is achieved by comparing the current score with the historical best score or ideal score, characterizing the immediate state of lung structure after treatment. The weighting coefficient is a parameter used to adjust the contribution of the three factors to the final evaluation coefficient. Specifically, it can be determined using expert experience assignment or machine learning optimization, achieving differentiated assessment needs for different patient groups or treatment stages under the constraint that the sum is 1.

[0095] Specifically, the model transforms raw data of different dimensions into standardized indicators through max-min normalization, eliminating the incomparability between dynamic and static state data within the treatment cycle. The ultrasound mean factor captures the average structural improvement generated by each intervention during treatment, reflecting the sustained effect of the treatment; the ultrasound difference factor measures the overall change in lung structure before and after treatment, reflecting the cumulative effect of the treatment plan; and the ultrasound factor directly characterizes the actual state of lung structure at the end of treatment, providing a current assessment benchmark. A comprehensive coefficient is generated by weighted summation of these three factors, balancing the contribution ratio of dynamic changes during treatment and the final state under the constraint of the weighting coefficient. This model combines structural change data in the time dimension with real-time state data in the spatial dimension through a dynamic parameter fusion mechanism, forming a multi-dimensional quantitative assessment indicator.

[0096] Compared to existing technologies, traditional methods typically assess lung structure using only a single ultrasound score or simple difference calculations, failing to distinguish the correlation between staged improvements during treatment and the final outcome. Existing technologies lack systematic analysis of the dynamic changes in multiple ultrasound scores throughout the treatment cycle and do not consider the weighting of data from different treatment stages. This proposed approach constructs a three-factor model incorporating mean change, overall difference, and current status. While preserving independent information across each dimension, it establishes an adjustable collaborative analysis mechanism, overcoming the limitations of relying on a single static indicator for assessment.

[0097] Through the above technical solution, this application can generate a quantitative comprehensive evaluation coefficient based on the dynamic changes in lung ultrasound scoring data during the treatment cycle and the current structural status. This coefficient can accurately reflect the contribution of different treatment stages to the improvement of lung structure, identify cases of abnormal fluctuations or stagnant changes, and provide data support for adjusting respiratory support parameters or developing individualized follow-up plans based on the structural change dimension. By eliminating dimensional differences and dynamic weight allocation, the risk of misjudgment caused by ignoring fluctuations in the treatment process or over-reliance on a single assessment is reduced, thereby improving the monitoring sensitivity of lung structural changes in patients with complex conditions.

[0098] Preferably, the steps for constructing a pulmonary function evaluation model and outputting pulmonary function evaluation coefficients based on the prone ventilation coefficient, the patient's APACHE II score (Acute Physiology and Chronic Health Evaluation II), and the lung ultrasound evaluation coefficient are as follows:

[0099] The APACHE II score of the patient was processed by maximum-min normalization to obtain the APACHE II score index;

[0100] A lung function evaluation model was constructed based on the APACHE II score index, prone ventilation coefficient, blood oxygenation evaluation coefficient, and lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output.

[0101] The lung function assessment model is represented as follows:

[0102] P = c1α + c2βexp(-k A A)+c3γexp(-k A A)

[0103] Wherein, P represents the lung function evaluation coefficient, α represents the prone ventilation coefficient, β represents the blood oxygenation evaluation coefficient, γ represents the lung ultrasound evaluation coefficient, A represents the APACHE II score index, and k A c represents the APACHE II score index decay factor. i Denotes the weight coefficients and ∑c i =1, meaning the improvement in blood oxygenation is directly proportional to the lung function evaluation coefficient.

[0104] Among them, the APACHE II score index refers to the standardized numerical value of the Acute Physiology and Chronic Health Score II (APSII) converted into a maximum-minimum normalization. Specifically, a linear scaling method can be used to map the original score to the [0,1] interval to eliminate dimensional differences between different patients. The prone ventilation coefficient is a quantitative indicator reflecting the effect of prone ventilation intervention. It can be achieved through normalization of prone ventilation frequency and total duration, and geometric mean calculation, used to characterize the direct impact of ventilation support on lung function. The blood oxygenation evaluation coefficient is an assessment indicator constructed based on the dynamic changes of the blood oxygenation index. Specifically, it can be achieved through a weighted combination of the mean, total, and post-treatment blood oxygenation difference within the treatment cycle, used to quantify the degree of blood oxygenation improvement. The lung ultrasound evaluation coefficient is an assessment indicator integrating lung ultrasound scores and diaphragmatic function parameters. Specifically, it can be achieved through a linear weighted combination of the mean, total, and post-treatment ultrasound score differences, used to reflect the improvement status of lung structure. The exponential decay function refers to an exponential weighting adjustment mechanism with the APACHE II score index as the independent variable. Specifically, it can be achieved by using exp(-k A A) Form implementation, used to dynamically suppress the assessment weights of blood oxygen and ultrasound indicators when the baseline health status deteriorates.

[0105] Specifically, the APACHE II score, after being converted into a standardized index through max-min normalization, is input into the pulmonary function assessment model along with the prone ventilation coefficient, blood oxygenation evaluation coefficient, and lung ultrasound evaluation coefficient. The prone ventilation coefficient participates directly in the calculation as an independent term, reflecting the immediate effect of ventilation intervention. The blood oxygenation evaluation coefficient and lung ultrasound evaluation coefficient are correlated with the APACHE II score index through exponential decay functions. When the patient's baseline health status deteriorates, an increase in the APACHE II score index leads to a decrease in the index value, thereby reducing the contribution weight of blood oxygenation and ultrasound indicators. Weighting coefficient constraints ensure the balance of the three types of indicators—prone ventilation, blood oxygenation improvement, and structural changes—avoiding a single indicator dominating the assessment results. This model achieves dynamic coupling between physiological characteristics and treatment response through the APACHE II score index decay factor, making the assessment results more closely reflect the patient's actual condition.

[0106] Compared to existing technologies, traditional methods rely on a single static indicator to assess lung function, failing to consider the impact of the patient's baseline health status on treatment response. For example, in existing technologies, the improvement in blood oxygenation may be overestimated, leading to assessment bias in patients with severe underlying diseases. This approach dynamically adjusts the weights of blood oxygenation and ultrasound indicators using the APACHE II scoring index, automatically reducing the contribution of relevant indicators when the patient's baseline condition deteriorates, thereby avoiding the problem of inflated assessments caused by interference from underlying diseases. Simultaneously, the model integrates three types of heterogeneous data: ventilation intervention, dynamic blood oxygenation, and structural changes, overcoming the limitation of traditional methods' single data dimension.

[0107] Through the above technical solutions, this application can effectively identify treatment response biases caused by differences in baseline health status, improving the accuracy of high-risk patient screening. By dynamically suppressing the weight of blood oxygenation and ultrasound indicators when baseline status deteriorates, the risk of inflated assessment results is reduced, avoiding misjudgments of the degree of improvement in critically ill patients. The multi-dimensional indicator fusion mechanism can prevent abnormal fluctuations in a single indicator from interfering with the overall assessment, improving the robustness of pulmonary function evaluation in complex cases. The introduction of the exponential decay function enables the follow-up strategy to be automatically adjusted according to the individualized physiological characteristics of patients, realizing a technical transformation from static rules to dynamic optimization in high-risk patient management.

[0108] Preferably, the blood oxygenation index is obtained in the following way:

[0109] Obtain arterial blood oxygen partial pressure, arterial blood carbon dioxide partial pressure, and arterial blood pH value;

[0110] Import arterial blood oxygen partial pressure into the formula Obtain the arterial blood oxygen partial pressure index, O cur Indicates the partial pressure of oxygen in arterial blood, O opt This represents the ideal value of arterial blood oxygen partial pressure;

[0111] Arterial blood carbon dioxide partial pressure is incorporated into the formula. Obtain the partial pressure index of carbon dioxide in arterial blood, CO cur Indicates the partial pressure of carbon dioxide in arterial blood, CO opt This represents the ideal value of arterial blood carbon dioxide partial pressure;

[0112] Import arterial blood pH value into the formula The pH index of arterial blood is obtained from the sample. pH represents the pH value of arterial blood. opt ΔpH represents the standard arterial blood pH value. max This indicates the maximum permissible deviation of arterial blood pH.

[0113] A blood oxygenation index is constructed based on the arterial blood oxygen partial pressure index, arterial blood carbon dioxide partial pressure index, and arterial blood pH index. The blood oxygenation index is then output as follows:

[0114] B = d1O + d2(1-CO) + d3(1-pH) ind )

[0115] Where B represents the blood oxygenation index, O represents the arterial blood oxygen partial pressure index, CO represents the arterial blood carbon dioxide partial pressure index, and pH... ind Indicates the pH index of arterial blood, d i Denotes the weight coefficients and ∑d i =1, the higher the blood oxygenation index, the better the oxygenation status.

[0116] The arterial blood oxygen partial pressure index is a standardized parameter calculated by the ratio of the current measured value to the ideal value. Specifically, it can be achieved by dividing the measured value by a preset ideal value and combining this with a threshold limit, quantifying the degree to which the oxygen partial pressure approaches the ideal state. The arterial blood carbon dioxide partial pressure index is a parameter calculated by the ratio of the measured value to the ideal value and then inversely processed. Specifically, it can be achieved by dividing the measured value by the ideal value and then subtracting the ideal value from 1, reflecting the negative impact of carbon dioxide retention. The arterial blood pH index is a parameter calculated by dividing the absolute deviation of the current pH value from the standard value by the maximum permissible deviation. Specifically, it can be achieved by the ratio of the absolute value of the difference between the measured value and the standard value to a preset upper limit of deviation, characterizing the degree of acid-base imbalance. The weighting coefficient is a proportional factor used to adjust the contribution of each parameter to the final index. Specifically, it can be achieved using a weighted allocation method after normalization to ensure the balance of the influence of each indicator.

[0117] Specifically, three basic physiological parameters—arterial blood oxygen partial pressure, carbon dioxide partial pressure, and pH—were collected and standardized. For oxygen partial pressure, a ratio of the measured value to the ideal value, combined with a threshold cutoff, was used to reflect the actual oxygenation level while avoiding interference from abnormally high values. For carbon dioxide partial pressure, a ratio was also used, but excessively high partial pressure was converted into a negative indicator using the (1-CO) form, strengthening its inhibitory effect on oxygenation. For pH, the degree of acid-base imbalance was quantified by calculating the deviation ratio from the standard value. These three standardized indicators reflect blood oxygenation status from three dimensions: oxygen supply, carbon dioxide clearance, and internal environment stability. A comprehensive index was then generated through weighted summation. The design with a total weight coefficient of 1 ensures a reasonable distribution of the influence of each dimension, avoiding a single parameter dominating the assessment results, thereby achieving a dynamic quantitative assessment of the patient's individualized oxygenation status.

[0118] Compared to existing technologies, traditional methods typically assess oxygenation status using only single parameters such as arterial blood oxygen partial pressure or oxygenation index, neglecting the influence of carbon dioxide partial pressure and acid-base balance on overall oxygenation. This proposed method, however, quantifies and integrates the positive contribution of oxygen partial pressure, the negative effect of carbon dioxide retention, and the inhibitory effect of pH abnormalities through multi-parameter fusion and standardization, forming a multidimensional dynamic evaluation index. Existing technologies lack a systematic compensation mechanism for excessively high carbon dioxide partial pressure and acid-base imbalance, leading to assessment biases in patients with compensated respiratory failure or mixed acid-base imbalances. This proposed method, through reverse processing and bias ratio calculation, accurately reflects the true oxygenation level under these complex pathological conditions.

[0119] Through the aforementioned technical solution, this application addresses the limitations of traditional assessment methods that rely on single indicators. By integrating multi-dimensional physiological parameters and implementing dynamic standardization, it achieves a comprehensive quantitative assessment of blood oxygenation status. For patients with carbon dioxide retention or acid-base imbalance, it accurately identifies abnormal oxygenation status, avoiding the risk of masking overall oxygenation dysfunction due to a single normal indicator. Furthermore, the normalized design of the weighting coefficients ensures the comparability of assessment results across different patient groups or pathological states, providing a reliable basis for monitoring individualized treatment responses.

[0120] Preferably, the lung ultrasound scoring index is obtained in the following way:

[0121] The ratio of the diaphragm thickness at the end of inspiration to the diaphragm thickness at the end of inspiration and the diaphragm thickness at the end of inspiration is calculated to obtain the diaphragm thickening rate.

[0122] Import the diaphragm thickening rate into the formula The diaphragm thickening rate index is obtained from the data. DTF represents the diaphragm thickening rate. opt This indicates the reference value for the diaphragm thickening rate;

[0123] Import the diaphragmatic movement (the distance the diaphragm moves between deep expiration and inspiration, in millimeters) into the formula. The diaphragm movement index (DE) is obtained from the data. DE represents the degree of diaphragm movement. opt This indicates the reference value for diaphragmatic mobility;

[0124] Summarize the LUS scores under the 12-zone method The LUS score index is obtained by comparing it with the maximum sum of scores.

[0125] A lung ultrasound scoring model is constructed based on the LUS score index, diaphragm movement index, and diaphragm thickening rate index to output a lung ultrasound scoring index. The lung ultrasound scoring model is expressed as follows:

[0126] L=e1(1-LUS ind )+e2DE ind +e3DTF ind

[0127] Where L represents the Lung Ultrasound Score Index, LUS ind Denotes the LUS rating index, DE ind The Diaphragmatic Movement Index (DTF) represents the diaphragmatic movement index. ind The diaphragm thickening rate index, e i Denotes the weight coefficients and ∑e i =1, where a higher lung ultrasound score index indicates better oxygenation.

[0128] The diaphragm thickening rate quantifies diaphragmatic contraction function by calculating the rate of change in diaphragm thickness during the respiratory cycle. Specifically, it can be achieved by using ultrasound to measure the percentage difference in diaphragm thickness between the end of inspiration and end of expiration relative to the end-expiratory thickness, dynamically reflecting the diaphragm's mechanical properties. The diaphragm movement index assesses diaphragmatic mobility by standardizing the diaphragm displacement distance between deep expiration and inspiration. This can be achieved by using millimeter-level displacement measurements combined with pre-set reference values ​​for normalization, eliminating the influence of individual anatomical differences on the assessment results. The LUS score index characterizes the extent of lung lesions by the ratio of the sum of the 12 lung ultrasound scores to the maximum possible score. This can be achieved by summing the scores from 0 to 3 for each region and dividing by 36 for standardization, systematically assessing the spatial distribution characteristics of lung structural abnormalities.

[0129] Specifically, by continuously measuring changes in diaphragmatic thickness during the respiratory cycle and calculating the thickening rate, which is then converted into an exponential form, the dynamic trends of diaphragmatic contraction function can be captured. Simultaneously, comparing the diaphragmatic movement distance with reference values ​​to generate a movement index allows for a quantitative assessment of whether the diaphragmatic movement amplitude meets physiological requirements. Combining this with the ratio of the total LUS score across 12 regions systematically reflects the spatial accumulation of pulmonary consolidation, exudation, and other lesions. Finally, by weighted fusion of the three indices, a comprehensive evaluation index reflecting both diaphragmatic function and lung structural changes is constructed, achieving a shift from single-structure assessment to a synergistic structure-function analysis. For example, a decrease in the diaphragmatic thickening rate index while the movement index remains normal may indicate a compensatory breathing pattern; conversely, a simultaneous increase in the LUS score index suggests worsening lung structural damage requiring priority intervention.

[0130] Compared to existing technologies, traditional methods rely solely on single static parameters such as LUS scores or diaphragm thickness, failing to distinguish between compensated and decompensated diaphragmatic functions and lacking quantitative assessment of lesion spatial distribution. This proposed approach establishes a time-series assessment model of diaphragmatic biomechanics by combining dynamic thickening rate calculation with mobility standardization. Furthermore, it introduces a multi-regional scoring system to eliminate the interference of local lesions on overall assessment, overcoming the shortcomings of traditional ultrasound methods, such as limited parameter dimensions and insufficient dynamic response.

[0131] Through the above technical solution, this application can monitor the coupling relationship between diaphragmatic contraction function and changes in lung structure in real time, accurately identify the difference between compensatory breathing patterns and pathological functional failure, and avoid assessment bias caused by misjudgment of local lesions. At the same time, by quantifying the synergistic change trend of diaphragmatic mobility and thickening rate, it provides data support for the development of individualized respiratory rehabilitation programs, effectively improving the assessment accuracy of complex pulmonary pathological states.

[0132] Preferably, the follow-up frequency optimization model is expressed as:

[0133] F tar=F0+δ(μ-P cur )

[0134] Among them, F tar The target follow-up frequency is represented by F0, the baseline follow-up frequency is represented by δ, the follow-up frequency adjustment scaling factor is represented by μ, and the reference value of the lung function evaluation coefficient is represented by P. cur This represents the current lung function evaluation coefficient.

[0135] The baseline follow-up frequency refers to the initial follow-up interval preset based on the patient's underlying condition, which can be achieved using routine monitoring cycles recommended by clinical guidelines, such as once a week or once a day. The follow-up frequency adjustment scaling factor is a proportional coefficient used to control the magnitude of changes in the follow-up frequency. It can be set to a fixed value through clinical data fitting or expert experience, and its function is to convert the deviation of the pulmonary function evaluation coefficient into an operable frequency adjustment. The pulmonary function evaluation coefficient reference value refers to the preset threshold for achieving the target pulmonary function status, which can be determined using data from healthy individuals or treatment target values, for example, 0.8, used to determine whether the patient's current status deviates from the expected target. The current pulmonary function evaluation coefficient refers to the real-time pulmonary function assessment result calculated through a multi-dimensional model, with a value range, for example, from 0 to 1; a higher value indicates a better pulmonary function status.

[0136] Specifically, this model directly links the dynamic assessment results of lung function with the follow-up strategy through a linear relationship. When the current lung function evaluation coefficient is lower than the reference value, the difference term (μ-P) is applied. cur If the scaling factor is positive, the target follow-up frequency is increased from the baseline value, thereby strengthening the monitoring density for patients whose condition is worsening. For example, if the baseline follow-up frequency is once a week, the scaling factor is 0.5 times / week per unit coefficient difference, and the current coefficient differs from the reference value by 0.4, then the target frequency is increased to 1.2 times per week. Conversely, if the current coefficient is higher than the reference value, the difference term is negative, and the target frequency is reduced accordingly, minimizing resource consumption for stable patients. The scaling factor limits the adjustment range, avoiding drastic changes in follow-up frequency due to short-term index fluctuations. For example, when the coefficient difference is 0.2, the frequency is adjusted by only 0.1 times / week, ensuring the stability of the strategy.

[0137] Compared to existing technologies, traditional follow-up strategies are typically based on fixed periods or subjective judgment, failing to quantify real-time changes in patients' lung function. For example, existing technologies may use daily follow-up for all critically ill patients and weekly follow-up for mildly ill patients, but cannot dynamically adjust for individuals with abnormal treatment responses or sudden changes in their condition. This approach establishes a mathematically driven model to transform multi-dimensional lung function evaluation results into frequency adjustment quantities, enabling the follow-up strategy to automatically optimize according to the patient's condition. This avoids monitoring delays caused by fixed periods for high-risk patients and reduces unnecessary follow-up for patients with stable conditions.

[0138] Through the above technical solution, this application achieves dynamic optimization of follow-up strategies based on the patient's real-time lung function status. For patients whose lung function does not meet the target, increasing the follow-up frequency can promptly detect signs of disease deterioration and initiate intervention; for patients whose lung function improves, reducing the follow-up frequency can reduce the consumption of medical resources. This model, through a quantitative adjustment mechanism, solves the problems of monitoring lag or resource waste caused by rigid follow-up in traditional methods, and is particularly suitable for complex cases with significant individual differences in treatment response.

[0139] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0140] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating lung function, characterized in that, Includes the following steps: A prone ventilation model was constructed based on the number of prone ventilations and the total duration of prone ventilation, and the prone ventilation coefficient was output. Based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment, a blood oxygenation evaluation model is constructed and a blood oxygenation evaluation coefficient is output. Based on the mean difference of lung ultrasound score index before and after lung treatment within the treatment cycle, the difference of lung ultrasound score index before and after the treatment cycle, and the lung ultrasound score index after the treatment cycle, a lung ultrasound evaluation model is constructed and lung ultrasound evaluation coefficients are output. A lung function evaluation model was constructed based on the prone ventilation coefficient, the blood oxygenation evaluation coefficient under the patient's APACHE II score, and the lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output. A follow-up frequency optimization model is constructed based on lung function evaluation coefficients and baseline follow-up frequency to determine the target follow-up frequency.

2. The method for evaluating lung function according to claim 1, characterized in that, The follow-up frequency optimization model is expressed as follows: F tar =F0+δ(μ-P cur ) Among them, F tar The target follow-up frequency is represented by F0, the baseline follow-up frequency is represented by δ, the follow-up frequency adjustment scaling factor is represented by μ, and the reference value of the lung function evaluation coefficient is represented by P. cur This represents the current lung function evaluation coefficient.

3. The method for evaluating lung function according to claim 1, characterized in that, The steps for constructing a prone ventilation model and outputting the prone ventilation coefficient based on the number of prone ventilations and the total duration of prone ventilation are as follows: The number of prone ventilations was obtained by performing a maximum-min normalization process; The duration factor is obtained by comparing the total duration of prone ventilation with the maximum permissible total duration. A prone ventilation model is constructed based on the frequency factor and duration factor to output the prone ventilation coefficient. The prone ventilation model is expressed as follows: Where α represents the prone ventilation coefficient, N ind T represents the frequency factor. ind The duration factor is indicated by the prone ventilation coefficient, which is directly proportional to the ventilation effect.

4. The method for evaluating lung function according to claim 1, characterized in that, The steps for constructing a blood oxygenation evaluation model and outputting the blood oxygenation evaluation coefficient based on the mean difference in blood oxygenation index before and after lung treatment within the treatment cycle, the difference in blood oxygenation index before and after the treatment cycle, and the post-treatment blood oxygenation index are as follows: The mean difference of blood oxygenation index before and after lung treatment within the treatment cycle, the difference of blood oxygenation index before and after the treatment cycle, and the blood oxygenation index after treatment are processed by maximum-min normalization to obtain the mean blood oxygenation factor, the difference of blood oxygenation factor, and the blood oxygenation factor. A blood oxygenation evaluation model is constructed based on the mean blood oxygenation factor, the difference blood oxygenation factor, and the blood oxygenation factor, and outputs the blood oxygenation evaluation coefficient. The blood oxygenation evaluation model is expressed as follows: β=a1ΔB mean,ind +a2ΔB total,ind +a3B end,ind Where β represents the blood oxygenation evaluation coefficient, ΔB mean,ind ΔB represents the mean blood oxygenation factor. total,ind B represents the oxygenation difference factor. end,ind Indicates blood oxygenation factor, a i Represents the weight coefficients and ∑a i =1, the blood oxygen evaluation coefficient is directly proportional to the blood oxygen improvement effect.

5. The method for evaluating lung function according to claim 1, characterized in that, The steps for constructing a lung ultrasound evaluation model and outputting lung ultrasound evaluation coefficients based on the mean difference in lung ultrasound score index before and after treatment within the treatment cycle, the difference in lung ultrasound score index before and after the treatment cycle, and the lung ultrasound score index after the treatment cycle are as follows: The mean difference of lung ultrasound score index before and after treatment within the treatment cycle, the difference of lung ultrasound score index before and after treatment cycle, and the lung ultrasound score index after treatment cycle are processed by maximum-min normalization to obtain the ultrasound mean factor, ultrasound difference factor, and ultrasound factor. A lung ultrasound evaluation model is constructed based on the ultrasound mean factor, ultrasound difference factor, and ultrasound factor to output lung ultrasound evaluation coefficients. The lung ultrasound evaluation model is expressed as follows: γ=b1ΔL mean,ind +b2ΔL total,ind +b3L end,ind Where γ represents the lung ultrasound evaluation coefficient, ΔL mean,ind ΔL represents the ultrasound mean factor. total,ind L represents the ultrasound difference factor. end,ind Indicates the ultrasound factor, b i Represents the weight coefficients and ∑b i =1, the lung ultrasound evaluation coefficient is directly proportional to the improvement effect of blood oxygenation.

6. The method for evaluating lung function according to claim 1, characterized in that, The steps for constructing a pulmonary function evaluation model and outputting pulmonary function evaluation coefficients based on prone ventilation coefficient, patient APACHE II score, and lung ultrasound evaluation coefficient are as follows: The APACHE II score of the patient was processed by maximum-min normalization to obtain the APACHE II score index; A lung function evaluation model was constructed based on the APACHE II score index, prone ventilation coefficient, blood oxygenation evaluation coefficient, and lung ultrasound evaluation coefficient, and the lung function evaluation coefficient was output. The lung function assessment model is represented as follows: P=c1α+c2βexp(-k A A)+c3γexp(-k A A) Wherein, P represents the lung function evaluation coefficient, α represents the prone ventilation coefficient, β represents the blood oxygenation evaluation coefficient, γ represents the lung ultrasound evaluation coefficient, A represents the APACHE II score index, and k A c represents the APACHE II score index decay factor. i Denotes the weight coefficients and ∑c i =1, meaning the improvement in blood oxygenation is directly proportional to the lung function evaluation coefficient.

7. The method for evaluating lung function according to any one of claims 1-6, characterized in that, The blood oxygenation index is obtained as follows: Obtain arterial blood oxygen partial pressure, arterial blood carbon dioxide partial pressure, and arterial blood pH value; Import arterial blood oxygen partial pressure into the formula Obtain the arterial oxygen partial pressure index, O cur Indicates the partial pressure of oxygen in arterial blood, O opt This represents the ideal value of arterial blood oxygen partial pressure; Arterial blood carbon dioxide partial pressure is incorporated into the formula. Obtain the partial pressure index of carbon dioxide in arterial blood, CO cur Indicates the partial pressure of carbon dioxide in arterial blood, CO opt This represents the ideal value of arterial blood carbon dioxide partial pressure; Import arterial blood pH value into the formula The pH index of arterial blood is obtained from the sample. pH represents the pH value of arterial blood. opt ΔpH represents the standard arterial blood pH value. max This indicates the maximum permissible deviation of arterial blood pH. A blood oxygenation index is constructed based on the arterial blood oxygen partial pressure index, arterial blood carbon dioxide partial pressure index, and arterial blood pH index. The blood oxygenation index is then output as follows: B=d1O+d2(1-CO)+d3(1-pH ind ) Where B represents the blood oxygenation index, O represents the arterial blood oxygen partial pressure index, CO represents the arterial blood carbon dioxide partial pressure index, and pH... ind Indicates the pH index of arterial blood, d i Denotes the weight coefficients and ∑d i =1, the higher the blood oxygenation index, the better the oxygenation status.

8. The method for evaluating lung function according to any one of claims 1-6, characterized in that, The lung ultrasound scoring index is obtained as follows: The ratio of the diaphragm thickness at the end of inspiration to the diaphragm thickness at the end of inspiration and the diaphragm thickness at the end of inspiration is calculated to obtain the diaphragm thickening rate. Import the diaphragm thickening rate into the formula The diaphragm thickening rate index is obtained from the data. DTF represents the diaphragm thickening rate. opt This indicates the reference value for the diaphragm thickening rate; Import diaphragmatic movement into the formula The diaphragm movement index (DE) is obtained from the data. DE represents the degree of diaphragm movement. opt This indicates the reference value for diaphragmatic mobility; Summarize the LUS scores under the 12-zone method The LUS score index is obtained by comparing it with the maximum sum of scores. A lung ultrasound scoring model is constructed based on the LUS score index, diaphragm movement index, and diaphragm thickening rate index to output a lung ultrasound scoring index. The lung ultrasound scoring model is expressed as follows: L=e1(1-LUS ind )+e2DE ind +e3DTF ind Where L represents the Lung Ultrasound Score Index, LUS ind Denotes the LUS rating index, DE ind The Diaphragmatic Movement Index (DTF) represents the diaphragmatic movement index. ind The diaphragm thickening rate index, e i Denotes the weight coefficients and ∑e i =1, where a higher lung ultrasound score index indicates better oxygenation.

Images