A Modeling Method for Cardiac Stroke Volume Measurement Based on Ultrasound Imaging

By using fully automated three-dimensional ultrasound volume imaging as a reference, a calibration model was established to optimize Doppler stroke volume measurement, which solved the measurement error problem of Doppler method in the Chinese population and achieved higher measurement accuracy and quality of diagnosis and treatment.

CN123120360APending Publication Date: 2026-11-17JINING NO 1 PEOPLES HOSPITAL (JINING ACAD OF MEDICAL SCI)
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611454014.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-18
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing transthoracic two-dimensional Doppler ultrasound method has errors in measuring stroke volume, which is particularly difficult to verify accurately in the Chinese population, and lacks reliable reference standards, thus affecting the quality of diagnosis and treatment.

Method used

Using the stroke volume measurement results of fully automated three-dimensional ultrasound volume imaging as the standard, a calibration model was established by analyzing the linear relationship between the diameter of the left ventricular outflow tract and the time integral of the flow velocity, and the measurement method of Doppler method was optimized.

Benefits of technology

This improved the accuracy of Doppler measurement of stroke volume in the Chinese population, reduced measurement bias, and enhanced the quality of diagnosis and treatment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123120360A_ABST
    Figure CN123120360A_ABST
Patent Text Reader

Abstract

A method for modeling stroke volume measurement based on ultrasound imaging includes the following steps: extracting transthoracic two-dimensional Doppler measurement data for each subject from a hospital database, including left ventricular outflow tract diameter and left ventricular outflow tract flow velocity time integral, and calculating the corresponding stroke volume; extracting fully automated three-dimensional ultrasound data for each subject from an ultrasound database, and recording the corresponding stroke volume as ; analyzing the consistency between and , and determining whether the clinical consistency between the two is within an acceptable range based on the calculation results. The ultrasound imaging-based stroke volume measurement modeling method provided by this invention can improve the accuracy of transthoracic two-dimensional Doppler left ventricular outflow tract measurement by using the stroke volume measurement results of fully automated three-dimensional ultrasound volume imaging as a standard.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of cardiac stroke volume measurement technology, specifically relating to a modeling method for cardiac stroke volume measurement based on ultrasound imaging. Background Technology

[0002] Stroke volume (SV) is a core parameter connecting cardiac pumping function and systemic hemodynamic status. Accurate quantitative assessment of SV plays a crucial role in evaluating treatment efficacy in clinical scenarios such as perioperative fluid management, monitoring the effectiveness of chronic heart failure treatment, and cardiac resynchronization therapy. Transthoracic two-dimensional Doppler left ventricular outflow tract (LVOT) measurement (hereinafter referred to as Doppler) is currently the most widely used non-invasive method for SV measurement in clinical practice, with advantages including ease of operation and real-time performance. In clinical practice, when patient positioning is limited or lung air interferes with ultrasound imaging, making other methods difficult to implement, LVOT is often the only feasible non-invasive method for monitoring cardiac function. However, the measurement principle of this method is based on several ideal model assumptions: for example, a circular cross-section of the left ventricular outflow tract, uniform flow velocity distribution, sound pressure parallel to the blood flow direction, and stable heart rhythm. In reality, the cross-section of the left ventricular outflow tract is approximately elliptical, and the flow velocity in LVOT is not completely uniform. Therefore, these factors collectively lead to a certain degree of error in SV measurement using the LVOT method. Since the accuracy of SV measurement directly affects the quality of diagnosis and treatment in actual clinical work, it is necessary to improve the accuracy of SV measurement. However, previous studies validating the accuracy of LVOT (Low Ventricular Transmission Therapy) have mostly been based on Western population samples, mainly focusing on diseased individuals, with fewer studies on healthy individuals. Furthermore, the left ventricular volume and SV of Asian populations are significantly lower than those of Western populations. Therefore, it is currently difficult to systematically validate and improve the accuracy of two-dimensional Doppler LVOT SV measurement in healthy Chinese populations.

[0003] To improve the accuracy of LVOT (Low Volume Doppler) measurement, a reliable reference standard is essential. While cardiac magnetic resonance imaging (CMR) is currently the recognized gold standard for quantitative cardiac assessment, its high equipment cost, long examination time, and high examination fees limit its widespread application in clinical studies in healthy populations. In recent years, three-dimensional ultrasound technology has made significant progress. Fully automated three-dimensional ultrasound volume imaging (HeartModel) uses adaptive analysis algorithms to automatically identify the endocardial boundary. Multiple studies have confirmed that its SV measurement results are highly consistent with CMR. However, due to the higher cost of its equipment compared to ordinary cardiac ultrasound diagnostic instruments, it is less widely used in practical applications compared to LVOT. Nevertheless, it can serve as a non-invasive, high-precision, and accessible reference standard in clinical studies in healthy populations. Therefore, to improve the accuracy of LVOT measurement in the Chinese population, it is necessary to optimize the LVOT measurement method using the SV measurement results of fully automated three-dimensional ultrasound volume imaging (HeartModel) as the standard. This invention addresses this technical problem. Summary of the Invention

[0004] This invention provides a modeling method for measuring stroke volume based on ultrasound imaging. It can improve the accuracy of stroke volume measurement in the Chinese population by using the results of fully automated three-dimensional ultrasound volume imaging as the standard, thereby helping to improve the quality of diagnosis and treatment.

[0005] A method for modeling cardiac stroke volume measurement based on ultrasound imaging includes the following steps:

[0006] Transthoracic two-dimensional Doppler ultrasound data for each subject were extracted from the hospital database, including left ventricular outflow tract diameter and left ventricular outflow tract velocity time integral, and the corresponding stroke volume was calculated based on these data. ;

[0007] Fully automated three-dimensional ultrasound data for each subject were extracted from the ultrasound database, and the corresponding stroke volume was recorded as follows: ;

[0008] analyze and The consistency between the two was assessed, and the clinical consistency between them was judged to be within an acceptable range based on the calculation results.

[0009] calculate and Bias value of measurement results and analysis and The correlation;

[0010] Using the diameter of the left ventricular outflow tract and the time integral of the left ventricular outflow tract velocity as independent variables, analyze their effects on... The degree of impact;

[0011] The optimal cutoff value for the diameter of the left ventricular outflow tract was calculated, and several subjects were divided into high-risk and low-risk groups based on this value.

[0012] For the high-risk group, the left ventricular outflow tract diameter, left ventricular outflow tract velocity time integral, Establish a calibration model;

[0013] The safety of the correction results is evaluated based on the obtained correction model.

[0014] Furthermore, the analysis and The correlation is determined in the following ways:

[0015] Using the Pearson product-moment correlation coefficient formula, the correlation coefficients of all subjects were analyzed. , The Pearson correlation coefficient r is calculated using this method; where, , r is used to measure and The direction and intensity of the linear correlation between them;

[0016] The slope of the linear regression equation can be calculated using the following formula. :

[0017] ;

[0018] in, yes The sample standard deviation yes The sample standard deviation;

[0019] The intercept of the linear regression equation can be calculated using the following formula. :

[0020] ;

[0021] in, yes The sample mean; yes The sample mean;

[0022] Based on the calculation results, the corresponding linear regression equation is obtained:

[0023] .

[0024] Furthermore, the analysis and The consistency between the two, and the method for judging whether the clinical consistency between them is within an acceptable range based on the calculation results, is as follows:

[0025] according to and The specific values ​​are calculated using Bland-Altman spectroscopy to determine the average bias, 95% compliance limit, compliance range width, and percentage error. If the calculated percentage error is greater than 30%, it is considered... and The consistency between them is poor, and they cannot be directly interchanged in clinical practice.

[0026] Furthermore, the analysis uses the left ventricular outflow tract diameter and the time integral of the left ventricular outflow tract velocity as independent variables to examine their effects on... The degree of influence is determined as follows;

[0027] For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The diameter of the left ventricular outflow tract. The intercept is... For coefficients;

[0028] For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The time integral of the left ventricular outflow tract velocity. The intercept is... For coefficients;

[0029] Will , At the same time, as an independent variable, and Perform linear regression analysis; if the results show that both are related to... Independent correlation means , All Independent influencing factors.

[0030] Furthermore, the process of calculating the optimal cutoff value for the left ventricular outflow tract diameter and using this value to divide several subjects into high-risk and low-risk groups is as follows:

[0031] Using the left ventricular outflow tract diameter for each subject as a state variable and the clinically significant bias value as a state variable, ROC curves were plotted, and the area under the curve and Youden index were calculated accordingly.

[0032] Based on the size of the area under the curve, determine the relationship between the diameter of the left ventricular outflow tract and the curve. The ability to distinguish between large and small values; and the left ventricular outflow tract diameter corresponding to the largest Youden index is taken as the optimal cutoff value;

[0033] Subjects whose left ventricular outflow tract diameter was smaller than the optimal cutoff value were classified as the low-risk group, while the remaining subjects were classified as the high-risk group.

[0034] Furthermore, the method for establishing the calibration model is as follows:

[0035] Based on the obtained linear regression equation, the following two methods are used to... Perform the correction and record the correction result as follows: :

[0036] ;

[0037] in, It is the corrected stroke volume based on the diameter of the left ventricular outflow tract.

[0038] ;

[0039] in, It is the corrected stroke volume based on the time integral of the left ventricular outflow tract velocity.

[0040] Based on the above calculation model, calculate the corrected bias values ​​respectively. :

[0041] ;

[0042] Based on the calculation results, the average bias value and bias rate of the whole sample and the high-risk group are calculated, and the best correction model is selected accordingly.

[0043] Furthermore, the process for assessing the security of the correction results is as follows:

[0044] The selected optimal correction model is used to calculate the data in the low-risk group that are below the clinically significant bias value. Based on the calculation results, it is determined whether any new significant bias will be added.

[0045] Furthermore, the aforementioned The calculation method is as follows:

[0046] according to and The size, for Perform the calculation, if > Then the following calculations are performed:

[0047] .

[0048] The technical effects of this invention are as follows:

[0049] This application uses the SV measurement results from fully automated three-dimensional ultrasound volume imaging. As a standard, the SV measurement results of the Doppler method are compared. Correction is performed. This application calculates the left ventricular outflow tract diameter and left ventricular outflow tract velocity time integral in the Doppler method, and compares them with the bias value. The linear relationship between the parameters is used to establish a subsequent calibration model. Furthermore, both calibration models in this application are based on conventional measurement parameters, requiring no additional hardware or measurement procedures to obtain more accurate measurement results. Thus, the Doppler measurement results are calibrated in the above manner, improving measurement accuracy. Attached Figure Description

[0050] Figure 1 This is a table showing the baseline characteristics and measurement indicators of the research object of this invention.

[0051] Figure 2 In this invention and The correlation scatter plot.

[0052] Figure 3 This is a Bland-Altman consistency analysis diagram from the present invention.

[0053] Figure 4 This is a scatter plot showing the correlation between the LVOT diameter and ΔSV in this invention.

[0054] Figure 5 This is a scatter plot showing the correlation between VTI and ΔSV in this invention.

[0055] Figure 6 This is the ROC curve diagram in this invention. Detailed Implementation

[0056] The technical solution of the present invention will now be clearly and completely described in conjunction with specific embodiments and accompanying drawings.

[0057] See Figures 1-6 A method for modeling cardiac stroke volume measurement based on ultrasound imaging includes the following steps:

[0058] Transthoracic two-dimensional Doppler ultrasound data for each subject were extracted from the hospital database, including left ventricular outflow tract diameter and left ventricular outflow tract velocity time integral, and the corresponding stroke volume was calculated based on these data. ;

[0059] Fully automated three-dimensional ultrasound data for each subject were extracted from the ultrasound database, and the corresponding stroke volume was recorded as follows: ;

[0060] analyze and The consistency between the two was assessed, and the clinical consistency between them was judged to be within an acceptable range based on the calculation results.

[0061] calculate and Bias value of measurement results and analysis and The correlation;

[0062] Using the diameter of the left ventricular outflow tract and the time integral of the left ventricular outflow tract velocity as independent variables, analyze their effects on... The degree of impact;

[0063] The optimal cutoff value for the diameter of the left ventricular outflow tract was calculated, and several subjects were divided into high-risk and low-risk groups based on this value.

[0064] For the high-risk group, the left ventricular outflow tract diameter, left ventricular outflow tract velocity time integral, Establish a calibration model;

[0065] The safety of the correction results is evaluated based on the obtained correction model.

[0066] See Figure 1 The parameters in this protocol are named as follows: LVOT: Left ventricular outflow tract; VTI: Flow velocity-time integral; SV: Stroke volume; EDV: Left ventricular end-diastolic volume; ESV: Left ventricular end-systolic volume; EF: Ejection fraction; EDVI: Left ventricular end-diastolic volume index (EDV / body surface area); ESVI: Left ventricular end-systolic volume index (ESV / body surface area); SVI: Stroke volume index (SV / body surface area); HR: Heart rate; CO: Cardiac output; CI: Cardiac index. Continuous variables in this protocol are expressed as mean + standard deviation.

[0067] Furthermore, the analysis and The correlation is determined in the following ways:

[0068] Using the Pearson product-moment correlation coefficient formula, the correlation coefficients of all subjects were analyzed. , The Pearson correlation coefficient r is calculated using this method; where, , r is used to measure and The direction and intensity of the linear correlation between them;

[0069] The slope of the linear regression equation can be calculated using the following formula. :

[0070] ;

[0071] in, yes The sample standard deviation yes The sample standard deviation;

[0072] The intercept of the linear regression equation can be calculated using the following formula. :

[0073] ;

[0074] in, yes The sample mean; yes The sample mean;

[0075] Based on the calculation results, the corresponding linear regression equation is obtained:

[0076] .

[0077] In this embodiment, it can be based on Figure 1 The data collected is used to calculate the relevant values ​​mentioned above. In this embodiment... The value was (74.5±14.1) mL. The result was (63.7 ± 11.3) mL. The difference in SV between the two methods. The value was (10.8±11.1) m, ranging from -6.6 to 37.8 m. The t-test showed a significant difference between the two values. The calculation results are higher than The mean difference in the calculated results was 10.80 mL, 95% CI: 7.90–13.71 mL, t = 7.290, P < 0.001, indicating that the Doppler measurement... There is a systemic overestimation.

[0078] See Figure 2 According to Pearson's correlation analysis, and It showed a significant positive correlation (r=0.640, (P<0.001), based on the above values ​​and calculation method, the following can be calculated in this embodiment: and The regression equation is:

[0079] ;

[0080] in, Figure 2 The horizontal axis in the middle is The unit is mL; the vertical axis is The unit is mL; the red solid line is the regression line, and the gray dashed line is the y=x reference line.

[0081] The r mentioned above is used to measure the direction and strength of the linear correlation between two continuous variables. The value range is [-1, 1]. Positive numbers represent positive correlation, and negative numbers represent negative correlation. The closer the absolute value is to 1, the stronger the linear correlation. In this example, r is 0.64, which belongs to a moderately strong positive linear correlation. The value is the square of r, in this embodiment. =0.409 means: Of the numerical differences, 40.9% can be attributed to... This can be explained by linear changes. P < 0.001 indicates that: it is believed that... and The positive correlation between them truly exists.

[0082] Furthermore, the analysis and The consistency between the two, and the method for judging whether the clinical consistency between them is within an acceptable range based on the calculation results, is as follows:

[0083] according to and The specific values ​​are calculated using Bland-Altman spectroscopy to determine the average bias, 95% compliance limit, compliance range width, and percentage error. If the calculated percentage error is greater than 30%, it is considered... and The consistency between them is poor, and they cannot be directly interchanged in clinical practice.

[0084] See Figure 3 In this embodiment, the average bias is 10.80 mL; the 95% concordance limit is -10.93 m to 32.53 m, meaning that for 95% of subjects, the measurement difference between the two methods will fall within this range, i.e., in most cases, the deviation between the two methods will not exceed this range; the concordance range width is 43.47 mL, meaning that the total span of the 95% concordance limit is 43.47 mL, and the narrower the span, the better the concordance; the percentage error is 31.5%.

[0085] In this embodiment, 30% is the clinical consistency threshold, which is an internationally accepted standard. Since the percentage error in this embodiment is greater than 30%, it is therefore considered... and The consistency between the two is poor, with significant discrepancies, making direct interchangeability in clinical practice impossible. Therefore, further research is needed. Make corrections to make it more accurate.

[0086] Figure 3 The horizontal axis in the middle is and The mean; the vertical axis is The solid red line represents the mean of bias; the dashed red line represents the 95% consensus threshold; and the dashed blue line represents the clinically significant bias threshold ±15 mL.

[0087] See Figure 4 , Figure 5 Furthermore, the analysis uses the left ventricular outflow tract diameter and the time integral of the left ventricular outflow tract velocity as independent variables to examine their effects on... The degree of influence is determined as follows;

[0088] For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The diameter of the left ventricular outflow tract. The intercept is... For coefficients;

[0089] For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The time integral of the left ventricular outflow tract velocity. The intercept is... For coefficients;

[0090] Will , At the same time, as an independent variable, and Perform linear regression analysis; if the results show that both are related to... Independent correlation means , All Independent influencing factors.

[0091] The purpose of the above content is to analyze , Will these two parameters affect...? and Measurement bias The size of . For single-factor analysis, in this embodiment, and The regression equation is: ( =0.109, r=0.331, P=0.013). The above indicates that the larger the LVOT diameter of the subject, the greater the SV difference ΔSV between the two methods. P=0.013<0.05, indicating that the above positive correlation is statistically significant and not due to random error.

[0092] and The regression equation is: ( =0.096).

[0093] Because univariate analysis can only explain the individual factors (LVOT diameter and VTI) and biases. While there is some correlation, LVOT diameter and VTI themselves may be interrelated; for example, a person with a wide outflow channel may also have a higher VTI. Therefore, it's difficult to determine the relationship between VTI and deviation. The correlation might be that people with larger VTIs also have larger LVOT diameters, indirectly leading to a larger bias. Therefore, it's necessary to include both independent variables in the model for calculation to see if a single factor can still independently influence the outcome. .

[0094] Figure 4 In the middle, the horizontal axis is Unit: cm; Vertical axis: The solid red line represents the regression line, and the dashed blue line represents the clinically significant bias value ±15 mL. Figure 5 In the middle, the horizontal axis represents VTI, in cm, and the vertical axis represents... The solid red line represents the regression line, and the dashed blue line represents the clinically significant bias value ±15 mL.

[0095] See Figure 6 Furthermore, the process of calculating the optimal cutoff value for the left ventricular outflow tract diameter and using this value to divide several subjects into high-risk and low-risk groups is as follows:

[0096] Using the left ventricular outflow tract diameter for each subject as a state variable and the clinically significant bias value as a state variable, ROC curves were plotted, and the area under the curve and Youden index were calculated accordingly.

[0097] Based on the size of the area under the curve, determine the relationship between the diameter of the left ventricular outflow tract and the curve. The ability to distinguish between large and small values; and the left ventricular outflow tract diameter corresponding to the largest Youden index is taken as the optimal cutoff value;

[0098] Subjects whose left ventricular outflow tract diameter was smaller than the optimal cutoff value were classified as the low-risk group, while the remaining subjects were classified as the high-risk group.

[0099] As discussed above, since a larger LVOT diameter leads to greater measurement bias, a specific diameter cutoff value is needed to predict whether patients will exhibit clinically unacceptable bias. Because VTI values ​​are influenced by many other factors and are inherently unstable, this embodiment uses LVOT diameter as the predictor variable. ROC curves were plotted for the state variable. The LVOT diameter is the anatomical width of the left ventricular outflow tract. Its value is very stable within the same disease cycle and does not change much with heart rate or myocardial contractility. The measurement boundary is easy to identify and the operation is simple. Repeated measurements of the same patient show little variation.

[0100] See Figure 6 The results in this embodiment show that the AUC (Area Under the Curve) is 0.748 (95% CI: 0.616-0.869), suggesting that the LVOT diameter is important for prediction. Significant bias has moderate discriminative power. (Based on the Youden index) The optimal cutoff value was 2.10 cm, corresponding to a sensitivity of 65.0% and a specificity of 80.6%. The AUC (Area Under the Curve) is calculated by integrating the area under the ROC curve and the horizontal axis. An AUC of 0.748 indicates that, given a random sample of one subject with significant bias and one without significant bias, the probability of correctly distinguishing the more biased subject based on the LVOT diameter is 74.8%. The Youden index is calculated as: Sensitivity + Specificity - 1; a higher value indicates better overall discrimination performance.

[0101] Due to transthoracic two-dimensional Doppler ultrasound Measurement is widely used in clinical practice. However, uniformly correcting SV values ​​for all patients could potentially introduce new errors into patients with already small deviations. Therefore, to avoid this situation and improve the accuracy of correction, this embodiment divides the study subjects into high-risk and low-risk groups based on the LVOT diameter cutoff value of 2.10 cm determined by the ROC curve. Only the high-risk group is subjected to [the correction / correction]. Numerical correction. In this embodiment, after risk stratification, the high-risk group ( A total of 20 cases (35.7% of the total sample) with a diameter >2.10 cm were included, representing a significant bias rate of 65.0%; this portion was used as the primary target for correction. Low-risk group ( A total of 36 cases (64.3% of the total sample) were <2.10 cm, with a significant bias rate of 19.4%; no correction was performed.

[0102] Figure 6 The horizontal axis represents 1-specificity; the vertical axis represents sensitivity; the blue curve is the ROC curve; the gray dashed line is the random guessing line; the red dot is the optimal cutoff point, and the corresponding value is the optimal cutoff value.

[0103] Furthermore, the method for establishing the calibration model is as follows:

[0104] Based on the obtained linear regression equation, the following two methods are used to... Perform the correction and record the correction result as follows: :

[0105] ;

[0106] in, It is the corrected stroke volume based on the diameter of the left ventricular outflow tract.

[0107] ;

[0108] in, It is the corrected stroke volume based on the time integral of the left ventricular outflow tract velocity.

[0109] Based on the above calculation model, calculate the corrected bias values ​​respectively. :

[0110] ;

[0111] Based on the calculation results, the average bias value and bias rate of the whole sample and the high-risk group are calculated, and the best correction model is selected accordingly.

[0112] In this embodiment, before correction, the overall significant bias rate was 35.7%, and the significant bias rate of the high-risk group was 65.0%.

[0113] for Regarding VTI, after correction, the overall mean bias decreased to 5.65 mL, the overall significant bias rate decreased to 17.9%, the mean bias in the high-risk group decreased to 7.59 mL, and the significant bias rate decreased to 15.0%.

[0114] Based on the above, it can be seen that the mean bias and significant bias rates of the two correction schemes were the same in the high-risk group (7.59 mL and 15.0%, respectively). The main difference lies in the overall level. Correction can more effectively reduce systematic bias, while VTI correction is superior in reducing significant clinical bias.

[0115] Since reducing the significant bias rate has higher clinical value than further reducing the mean bias in clinical applications, and VTI correction can avoid overcorrection for some cases, this embodiment selects the regression formula with VTI as the variable as the best correction model.

[0116] Furthermore, the process for assessing the security of the correction results is as follows:

[0117] The selected optimal correction model is used to calculate the data in the low-risk group that are below the clinically significant bias value. Based on the calculation results, it is determined whether any new significant bias will be added.

[0118] In this embodiment, for the low-risk group ( There was no significant bias in <2.10cm) Corrected calculations were performed on individuals, and the results showed that only 2 out of the 29 originally safe individuals developed new significant biases (| (≥15mL), the absolute values ​​of bias were 15.3mL and 17.9mL, respectively, with a mean of 16.6mL. Therefore, using the above correction formula, even if an individual without significant bias is accidentally injured, the degree of bias is relatively mild, and it has good safety.

[0119] Furthermore, the aforementioned The calculation method is as follows:

[0120] according to and The size, for Perform the calculation, if > Then the following calculations are performed:

[0121] .

[0122] The difference between the stroke volume measurements of the two methods, also known as the bias value, indicates a greater absolute value and thus a more severe measurement deviation between the two methods.

[0123] This scheme uses the SV measurement results from fully automated three-dimensional ultrasonic volume imaging. As a standard, the SV measurement results of the Doppler method are compared. By performing corrections to obtain the optimal correction formula, the bias between the two can be reduced, thereby improving the accuracy of Doppler measurements.

[0124] The above are merely exemplary embodiments of the present invention and do not limit the patent scope of the present invention. All equivalent transformations made using the contents of the present invention specification and drawings under the technical concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for modeling cardiac stroke volume measurement based on ultrasound imaging, characterized in that, Includes the following steps: Transthoracic two-dimensional Doppler ultrasound data for each subject were extracted from the hospital database, including left ventricular outflow tract diameter and left ventricular outflow tract velocity time integral, and the corresponding stroke volume was calculated based on these data. ; Fully automated three-dimensional ultrasound data for each subject were extracted from the ultrasound database, and the corresponding stroke volume was recorded as follows: ; analyze and The consistency between the two was assessed, and the clinical consistency between them was judged to be within an acceptable range based on the calculation results. calculate and Bias value of measurement results and analysis and The correlation; Using the diameter of the left ventricular outflow tract and the time integral of the left ventricular outflow tract velocity as independent variables, analyze their effects on... The extent of the impact; The optimal cutoff value for the diameter of the left ventricular outflow tract was calculated, and several subjects were divided into high-risk and low-risk groups based on this value. For the high-risk group, the left ventricular outflow tract diameter, left ventricular outflow tract velocity time integral, Establish a calibration model; The safety of the correction results is evaluated based on the obtained correction model.

2. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 1, characterized in that, The analysis and The correlation is determined in the following ways: Using the Pearson product-moment correlation coefficient formula, the correlation coefficients of all subjects were analyzed. , The Pearson correlation coefficient r is calculated using the following method; where, , r is used to measure and The direction and intensity of the linear correlation between them; The slope of the linear regression equation can be calculated using the following formula. : ; in, yes The sample standard deviation yes The sample standard deviation; The intercept of the linear regression equation can be calculated using the following formula. : ; in, yes The sample mean; yes The sample mean; Based on the calculation results, the corresponding linear regression equation is obtained: 。 3. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 1, characterized in that, The analysis and The consistency between the two, and the method for judging whether the clinical consistency between them is within an acceptable range based on the calculation results, is as follows: according to and The specific values ​​are calculated using Bland-Altman spectroscopy to determine the average bias, 95% compliance limit, compliance range width, and percentage error. If the calculated percentage error is greater than 30%, it is considered... and The consistency between them is poor, and they cannot be directly interchanged in clinical practice.

4. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 2, characterized in that, The analysis uses the left ventricular outflow tract diameter and the time integral of the left ventricular outflow tract velocity as independent variables to examine their impact on... The degree of influence is determined as follows; For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The diameter of the left ventricular outflow tract. The intercept is... For coefficients; For all subjects ( , Perform linear regression analysis to obtain the corresponding linear regression equation. and judge and Whether they are positively correlated, among which, The time integral of the left ventricular outflow tract velocity. The intercept is... For coefficients; Will , At the same time, as an independent variable, and Perform linear regression analysis; if the results show that both are related to... Independent correlation means , All Independent influencing factors.

5. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 4, characterized in that, The process of calculating the optimal cutoff value for the left ventricular outflow tract diameter and using this value to divide several subjects into high-risk and low-risk groups is as follows: Using the left ventricular outflow tract diameter for each subject as a state variable and the clinically significant bias value as a state variable, ROC curves were plotted, and the area under the curve and Youden index were calculated accordingly. Based on the size of the area under the curve, determine the relationship between the diameter of the left ventricular outflow tract and the curve. The ability to distinguish between large and small values; and the left ventricular outflow tract diameter corresponding to the largest Youden index is taken as the optimal cutoff value; Subjects whose left ventricular outflow tract diameter was smaller than the optimal cutoff value were classified as the low-risk group, while the remaining subjects were classified as the high-risk group.

6. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 5, characterized in that, The method for establishing the calibration model is as follows: Based on the obtained linear regression equation, the following two methods are used to... Perform the correction and record the correction result as follows: : ; in, It is the corrected stroke volume based on the diameter of the left ventricular outflow tract. ; in, It is the corrected stroke volume based on the time integral of the left ventricular outflow tract velocity. Based on the above calculation model, calculate the corrected bias values ​​respectively. : ; Based on the calculation results, the average bias value and bias rate of the whole sample and the high-risk group are calculated, and the best correction model is selected accordingly.

7. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 6, characterized in that, The process for assessing the security of the correction results is as follows: The selected optimal correction model is used to calculate the data in the low-risk group that are below the clinically significant bias value. Based on the calculation results, it is determined whether a new significant bias will be added.

8. The method for modeling cardiac stroke volume measurement based on ultrasound imaging according to claim 1, characterized in that, The The calculation method is as follows: according to and The size, for Perform the calculation, if > Then the following calculations are performed: 。