Application of heart rate variability multi-parameter in anterior circulation single subcortical infarction system

Abstract

The application relates to the technical field of ischemic stroke prediction, and provides application of heart rate variability multi-parameters in a system of anterior circulation single subcortical infarction. SS , MSE AUC , CMSE SS , CMSE LS or CMSE AUC ; wherein, MSE SS : small-scale multi-scale entropy index; MSE AUC : multi-scale entropy area under the curve; CMSE SS : small-scale composite multi-scale entropy index; CMSE LS : large-scale composite multi-scale entropy index; and CMSE AUC : composite multi-scale entropy area under the curve. The application proves that the heart rate variability multi-parameters are independently related to whether the anterior circulation single subcortical infarction patient has progressive cerebral infarction through systematic analysis of the correlation between linear and nonlinear HRV parameters and progressive cerebral infarction, and the heart rate variability multi-parameters combined with conventional clinical indexes can improve the prediction efficiency of traditional clinical-image-laboratory indexes, and the heart rate variability multi-parameters include MSE SS , MSE AUC , CMSE SS , CMSE LS or CMSE AUC .

Description

Technical Field

[0001] This invention belongs to the field of ischemic stroke prediction technology, and particularly relates to the application of heart rate variability multiparameters in the anterior circulation single subcortical infarction system. Background Technology

[0002] Single subcortical infarction (SSI) accounts for approximately one-third of all ischemic strokes. It refers to infarction caused by the occlusion of a single perforating artery, involving subcortical structures such as the basal ganglia, internal capsule, and thalamus. Clinically, SSI often presents as limb weakness, sensory disturbances, or dysarthria, with relatively mild neurological deficits. Although the overall prognosis of SSI is relatively good, 20-40% of patients may experience early neurological deterioration (END). Causes include progressive infarction (PI), recurrent cerebral ischemia, increased intracranial pressure, and secondary parenchymal hemorrhage, with PI being the most common cause.

[0003] Intense pulse variability (PI) significantly impacts the functional prognosis of stroke patients with sudden cardiac arrest (SSI), making early identification of its predictors crucial for targeted intervention. Studies have confirmed that initial National Institutes of Health Stroke Scale (NIHSS) scores, ≥3 axial infarct layers, dyslipidemia, and insulin resistance are all closely associated with PI. Heart rate variability (HRV) refers to the temporal variation between consecutive heartbeats and is a non-invasive biomarker reflecting autonomic nervous function. Traditional linear HRV analysis assesses autonomic nervous tension using time-domain and frequency-domain indicators; however, nonlinear parameters such as Higuchi fractal dimension (HFD), Poincaré plots, multiscale entropy (MSE), and composite multiscale entropy (CMSE) are superior to traditional linear methods in quantifying HRV, especially in complex physiological states. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides the application of heart rate variability multi-parameters in anterior circulation single subcortical infarction systems, with the aim of solving the problems mentioned in the background art.

[0005] In a first aspect, this invention provides the application of multiple heart rate variability parameters in constructing a prediction system for single subcortical infarction in the anterior circulation, wherein the multiple heart rate variability parameters include MSE.SS MSE AUC CMSE SS CMSE LS or CMSE AUC ; Among them, MSE SS Small-scale multi-scale entropy exponent; MSE AUC Area under the multiscale entropy curve (CMSE) SS : Composite small-scale multi-scale entropy index; CMSE LS : Composite large-scale multi-scale entropy index; CMSE AUC Area under the composite multiscale entropy curve:

[0006] Furthermore, the prediction system includes an electrocardiogram (ECG) acquisition component.

[0007] Furthermore, the ECG acquisition component acquires multiple parameters of the subject's heart rate variability by using a dynamic electrocardiograph to record the subject's continuous ECG signals for 24 hours.

[0008] Furthermore, the subject was a patient with a single subcortical infarction in the anterior circulation.

[0009] Furthermore, by combining multiple parameters of heart rate variability and routine clinical indicators, we can help predict whether patients with a single subcortical infarction in the anterior circulation will develop progressive cerebral infarction; the routine clinical indicators are the initial NIHSS score, serum creatinine, and the number of axial planes of infarction ≥3.

[0010] Secondly, the present invention provides a computer program for predicting single subcortical infarction in the anterior circulation. When the computer program is executed, it performs the following specific steps: recording the subject's continuous electrocardiogram signal for 24 hours using a dynamic electrocardiograph; identifying the R wave peak after standardized preprocessing; calculating adjacent RR intervals and constructing an RR interval time series; calculating a nonlinear HRV index; and obtaining multiple parameters of heart rate variability, including MSE. SS MSE AUC CMSE SS CMSE LS or CMSE AUC ; The basic model was constructed based on conventional clinical indicators, namely, initial NIHSS score, creatinine, and number of axial infarct layers ≥3. The basic model is respectively related to MSE SS MSE AUC CMSE SS CMSE LS or CMSE AUC Jointly construct a prediction model.

[0011] Furthermore, the 24-hour ECG signal was segmented into 1-hour intervals for MSE analysis, and the MSE was calculated. SS and MSE AUC .

[0012] Furthermore, the CMSE algorithm is used to calculate CMSE. SS CMSE LS or CMSE AUC .

[0013] This invention has the following beneficial effects: Through systematic analysis of the association between linear and nonlinear HRV parameters and progressive cerebral infarction, the results confirm the multiparameter heart rate variability (MSE) SS MSE AUC CMSE SS CMSE LS or CMSE AUC Heart rate variability was independently associated with whether patients with a single subcortical infarction in the anterior circulation would develop progressive cerebral infarction, and the combination of multiple parameters of heart rate variability with conventional clinical indicators could improve the predictive efficacy of traditional clinical-imaging-laboratory indicators. Attached Figure Description

[0014] Exemplary embodiments of the present invention can be more fully understood by referring to the following figures: Figure 1 This is a three-dimensional distribution map of MSE values ​​in Embodiment 1 of the present invention.

[0015] Figure 2 This is the basic model of Embodiment 2 of the present invention and its combined heart rate variability multi-parameter prediction efficacy analysis results. Detailed Implementation

[0016] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present 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 of the present invention and are not intended to limit the present invention.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention.

[0018] Materials and Methods: 1. Study subjects: Patients admitted to the stroke unit of the First Affiliated Hospital of Nanchang University within 48 hours of onset between January 2020 and October 2022 were selected consecutively.

[0019] Inclusion criteria: (1) First diffusion-weighted imaging (DWI) completed within 48 hours of onset; (2) DWI diagnosis of ACSSI consistent with clinical neurological deficits; (3) Second DWI or computed tomography (CT) completed when neurological function deteriorates.

[0020] Exclusion criteria: (1) Comorbid atrial fibrillation (AF), symptomatic heart failure, or acute myocardial infarction; (2) Failure to complete 24-hour Holter monitoring within 3 days of admission or poor ECG signal quality; (3) Received intravenous thrombolysis or endovascular treatment; (3) Suspected cardioembolism, arterial-arterial embolism, or other definite causes (Moyamoya disease, aortic dissection, vasculitis, etc.); (4) Worsening of neurological deficits before the first DWI examination; (5) Incomplete imaging, laboratory, or follow-up data.

[0021] The protocol has been approved by the Ethics Committee of the First Affiliated Hospital of Nanchang University, review opinion number: IIT

[2023] Lin Lun Shen No. 364, and all participants have signed written informed consent forms.

[0022] 2. Data collection: (1) Collect baseline data: age, sex, medical history (hypertension, diabetes, stroke), initial NIHSS score; (2) Laboratory parameters within 24 hours of admission: neutrophil count, lymphocyte count, blood urea nitrogen (BUN), creatinine, uric acid, fasting blood glucose, total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), homocysteine, fibrinogen, and D-dimer. The neutrophil-to-lymphocyte ratio (NLR) was also calculated. All patients underwent standardized imaging assessments according to previous study protocols. The Fazekas 4-point scale was used to assess leukoaraiosis, and the number of axial infarct slices and the maximum diameter on DWI were recorded.

[0023] (3) Definition: Anterior circulation single subcortical infarction (ACSSI): A single subcortical infarction supplied by the lenticulostriate artery, Heubner's recurrent artery, or the anterior choroidal artery.

[0024] Progressive infarction (PI): an increase of ≥1 point in limb motor strength score or ≥2 points in NIHSS total score within 7 days of admission, and a second DWI or CT scan confirms enlargement of the original infarct lesion.

[0025] ECG data acquisition and HRV analysis: The patient's ECG signal was recorded continuously for 24 hours using a dynamic electrocardiograph. After standardized preprocessing, the R wave peak was identified, the adjacent RR intervals were calculated, and the RR interval time series was constructed. Based on this, linear and nonlinear HRV indices were calculated.

[0026] Linear HRV analysis Time-domain analysis: Three core indicators were measured: standard deviation of all normal-to-normal RR intervals (SDNN), standard deviation of the averages of all normal-to-normal RR intervals in all 5-min segments (SDANN), and root mean square of differences of adjacent normal-to-normal RR intervals (RMSSD).

[0027] Frequency domain analysis: The Welch method was used to convert the RR interval time series into power spectral density to quantify autonomic neural modulation. Frequency bands were divided according to international standards: low frequency (LF, 0.04~0.15Hz), high frequency (HF, 0.15~0.4Hz), and total power (TP, ≤0.4Hz). LF, HF, TP, and the LF / HF ratio were calculated.

[0028] Nonlinear HRV Analysis Poincaré plot: Presents continuous RR intervals as two-dimensional scatter points, and extracts the standard deviation of the minor axis (SD1), the standard deviation of the major axis (SD2), and the SD1 / SD2 ratio after fitting an ellipse.

[0029] HFD: By constructing subsequences at different scales and calculating curve lengths, the fractal characteristics of time series are quantified.

[0030] Approximate entropy (ApEn) and sample entropy (SampEn): used to quantify the regularity and complexity of time series. SampEn eliminates self-matching bias and has better stability.

[0031] The 24-hour electrocardiogram (ECG) signal was segmented into 1-hour intervals for MSE analysis, and a three-dimensional distribution map of the 24-hour MSE values ​​was plotted. Figure 1 As shown. Calculate the hourly small-scale (τ=1-5) MSE. SS Large-scale (τ=16-20) large-scale multiscale entropy index (MSE) LS ) and MSE AUC The 24-hour average was used as the final indicator. The CMSE was calculated using the CMSE algorithm. SS CMSE LS CMSE AUC .

[0032] 3. Statistical methods Continuous variables were expressed as mean ± standard deviation or median (interquartile range, IQR), and comparisons between groups were performed using t-tests or Mann-Whitney U tests. Categorical variables were expressed as number of cases (percentages), and comparisons between groups were performed using χ² tests or Fisher's exact test. Multivariate logistic regression analysis was used to analyze the independent association between HRV parameters and PI. ROC curves were used to evaluate the predictive power of each model for PI. Statistical analysis was performed using SPSS 26.0 software, and P < 0.05 was considered statistically significant.

[0033] Example 1: From January 2020 to October 2022, a total of 1,035 patients with acute coronary syndrome (ACSSI) were screened. 175 patients who received intravenous thrombolysis or endovascular treatment were excluded; 15 patients with suspected cardiogenic or arterial-arterial embolism or other definite causes were excluded; 98 patients with deteriorating neurological function before the first DWI were excluded; 109 patients had incomplete data; 16 patients had atrial fibrillation, heart failure, or myocardial infarction; and 305 patients who did not undergo 24-hour Holter monitoring or had poor ECG signal quality were excluded. Finally, 317 patients were included, including 58 patients in the progressive cerebral infarction group and 259 patients in the non-progressive cerebral infarction group.

[0034] Table 1 shows the comparison of baseline demographic, clinical, laboratory, and imaging data between the progressive cerebral infarction group and the non-progressive cerebral infarction group. The results showed that compared with the non-progressive cerebral infarction group, the progressive cerebral infarction group had higher initial NIHSS scores, lower creatinine levels, and a higher proportion of infarct axial slices ≥3, all of which were statistically significant (P<0.05).

[0035] Table 1. Comparison of baseline demographic, clinical, laboratory, and imaging data between the two groups.

[0036] In Table 1, PI: progressive cerebral infarction; NIHSS: National Institutes of Health Stroke Scale; NLR: neutrophil / lymphocyte ratio; BUN: blood urea nitrogen; TC: total cholesterol; TG: triglycerides; HDL-c: high-density lipoprotein cholesterol; LDL-c: low-density lipoprotein cholesterol; IQR: interquartile range; SD: standard deviation. *P<0.05.

[0037] Table 2 shows a comparison of the multi-parameter heart rate variability between the progressive cerebral infarction group and the non-progressive cerebral infarction group. The results indicate that the linear parameter LF and the nonlinear parameter MSE... SS MSE LS MSEA UC CMSE SS CMSE LS CMSE AUC The differences were statistically significant (P<0.05), while the differences in other HRV parameters were not statistically significant (P>0.05).

[0038] Table 2. Comparison of multiple parameters of heart rate variability between the two groups.

[0039] In Table 2, PI represents progressive cerebral infarction; SDNN represents the standard deviation of the RR interval of all normal sinus beats; SDANN represents the standard deviation of the average RR interval for every 5 minutes throughout the entire process; RMSSD represents the root mean square of the difference between adjacent normal RR intervals; LF represents low frequency; HF represents high frequency; HFD represents the Higuchi fractal dimension; ApEn represents approximate entropy; SampEn represents sample entropy; and MSE represents the sample entropy. SS Small-scale multi-scale entropy exponent; MSE LS Composite large-scale multi-scale entropy exponent; MSE AUC Area under the multiscale entropy curve (CMSE) SS : Composite small-scale multi-scale entropy index; CMSE LS : Composite large-scale multi-scale entropy index; CMSE AUCArea under the composite multiscale entropy curve; IQR: interquartile range; SD: standard deviation. *P<0.05, the difference is statistically significant.

[0040] Example 2: Validation of multiparameters of heart rate variability in predicting single subcortical infarction in the anterior circulation 1. Using the nonlinear parameter of heart rate variability as a predictive index, the logistic regression validation of the independent association between multiple parameters of heart rate variability and progressive cerebral infarction (PI) is shown in Table 3. The results show that after adjusting for all confounding factors, the small-scale multiscale entropy index (MSE) is significantly higher than that of progressive cerebral infarction (PI). SS The area under the multiscale entropy curve (MSE) AUC Composite Small-Scale Multi-Scale Entropy Index (CMSE) SS Composite Large-Scale Multiscale Entropy Index (CMSE) LS ) and the area under the composite multiscale entropy curve (CMSE) AUC It was independently associated with progressive cerebral infarction (P<0.05).

[0041] Table 3. Logistic regression validation of the independent association between multiple parameters of heart rate variability and progressive cerebral infarction.

[0042] 2. Further analysis using receiver operating characteristic (ROC) curves was conducted to evaluate the predictive efficacy of heart rate variability (HRV) indicators for progressive cerebral infarction, such as... Figure 2 As shown. In predicting progressive cerebral infarction (PI) in patients with a single subcortical infarction in the anterior circulation, Model 1 (initial NIHSS score at admission, serum creatinine, and ≥3 axial infarct planes) had an area under the curve (AUC) of 0.684 (95% CI: 0.612–0.757), an optimal cutoff value of 0.1160, a sensitivity of 87.93%, and a specificity of 43.24%. Based on Model 1, combining it with MSE... SS The AUC was 0.705 (95% CI: 0.636–0.775), the optimal cutoff was 0.1898, the sensitivity was 75.86%, and the specificity was 62.93%; combined with MSE AUC The AUC was 0.708 (95% CI: 0.640–0.776), the optimal cutoff was 0.2088, the sensitivity was 70.69%, and the specificity was 69.88%; combined with CMSE SS The AUC was 0.708 (95% CI: 0.639–0.778), the optimal cutoff was 0.1914, the sensitivity was 75.86%, and the specificity was 64.09%; combined with CMSE LSThe AUC was 0.703 (95% CI: 0.634–0.772), the optimal cutoff was 0.1736, the sensitivity was 77.59%, and the specificity was 57.53%; combined with CMSE AUC The AUC was 0.716 (95% CI: 0.648–0.784), the optimal cutoff was 0.2203, the sensitivity was 68.97%, and the specificity was 73.36%.

[0043] 3. The specific steps for constructing a computer program based on a predictive model for predicting single subcortical infarction in the anterior circulation are as follows: When a computer program for predicting single subcortical infarction in the anterior circulation is executed, it performs the following specific steps: Recording the subject's continuous ECG signals for 24 hours using a Holter monitor; identifying R-wave peaks after standardized preprocessing; calculating adjacent RR intervals and constructing an RR interval time series; calculating the nonlinear HRV index; and obtaining multiple parameters of heart rate variability, including MSE. SS MSE AUC CMSE SS CMSE LS or CMSE AUC ; The basic model was constructed based on conventional clinical indicators, namely, initial NIHSS score, serum creatinine, and ≥3 axial infarct planes. The basic model is respectively related to MSE SS MSE AUC CMSE SS CMSE LS or CMSE AUC Jointly construct a prediction model.

[0044] All prediction models used binary logistic regression models, modeled using SPSS 26.0 software, to predict the risk probability of progressive stroke (PI) in patients with a single subcortical infarction in the anterior circulation.

[0045] Variable definition and assignment rules: (1) Dependent variable Y: Has progressive cerebral infarction (PI) occurred? Y=1: The patient has developed progressive cerebral infarction; Y=0: The patient has not experienced progressive cerebral infarction; P: Predicted probability of the patient developing progressive cerebral infarction, with a value range of 0. <P<1。

[0046] (2) The independent variables are shown in Table 4.

[0047] Table 4 Independent Variables

[0048] (3) General formula ①Logit linear transformation formula ; In the formula: Logit(P) is the Logit logarithmic transformation value of the probability of onset. In this embodiment, P represents the predicted probability of progressive cerebral infarction in patients with single subcortical infarction in the anterior circulation, and the value range is 0 < P < 1. Since the probability cannot be directly established as a linear equation, the probability in the 0~1 interval is mapped to the whole real number interval through the Logit logarithmic transformation. For the constant term of the regression equation, The regression coefficients corresponding to the first predictive indicator are: As the first predictive indicator, The regression coefficients corresponding to the second predictive indicator are... The second predictive indicator, The regression coefficients corresponding to the third predictive indicator are... The third predictive indicator, The regression coefficient is the value corresponding to the nth predictive indicator. This is the nth predictive indicator.

[0049] ② Formula for calculating the probability of an event occurring .

[0050] Complete formulas, experimental data, and judgment criteria for each prediction model: Model 1, the basic model, is constructed based on conventional clinical indicators (initial NIHSS score upon admission, serum creatinine, and ≥3 axial infarct planes); Model formula: , Results: AUC=0.684; Optimal cutoff value=0.1160; Risk assessment rule: when P≥0.1160, it is considered high risk of progressive cerebral infarction; when P<0.1160, it is considered low risk of progressive cerebral infarction.

[0051] Model 2, i.e., Model Construction; Model Formula: , Results: AUC = 0.705; Optimal cutoff value = 0.1898; Risk assessment rule: When P ≥ 0.1898, it is considered high risk of progressive cerebral infarction; when P < 0.1898, it is considered low risk of progressive cerebral infarction.

[0052] Model 3, i.e., the basic model Construction; Model Formula: , Results: AUC=0.708; Optimal cutoff value=0.2088; Risk assessment rule: when P≥0.2088, it is considered high risk of progressive cerebral infarction, and when P<0.2088, it is considered low risk of progressive cerebral infarction.

[0053] Model 4, i.e., Model Construction; Model Formula: , Results: AUC=0.708; Optimal cutoff value=0.1914; Risk assessment rule: when P≥0.1914, it is considered high risk of progressive cerebral infarction; when P<0.1914, it is considered low risk of progressive cerebral infarction.

[0054] Model 5, i.e., Model Construction; Model Formula: , Results: AUC=0.703; Optimal cutoff value=0.1736; Risk assessment rule: when P≥0.1736, it is considered high risk of progressive cerebral infarction, and when P<0.1736, it is considered low risk of progressive cerebral infarction.

[0055] Model 6, i.e., Model Construction; Model Formula: , Results: AUC=0.716; Optimal cutoff value=0.2203; Risk assessment rule: when P≥0.2203, it is considered high risk of progressive cerebral infarction, and when P<0.2203, it is considered low risk of progressive cerebral infarction.

[0056] Heart Rate Variability (HRV) Calculation Rules:

[0057] (1) RR interval series calculation rules

[0058] Assuming the dynamic electrocardiogram (ECG) captures continuous R-wave moments: Formula for calculating the period between adjacent cardiac cycles: The final RR interval time series is obtained as follows: .

[0059] (2) Calculation rules for multiscale entropy (MSE) series of indices

[0060] The 24-hour ECG signal was segmented into hourly intervals, and multi-scale entropy analysis was performed independently on each RR interval sequence. Scale division: small scale τ=1~5, and the average entropy value within the interval was calculated to obtain the MSE. SS For large-scale operations with τ = 16~20, the average entropy value within the interval is calculated to obtain the MSE. LS The area under the curve (MSE) is calculated by plotting the scale τ on the x-axis and the multi-scale entropy value on the y-axis.AUC Take the arithmetic mean of all 1-hour segment indicators for the whole day as the final 24-hour overall MSE parameter.

[0061] (3) Calculation rules for Composite Multiscale Entropy (CMSE) series of indices

[0062] Based on the RR interval time series, the composite multi-scale entropy algorithm is used to calculate the average entropy value at scales τ=1~5, thus obtaining the CMSE. SS The CMSE is obtained by averaging the entropy values ​​at scales τ = 16~20. LS Using scale τ as the abscissa and the composite multiscale entropy value as the ordinate, the area under the curve is calculated to obtain the CMSE. AUC .

[0063] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. The application of heart rate variability multiparameters in constructing a prediction system for single subcortical infarction in the anterior circulation, characterized by: Heart rate variability multiparameters including MSE SS MSE AUC CMSE SS CMSE LS or CMSE AUC ; Among them, MSE SS Small-scale multi-scale entropy exponent; MSE AUC Area under the multiscale entropy curve (CMSE) SS : Composite small-scale multi-scale entropy index; CMSE LS : Composite large-scale multi-scale entropy index; CMSE AUC Area under the composite multiscale entropy curve: Area index.

2. The application as described in claim 1, characterized in that: The prediction system includes an electrocardiogram (ECG) acquisition component.

3. The application as described in claim 2, characterized in that: The ECG acquisition component acquires multiple parameters of the subject's heart rate variability by using a dynamic electrocardiograph to record the subject's continuous ECG signals for 24 hours.

4. The application as described in claim 3, characterized in that: The subjects were patients with a single subcortical infarction in the anterior circulation.

5. The application as described in claim 4, characterized in that: By combining multiple parameters of heart rate variability and routine clinical indicators, this study helps predict whether patients with a single subcortical infarction in the anterior circulation will develop progressive cerebral infarction. The routine clinical indicators are the initial NIHSS score, serum creatinine, and the number of axial planes of infarction ≥3.

6. A computer program for predicting single subcortical infarction in the anterior circulation, characterized in that: When the computer program is executed, it performs the following specific steps: Recording the subject's continuous ECG signals for 24 hours using a dynamic electrocardiograph; identifying the R-wave peak after standardized preprocessing; calculating adjacent RR intervals and constructing an RR interval time series; calculating the nonlinear HRV index; and obtaining multiple parameters of heart rate variability, including MSE. SS MSE AUC CMSE SS CMSE LS or CMSE AUC ; The basic model was constructed based on conventional clinical indicators, namely, initial NIHSS score, serum creatinine, and ≥3 axial infarct planes. The basic model is respectively related to MSE SS MSE AUC CMSE SS CMSE LS or CMSE AUC Jointly construct a prediction model.

7. A computer program for predicting single subcortical infarction in the anterior circulation as described in claim 6, characterized in that: The 24-hour electrocardiogram (ECG) signal was divided into 1-hour segments for MSE analysis, and the MSE was calculated. SS and MSE AUC .

8. A computer program for predicting single subcortical infarction in the anterior circulation as described in claim 7, characterized in that: CMSE is calculated using the CMSE algorithm. SS CMSE LS or CMSE AUC .

Images