A rapid blood lipid detection method based on spectral analysis

Abstract

The present application relates to the technical field of blood lipid detection, and particularly relates to a blood lipid rapid detection method based on spectrum analysis. The method comprises the following steps: S1, judging the history of blood lipid abnormality; S2, investigating the multi-source interference factors; S3, analyzing the detection indexes; and S4, adjusting the preset examination period. The present application obtains the preset blood lipid diagnosis standard, historical detection data, demographic and physical examination information and spectrum detection indexes, and realizes the multi-element judgment on the existing blood lipid abnormality and the risk of blood lipid increase by fitting the blood lipid prediction curve and the health curve in real time, combining the multi-dimensional interference factor and the comorbidity correlation degree evaluation, and evaluating and selecting the corresponding detection mode according to the peak area and the peak displacement. The present application can effectively distinguish the primary and secondary hyperlipidemia, and dynamically adjust the detection period by selecting the infrared or metal Raman spectrum detection mode, so as to improve the detection efficiency and the early warning accuracy, and provide multiple choices for the clinical individual intervention scheme.

Description

Technical Field

[0001] This invention relates to the field of blood lipid detection technology, and in particular to a rapid blood lipid detection method based on spectral analysis. Background Technology

[0002] Dyslipidemia, also known as hyperlipidemia, is caused by lipid metabolism disorders, leading to abnormally high levels of lipids in the serum. It is closely related to diabetes, hypertension, and cardiovascular disease. Currently, the prevalence of hyperlipidemia among Chinese adults is 40.4%. Improving early detection rates and intervening through lifestyle and dietary changes, as well as medication, can effectively reduce the risk of cardiovascular disease. Therefore, combining digital healthcare technology with patient feedback data to match appropriate testing methods and optimizing the analysis of raw spectral data to achieve accurate and efficient blood lipid testing has become a research hotspot.

[0003] Raman spectroscopy, based on molecular vibrational scattering, can reflect changes in the structure and content of macromolecules such as proteins, lipids, nucleic acids, and sugars, making it suitable for non-invasive, real-time detection. It offers significant advantages over traditional biochemical assays due to its fast detection speed and simple sample processing. Quantitative analysis of peak range, intensity, and width using multivariate computational methods enables qualitative and quantitative detection of samples, providing a technical basis for clinical disease screening.

[0004] Chinese patent publication CN111329492A discloses a non-invasive blood lipid detection device and method based on near-infrared spectroscopy. This invention uses a system composed of supercontinuum laser, broadband light source, fiber optic beam splitter, and focusing lens to achieve efficient acquisition of near-infrared spectral signals from the human body detection site. It also utilizes a lock-in amplifier to synchronously acquire signals with a heart rate meter, improving signal stability and noise suppression, and ensuring data reliability. However, this invention relies too heavily on hardware construction and lacks in-depth data processing. It has poor ability to assess patients' historical blood lipid changes and individual differences, and its selection of qualitative and quantitative judgment factors for blood lipid abnormalities is relatively limited. Summary of the Invention

[0005] Therefore, this invention provides a rapid blood lipid detection method based on spectral analysis to overcome the problems of low data processing efficiency caused by the single data processing method and lack of individualized risk prediction and assessment capabilities in the prior art.

[0006] To achieve the above objectives, this invention provides a rapid blood lipid detection method based on spectral analysis, comprising:

[0007] Historical blood lipid data and basic indicator information are obtained using a preset examination cycle;

[0008] The historical blood lipid data is classified according to the preset blood lipid diagnostic criteria to obtain the results of the first indicator category or the second indicator category, and the corresponding detection method is selected based on the classification results.

[0009] When the historical blood lipid data is determined to be in the second indicator category, a multi-source interference factor screening is performed on the historical blood lipid data.

[0010] Based on the basic indicator information and the historical blood lipid data, it is determined whether the blood lipid level changes deviate from the normal state, and when the patient's blood lipid level changes deviate from the normal state, a multidimensional interference factor assessment is performed on the influencing factors of non-spontaneous blood lipid indicators to obtain the real-time comorbidity degree.

[0011] Based on the comparison results between the real-time comorbidity correlation degree and the preset standard comorbidity correlation degree range, the blood lipid detection method is selected, including the first predictive detection method and the second predictive detection method;

[0012] Among them, when the real-time co-morbidity correlation is greater than or equal to the minimum value of the standard interval, the first prediction detection method is used to perform feature analysis on the target detection molecule;

[0013] When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is adopted, and the degree of dyslipidemia risk is determined by analyzing the peak area and peak position deviation of the target characteristic peak, and it is determined whether to adjust the preset examination cycle.

[0014] The first predictive detection method is infrared spectroscopy detection, and the second predictive detection method is metal Raman spectroscopy detection.

[0015] Furthermore, historical blood lipid data are categorized according to preset blood lipid diagnostic criteria, and appropriate testing methods are selected based on the categorization results, including...

[0016] The blood lipid index category is determined according to the preset blood lipid diagnostic criteria, including the first index category and the second index category, and the corresponding test method is selected based on the classification result;

[0017] When the classification result is the first indicator category, select the first detection method.

[0018] When the classification result is the second indicator category, perform a multi-source interference factor screening step on the historical blood lipid data;

[0019] Furthermore, the steps for screening historical blood lipid data for multiple sources of confounding factors include:

[0020] Obtain basic indicator information, and based on this information and historical blood lipid data, obtain blood lipid prediction data curves and healthy blood lipid data curves. Based on the deviation value and the trend of deviation value changes, determine whether the patient's blood lipid level changes deviate from the normal state.

[0021] Obtain the deviation value change curve, fit a linear equation to the deviation value change curve, obtain the absolute value of the slope difference of the linear equation, and compare it with the standard slope value.

[0022] When the absolute value of the slope difference is greater than the standard slope value, it is determined that the trend of blood lipid level changes deviates from the normal state, and non-spontaneous blood lipid index influencing factors are classified and eliminated.

[0023] When the absolute value of the slope difference is less than or equal to the standard slope value, the trend of blood lipid level change is determined to be normal, and the current preset examination cycle is determined to be the re-examination cycle.

[0024] The basic indicator information includes demographic indicators and general physical examination indicators; any deviation value is the difference between the data point on the blood lipid prediction data curve and the data point on the healthy blood lipid data curve corresponding to the same time series.

[0025] Furthermore, the factors influencing non-spontaneous lipid indicators were classified and eliminated, including an assessment of the impact of multidimensional confounding factors.

[0026] A multidimensional interference factor impact assessment was performed based on basic indicator information, resulting in a multidimensional interference factor impact assessment score, which was then compared with the standard score of the multidimensional interference factor impact.

[0027] When the multidimensional interference factor influence assessment score is greater than or equal to the multidimensional interference factor influence standard score, the influence of all relevant interference factors is reduced to the normal range before blood lipid testing is performed.

[0028] When the multidimensional interference factor impact assessment score is lower than the multidimensional interference factor impact standard score, the impact of comorbidity parameters is assessed.

[0029] Furthermore, the classification and elimination of factors influencing non-spontaneous lipid indicators also includes an assessment of the impact of comorbidity, among which...

[0030] Several comorbidity parameters were obtained based on statistical analysis of past medical history data and clinical data in the database.

[0031] If no prior medical history is determined, the second predictive testing method will be implemented;

[0032] If a pre-existing medical history is determined, the impact of comorbidity on comorbidity parameters is further assessed to obtain the real-time comorbidity.

[0033] The real-time comorbidity correlation is compared with the standard comorbidity correlation interval. When the real-time comorbidity correlation is greater than or equal to the minimum value of the standard comorbidity correlation interval, the first predictive detection method is executed, and the corresponding detection content is adjusted according to the abnormal comorbidity parameters.

[0034] When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is executed.

[0035] Furthermore, the methods for performing predictive detection include,

[0036] The predictive detection methods include a first predictive detection method and a second predictive detection method;

[0037] The first predictive detection method is infrared spectroscopy for blood lipid detection, and the infrared spectroscopy detection content is adjusted based on the comorbidity parameters that have the greatest impact on the real-time comorbidity correlation. The second predictive detection method is metal Raman spectroscopy, and its detection process involves acquiring several detection indicators and corresponding detection analysis results, and determining whether to adjust the preset examination cycle based on the analysis results.

[0038] Furthermore, several detection indicators were obtained, including

[0039] The detection indicators include peak area deviation and peak displacement deviation;

[0040] Obtain the raw Raman spectral data and preprocess it to obtain the target Raman spectral data;

[0041] Based on the required detection content, the corresponding features of the target Raman spectral data are extracted to obtain several target feature peaks;

[0042] Perform peak area deviation analysis and peak displacement deviation analysis on any of the target characteristic peaks, and determine whether to adjust the preset inspection cycle based on the analysis results.

[0043] Furthermore, the preprocessing yields the target Raman spectral data, including:

[0044] Random noise in the original Raman spectrum was removed using a smoothing algorithm, and baseline drift was removed by polynomial fitting to obtain intermediate Raman spectral data;

[0045] The intermediate Raman spectral data are normalized to obtain the target Raman spectral data.

[0046] Furthermore, performing peak area deviation analysis and peak shift deviation analysis on any of the target characteristic peaks includes,

[0047] Based on the peak area prediction model, the standard peak area value of any of the target characteristic peaks under the basic index information is obtained;

[0048] Obtain the real-time peak area value and the real-time peak area difference of the target characteristic peak;

[0049] Among them, the real-time peak area difference is the absolute value of the difference between the real-time peak area value and the standard peak area value;

[0050] The real-time peak area difference is compared with the standard peak area difference.

[0051] If the real-time peak area difference is greater than the standard peak area difference, a risk result of dyslipidemia is obtained.

[0052] If the real-time peak area difference is less than or equal to the standard peak area difference, the standard peak displacement value corresponding to the target characteristic peak is obtained based on the peak displacement prediction model, and the real-time peak displacement value and real-time displacement offset of the target characteristic peak are obtained.

[0053] Among them, the real-time displacement offset is the absolute value of the difference between the real-time peak displacement value and the standard peak displacement value;

[0054] The real-time displacement offset is compared with the standard displacement offset.

[0055] If the real-time displacement offset is greater than the standard displacement offset, a Class II dyslipidemia risk result is obtained;

[0056] If the real-time displacement is less than or equal to the standard displacement, the risk level of dyslipidemia is considered normal.

[0057] Furthermore, determining whether to adjust the preset inspection cycle based on the analysis results includes:

[0058] When blood lipid levels are abnormal, adjust the preset examination cycle to the corrected examination cycle;

[0059] Among them, abnormal blood lipid indicators include Class I blood lipid abnormality risk results and Class II blood lipid abnormality risk results; when the abnormal result is a Class I blood lipid abnormality result, the corrected examination cycle is the product of the preset examination cycle and the first adjustment step size, the first adjustment step size is the difference between 1 and the first adjustment degree, and the first adjustment degree is the ratio of the difference between the real-time peak area value and the standard peak area value to the real-time peak area value.

[0060] When the abnormal result is a type II dyslipidemia result, the corrected examination cycle is the product of the preset examination cycle and the second adjustment step size. The second adjustment step size is the difference between 1 and the second adjustment degree. The second adjustment degree is the ratio of the difference between the real-time displacement offset and the standard displacement offset to the real-time displacement offset.

[0061] Compared with existing technologies, the beneficial effects of this invention lie in its ability to dynamically determine the risk of dyslipidemia by classifying and fitting predictive curves based on patients' historical blood lipid data and basic demographic and physical examination information, combining assessments of multi-source interference factors and comorbidity parameters, and selecting infrared or metal Raman spectroscopy detection methods according to the degree of comorbidity correlation. Peak area and peak position deviation analyses are performed on the target molecules, and the detection cycle is adjusted accordingly. This method not only improves the early detection rate and accuracy of dyslipidemia but also accurately distinguishes between secondary and primary hyperlipidemia, facilitating the development of personalized intervention plans, reducing the risk of cardiovascular events, and improving clinical testing efficiency. Attached Figure Description

[0062] Figure 1 This is a flowchart of a rapid blood lipid detection method based on spectral analysis, as described in an embodiment of the present invention.

[0063] Figure 2 This is a logic diagram for investigating multi-source interference factors in an embodiment of the present invention;

[0064] Figure 3 A logic decision diagram for obtaining several detection indicators in an embodiment of the present invention;

[0065] Figure 4 This is a logic diagram for determining whether to adjust the preset inspection cycle in an embodiment of the present invention. Detailed Implementation

[0066] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0067] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0068] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0069] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0070] Please see Figure 1 The flowchart shown is a rapid blood lipid detection method based on spectral analysis according to an embodiment of the present invention.

[0071] Historical blood lipid data and basic indicator information are obtained using a preset examination cycle;

[0072] The historical blood lipid data is classified according to the preset blood lipid diagnostic criteria to obtain the results of the first indicator category or the second indicator category, and the corresponding detection method is selected based on the classification results.

[0073] When the historical blood lipid data is determined to be in the second indicator category, a multi-source interference factor screening is performed on the historical blood lipid data.

[0074] Based on the basic indicator information and the historical blood lipid data, it is determined whether the blood lipid level changes deviate from the normal state, and when the patient's blood lipid level changes deviate from the normal state, a multidimensional interference factor assessment is performed on the influencing factors of non-spontaneous blood lipid indicators to obtain the real-time comorbidity degree.

[0075] Based on the comparison results between the real-time comorbidity correlation degree and the preset standard comorbidity correlation degree range, the blood lipid detection method is selected, including the first predictive detection method and the second predictive detection method;

[0076] Among them, when the real-time co-morbidity correlation is greater than or equal to the minimum value of the standard interval, the first prediction detection method is used to perform feature analysis on the target detection molecule;

[0077] When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is adopted, and the degree of dyslipidemia risk is determined by analyzing the peak area and peak position deviation of the target characteristic peak, and it is determined whether to adjust the preset examination cycle.

[0078] The first predictive detection method is infrared spectroscopy detection, and the second predictive detection method is metal Raman spectroscopy detection;

[0079] By classifying and fitting predictive curves based on patients' historical lipid data and basic demographic and physical examination information, and combining assessments of multi-source interference factors and comorbidity parameters, and selecting infrared or metal Raman spectroscopy detection methods according to the degree of comorbidity correlation, peak area and peak position deviation analyses are performed on the target molecules, and the detection cycle is adjusted to achieve dynamic assessment of dyslipidemia risk. This method not only improves the early detection rate and accuracy of dyslipidemia, but also accurately distinguishes between secondary and primary hyperlipidemia, which helps to develop personalized intervention plans, reduce the risk of cardiovascular events, and improve clinical testing efficiency.

[0080] Specifically, historical blood lipid data is categorized according to preset blood lipid diagnostic criteria, and appropriate testing methods are selected based on the categorization results.

[0081] The blood lipid index category is determined according to the preset blood lipid diagnostic criteria, including the first index category and the second index category, and the corresponding test method is selected based on the classification result;

[0082] When the classification result is the first indicator category, select the first detection method;

[0083] When the classification result is the second indicator category, perform a multi-source interference factor screening step on the historical blood lipid data;

[0084] In this embodiment, the hospital set the blood lipid diagnostic criteria as TC < 5.2, TG < 1.7, LDL < 3.4, and HDL > 1.0 / 1.3 according to clinical guidelines. All the above data are in mmol / L.

[0085] Obtain the patient's historical blood lipid data for the past twelve months and compare it with blood lipid diagnostic criteria;

[0086] If any of the above blood lipid data remains abnormal for more than three months, the patient has a lipid-related disease and the first testing method, namely venous blood sampling, will be used.

[0087] If all data are within the normal range, perform the multi-source interference factor investigation step to eliminate interference from external factors for subsequent blood lipid testing;

[0088] This step combines digital healthcare technology to track and analyze the collected patient medical data, intelligently assisting in the prevention of disease risks and improving the efficiency of patient data management. Specifically, preliminary screening through the history of lipid disorders can effectively exclude fluctuations in blood lipids caused by individual differences and exogenous factors such as environment and lifestyle, thereby ensuring the stability of data when subsequent multi-source interference factor screening and spectral detection methods are used. This method helps to identify abnormal blood lipid levels at an early stage, facilitates the development of individualized intervention strategies, and reduces the overall risk of cardiovascular disease and other metabolic syndromes, thus having important value for pre-clinical judgment.

[0089] See Figure 2 As shown, it is a logic decision diagram for investigating multi-source interference factors in an embodiment of the present invention;

[0090] Specifically, the steps for screening historical blood lipid data for multiple sources of confounding factors include:

[0091] Obtain basic indicator information, and based on this information and historical blood lipid data, obtain blood lipid prediction data curves and healthy blood lipid data curves. Based on the deviation value and the trend of deviation value changes, determine whether the patient's blood lipid level changes deviate from the normal state.

[0092] Obtain the deviation value change curve, fit a linear equation to the deviation value change curve, obtain the absolute value of the slope difference of the linear equation, and compare it with the standard slope value.

[0093] When the absolute value of the slope difference is greater than the standard slope value, it is determined that the trend of blood lipid level changes deviates from the normal state, and non-spontaneous blood lipid index influencing factors are classified and eliminated.

[0094] When the absolute value of the slope difference is less than or equal to the standard slope value, the trend of blood lipid level change is determined to be normal, and the current preset examination cycle is determined to be the re-examination cycle.

[0095] The basic indicator information includes demographic indicators and general physical examination indicators; any deviation value is the difference between the data point on the blood lipid prediction data curve and the data point on the healthy blood lipid data curve corresponding to the same time series.

[0096] In this embodiment, demographic indicators include age, sex, race, residential area, occupation, dietary habits, drinking habits, and exercise habits; general physical examination indicators include height, weight, body temperature, blood pressure, pulse, and BMI.

[0097] Matching healthy population samples were selected from the hospital database, and monthly low-density lipoprotein cholesterol data of the healthy population over the past year were extracted to fit a healthy blood lipid curve.

[0098] Wherein, the time series t = [1, 2, ... 12] months;

[0099] The linear regression equation for healthy low-density lipoprotein cholesterol is:

[0100] y 健康 =0.02t + 2.85

[0101] The patient's historical LDL cholesterol levels from 1 to 6 months were [3.5, 3.7, 4.0, 4.2, 4.5, 4.8];

[0102] The linear regression equation for predicting low-density lipoprotein cholesterol is as follows:

[0103] y预测 =0.26t + 3.3

[0104] The linear regression equation for the deviation value change curve is:

[0105] d i =β 趋势 t+β0=0.028t+0.452

[0106] According to the statistical data of the healthy group, the standard slope value is 0.02, the absolute value of the slope difference = 0.028 > 0.02, the trend of blood lipid level changes deviates from the normal state, and subsequent classification and elimination steps will be performed;

[0107] Through the above steps, this embodiment realizes a method for constructing a blood lipid prediction curve based on historical blood lipid data and basic indicator information, and comparing the slope with a health reference curve. This effectively determines whether the patient's blood lipid level changes deviate from the normal state, ensuring the accuracy and reliability of subsequent test data, thereby providing scientific support for individualized blood lipid intervention and cardiovascular disease prevention, and realizing multi-source interference screening and dynamic risk management of blood lipid abnormalities.

[0108] Specifically, the classification and elimination of factors influencing non-spontaneous lipid indicators includes an assessment of the impact of multidimensional confounding factors, among which...

[0109] A multidimensional interference factor impact assessment was performed based on basic indicator information, resulting in a multidimensional interference factor impact assessment score, which was then compared with the standard score of the multidimensional interference factor impact.

[0110] When the multidimensional interference factor influence assessment score is greater than or equal to the multidimensional interference factor influence standard score, the influence of all relevant interference factors is reduced to the normal range before blood lipid testing is performed.

[0111] When the multidimensional interference factor impact assessment score is less than the multidimensional interference factor impact standard score, the impact of comorbidity parameters is assessed.

[0112] In this embodiment, the steps for evaluating the impact of multidimensional interference factors are as follows:

[0113] The basic indicator information obtained includes the following three variables:

[0114] X1: Alcohol intake;

[0115] X2: High-fat diet intake;

[0116] X3: Duration of strenuous exercise;

[0117] To eliminate the influence of dimensions, all the above variables are standardized so that they meet a distribution with a mean of 0 and a standard deviation of 1. The standardization formula is:

[0118]

[0119] The multidimensional interference factor impact assessment model takes the following form:

[0120] Y = β1X1 + β2X2 + β3X3

[0121] Among them, X i For variables; μ i For variable X i The mean; σ i For variable X i Standard deviation; The standardized regression coefficients of each indicator obtained using standardized independent variables reflect the relative contribution of each confounding factor to the change in blood lipid levels.

[0122] when At this time, this factor leads to elevated blood lipids;

[0123] when When the value is less than or equal to 0, this factor inhibits the rise of blood lipids;

[0124] In this embodiment, the reference ranges for these three indicators for healthy individuals are as follows: X1 Alcohol intake ≤25g / day (male), X2 High-fat diet intake ≤30g / day, X3 Strenuous exercise time ≥3 hours / week;

[0125] The statistical parameters used for standardization of the healthy population were as follows: X1 mean alcohol intake of 15g / day, standard deviation of 8g / day; X2 mean high-fat diet intake of 25g / day, standard deviation of 10g / day; X3 mean duration of vigorous exercise of 4 hours / week, standard deviation of 2 hours / week.

[0126] This step effectively determines whether the patient's blood lipid levels deviate from the normal range, and screens and eliminates non-spontaneous influencing factors based on the results, ensuring the accuracy and reliability of subsequent test data, thereby providing scientific support for individualized blood lipid intervention and cardiovascular disease prevention.

[0127] Specifically, the classification and elimination of factors influencing non-spontaneous lipid indicators also includes an assessment of the impact of comorbidity, among which...

[0128] Several comorbidity parameters were obtained based on statistical analysis of past medical history data and clinical data in the database.

[0129] If no prior medical history is determined, the second predictive testing method will be implemented;

[0130] If a pre-existing medical history is determined, the impact of comorbidity on comorbidity parameters is further assessed to obtain the real-time comorbidity.

[0131] The real-time comorbidity correlation is compared with the standard comorbidity correlation interval. When the real-time comorbidity correlation is greater than or equal to the minimum value of the standard comorbidity correlation interval, the first predictive detection method is executed, and the corresponding detection content is adjusted according to the abnormal comorbidity parameters.

[0132] When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is executed.

[0133] In this embodiment, the formula for assessing the impact of comorbidity is:

[0134]

[0135] in,

[0136] D: Comorbidity assessment score;

[0137] w i The weight of the i-th factor is set according to historical data, with w1 = w2 = w3 = 1 / 3; i = 1...n, n = 3;

[0138] The standardized regression coefficients of each indicator obtained using the standardized independent variables are used to quantify the relative contribution of the i-th confounding factor to the change in blood lipid levels.

[0139] In this embodiment, the following intervals were obtained after large-sample verification using historical data:

[0140] When D < 0.3, it is a low correlation region, indicating that non-spontaneous interfering factors have little impact on blood lipid levels, and the patient's blood lipid abnormality is more likely to be primary lipolipemia.

[0141] When 0.3 ≤ D ≤ 0.7, the region is considered moderately correlated, indicating that the confounding factors have a moderate impact on blood lipid levels. In this case, further evaluation of comorbidities and other biochemical indicators is needed to differentiate between secondary components and primary factors.

[0142] When 0.7 is less than D, it is a high correlation region, indicating that non-spontaneous factors such as inflammation, endocrine disorders, pregnancy and other comorbid conditions have a significant impact on blood lipid levels. It is recommended to identify it as secondary lipemia and conduct blood lipid intervention testing after the interfering factors are eliminated or the condition is stable.

[0143] This step obtains the corresponding detection indicators, assigns different weights to each comorbid disease, and calculates a comprehensive score. In the context of blood lipid testing, this comorbidity indicator can be used to assess the extent to which other diseases, such as endocrine disorders, chronic inflammation, and liver diseases, affect a patient's blood lipid levels. This scoring system can not only quantitatively reflect the degree of interference of comorbidities on blood lipid testing, but also serve as an important basis for distinguishing between pathogenic and secondary hyperlipidemia.

[0144] Specifically, the methods for performing predictive detection include:

[0145] The predictive detection methods include a first predictive detection method and a second predictive detection method;

[0146] The first predictive detection method is infrared spectroscopy for blood lipid detection, and the infrared spectroscopy detection content is adjusted based on the comorbidity parameters that have the greatest impact on the real-time comorbidity correlation.

[0147] The second predictive detection method is metal Raman spectroscopy detection. Its detection process involves acquiring several detection indicators and corresponding detection analysis results, and determining whether to adjust the preset inspection cycle based on the analysis results.

[0148] In this embodiment, for familial hypercholesterolemia, infrared detection is used to detect the vibrational peaks of target receptor-related proteins and the corresponding nucleic acid vibrational bands.

[0149] For familial mixed-type dyslipidemia, infrared detection of the protein characteristic peaks of apolipoproteins is used;

[0150] This step selectively uses infrared spectroscopy to detect relevant molecules for different types of pathogenic lipid disorders, such as familial hypercholesterolemia and familial mixed dyslipidemia. Based on different disease characteristics, it accurately detects molecular markers related to pathological states, thereby ensuring that lipid testing is not just limited to the statistical level of lipid data. It provides a more accurate basis for diagnosis and classification, effectively improving diagnostic accuracy and helping medical experts provide personalized and precise treatment recommendations for high-risk patients.

[0151] See Figure 3 As shown, it is a logic decision diagram for obtaining several detection indicators in an embodiment of the present invention;

[0152] Specifically, obtaining several detection indicators includes,

[0153] The detection indicators include peak area deviation and peak displacement deviation;

[0154] Obtain the raw Raman spectral data and preprocess it to obtain the target Raman spectral data;

[0155] Based on the required detection content, the corresponding features of the target Raman spectral data are extracted to obtain several target feature peaks;

[0156] Perform peak area deviation analysis and peak displacement deviation analysis on any of the target characteristic peaks, and determine whether to adjust the preset inspection cycle based on the analysis results.

[0157] Specifically, the preprocessing to obtain the target Raman spectrum data includes,

[0158] Random noise in the original Raman spectrum was removed using a smoothing algorithm, and baseline drift was removed by polynomial fitting to obtain intermediate Raman spectral data;

[0159] The intermediate Raman spectral data are normalized to obtain the target Raman spectral data;

[0160] In this embodiment, the collected raw Raman spectral data is preprocessed by using the Savitzky-Golay smoothing filter algorithm, selecting an appropriate sliding window size and polynomial order to smooth the spectral data.

[0161] The low-signal region without significant peaks in the spectrum is determined by automatic or semi-automatic algorithms as the baseline reference interval;

[0162] Polynomial fitting is performed on the above baseline reference interval data. Usually, a quadratic or cubic polynomial is used to fit a smooth fitting curve, which serves as the background baseline.

[0163] The baseline curve obtained by fitting is subtracted point by point from the smoothed spectral data to obtain the corrected intermediate Raman spectral data.

[0164] Divide the intensity of each point in the intermediate spectral data by the maximum intensity value of the entire spectrum to normalize its value to between 0 and 1.

[0165] Through the above smoothing, baseline correction and normalization processes, this embodiment can ensure that random noise and baseline drift in the target Raman spectral data are effectively removed, so that each spectral data can accurately reflect the molecular vibration information in the sample, thereby providing accurate numerical references for subsequent peak area and peak position deviation analysis, as well as quantitative assessment of dyslipidemia risk.

[0166] Specifically, performing peak area deviation analysis and peak shift deviation analysis on any of the target characteristic peaks includes,

[0167] Based on the peak area prediction model, the standard peak area value of any of the target characteristic peaks under the basic index information is obtained;

[0168] Obtain the real-time peak area value and the real-time peak area difference of the target characteristic peak;

[0169] Among them, the real-time peak area difference is the absolute value of the difference between the real-time peak area value and the standard peak area value;

[0170] The real-time peak area difference is compared with the standard peak area difference.

[0171] If the real-time peak area difference is greater than the standard peak area difference, a risk result of dyslipidemia is obtained.

[0172] If the real-time peak area difference is less than or equal to the standard peak area difference, the standard peak displacement value corresponding to the target characteristic peak is obtained based on the peak displacement prediction model, and the real-time peak displacement value and real-time displacement offset of the target characteristic peak are obtained.

[0173] Among them, the real-time displacement offset is the absolute value of the difference between the real-time peak displacement value and the standard peak displacement value;

[0174] The real-time displacement offset is compared with the standard displacement offset.

[0175] If the real-time displacement offset is greater than the standard displacement offset, a Class II dyslipidemia risk result is obtained;

[0176] If the real-time displacement is less than or equal to the standard displacement, the risk level of dyslipidemia is considered normal.

[0177] In this embodiment, a confocal Raman spectrometer (Ar+ laser light source, 200mW power, 100x objective lens) was used at room temperature, with excitation wavelengths of 532, 785, and 830 nm. -1 The displacement range is 400-1800 cm. -1 Raman spectra of all samples were collected with an integration time of 20 s. The spectra were averaged four times. The LabSpec6 software package was used for spectral acquisition and analysis. A screening model was established using partial least squares method. The obtained spectral data and other data were analyzed using SPSS 21.0 software to complete the final analysis.

[0178] Peak area deviation analysis calculation formula:

[0179] PAD=|DPA-RPA|

[0180] Wherein, PAD is the real-time peak area difference, DPA is the peak area of ​​the real-time detected target feature peak, RPA is the standard peak area obtained by the peak area prediction model under the same basic index, and the peak area deviation threshold TA is set to 15 units.

[0181] If PAD > TA in the example, that is, DPA = 118 and RPA = 100, it is judged as a type of dyslipidemia risk, indicating that the lipid index has a significant deviation, which may show a high risk of pathogenic hyperlipidemia, and timely intervention or further testing of other related indicators is required.

[0182] If PAD≤TA in the embodiment, proceed to the next step of peak position deviation analysis;

[0183] Peak displacement deviation analysis calculation formula:

[0184] PSD = |DPS-RPS|

[0185] Wherein, PSD is the real-time peak shift value, DPS is the peak position of the real-time detected target characteristic peak, and RPS is the standard peak position, which is usually determined by healthy sample data or a prediction model. The peak shift deviation threshold TS is set to 2cm. -1 ;

[0186] In this embodiment, traditional blood lipid detection methods typically require complex sample processing and lengthy analysis procedures. Raman spectroscopy, compared to traditional biomarker methods, eliminates the need for sample labeling or pretreatment, simplifying the detection process and enabling rapid screening of large amounts of data. This is particularly beneficial for patients requiring frequent blood lipid level monitoring, such as diabetic patients or those at high risk of cardiovascular disease, assisting physicians in timely adjustments to treatment plans and optimizing patient health management. Regular monitoring of blood lipid levels using Raman spectroscopy can help detect high cholesterol or high triglycerides early, allowing for timely intervention and reducing the incidence of cardiovascular disease. For example, characteristic peaks of high cholesterol include C=C stretching vibrations: at approximately 1652 cm⁻¹. -1 The vibration at this location is produced by the stretching vibration of the C=C double bond in the cholesterol molecule; the C–H angular vibration occurs at approximately 1440 cm⁻¹. -1 and 1375cm -1 At this location, a distinct Raman signal is generated due to the C–H angular vibration of cholesterol; C–C stretching vibration: at approximately 1100 cm⁻¹ -1 The characteristic peaks of high triglycerides are indicated by the C–C stretching vibrations of cholesterol. These peaks also include C=O stretching vibrations: at approximately 1740 cm⁻¹. -1 At this location, the C=O stretching vibration of the ester group in the triglyceride molecule produces a characteristic peak; the C–H angular vibration: at approximately 1450 cm⁻¹ -1 and 1370cm -1 At a certain point, the C–H angular vibration of triglycerides produces a distinct Raman signal; the C–C stretching vibration: at approximately 1100 cm⁻¹ -1 At this location, the C–C stretching vibration of triglycerides will show a characteristic peak; therefore, based on Raman spectroscopy, the automated identification of disease risk levels can achieve timely screening of large amounts of data, improving the efficiency and accuracy of detection.

[0187] For example, if DPS = 1655cm -1 RPS = 1652cm -1 If PSD > TS, it is classified as a type II risk of dyslipidemia, which indicates that there are significant changes in the molecular structure of lipid indicators and a risk of secondary hyperlipidemia; if PSD ≤ TS, the deviation of the target characteristic peak in shift is within the normal range, and the degree of dyslipidemia risk is normal.

[0188] This step, based on the quantitative comparison method of DPA, RPA, DPS, and RPS, can finely classify the molecular characteristic data of blood lipids detected by metal Raman spectroscopy, thereby providing an accurate basis for subsequent adjustment of the detection cycle and improving the detection rate of dyslipidemia.

[0189] See Figure 4 As shown, this is a logic diagram for determining whether to adjust the preset inspection cycle according to an embodiment of the present invention.

[0190] Specifically, determining whether to adjust the preset inspection cycle based on the analysis results includes:

[0191] When blood lipid levels are abnormal, adjust the preset examination cycle to the corrected examination cycle;

[0192] Among them, abnormal blood lipid indicators include Class I blood lipid abnormality risk results and Class II blood lipid abnormality risk results; when the abnormal result is a Class I blood lipid abnormality result, the corrected examination cycle is the product of the preset examination cycle and the first adjustment step size, the first adjustment step size is the difference between 1 and the first adjustment degree, and the first adjustment degree is the ratio of the difference between the real-time peak area value and the standard peak area value to the real-time peak area value.

[0193] When the abnormal result is a type II dyslipidemia result, the corrected examination cycle is the product of the preset examination cycle and the second adjustment step size. The second adjustment step size is the difference between 1 and the second adjustment degree. The second adjustment degree is the ratio of the difference between the real-time displacement offset and the standard displacement offset to the real-time displacement offset.

[0194] In this embodiment, the first adjustment degree is calculated as the ratio of the real-time peak area difference (PAD) to the real-time peak area value (DPA):

[0195]

[0196] The first adjustment step size is:

[0197] S = 1 - A adj

[0198] The correction period is:

[0199] Correction cycle = Preset inspection cycle × S

[0200] If the preset inspection cycle is 30 days, the real-time peak area value = 100 units, the standard peak area value = 80 units, and the real-time peak displacement value = 1655 cm. -1 At this point, the first adjustment level = 0.2; the first adjustment step size = 0.8; and the correction period is 24 days.

[0201] If a type of dyslipidemia risk is established, with a significant abnormality in peak area, the patient needs to shorten the testing cycle to 24 days for more timely monitoring of lipid level changes and intervention. This step involves quantitatively analyzing the peak area and peak position of the target characteristic peak, comparing real-time detection data with preset standards, calculating the first and second adjustment levels, thereby determining the first and second adjustment step sizes, and correcting the original testing cycle. Shortening the retest cycle facilitates enhanced monitoring when abnormalities occur. This method achieves precise adjustment of the preset testing cycle based on dynamic quantitative analysis, strengthens clinical risk warning, and provides individualized intervention methods.

[0202] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0203] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A rapid blood lipid detection method based on spectral analysis, characterized in that, include, Historical blood lipid data and basic indicator information are obtained according to a preset examination cycle. The basic indicator information includes demographic indicators and general physical examination indicators. Demographic indicators include age, gender, race, residential area, occupation, dietary habits, drinking habits, and exercise habits. General physical examination indicators include height, weight, body temperature, blood pressure, pulse, and BMI. The historical blood lipid data is classified according to the preset blood lipid diagnostic criteria to obtain the results of the first indicator category or the second indicator category, and the corresponding detection method is selected based on the classification results. When the historical blood lipid data is determined to be in the second indicator category, a multi-source interference factor screening is performed on the historical blood lipid data. Based on the basic indicator information and the historical blood lipid data, it is determined whether the blood lipid level changes deviate from the normal state, and when the patient's blood lipid level changes deviate from the normal state, a multidimensional interference factor assessment is performed on the influencing factors of non-spontaneous blood lipid indicators to obtain the real-time comorbidity degree. Based on the comparison results between the real-time comorbidity correlation degree and the preset standard comorbidity correlation degree range, the blood lipid detection method is selected, including the first predictive detection method and the second predictive detection method; Among them, when the real-time co-morbidity correlation is greater than or equal to the minimum value of the standard interval, the first prediction detection method is used to perform feature analysis on the target detection molecule; When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is adopted, and the degree of dyslipidemia risk is determined by analyzing the peak area and peak position deviation of the target characteristic peak, and it is determined whether to adjust the preset examination cycle. The first predictive detection method is infrared spectroscopy detection, and the second predictive detection method is metal Raman spectroscopy detection; The first indicator category result is any blood lipid data that has been abnormal for more than three months; The results for the second indicator category show that all blood lipid data are within the normal range.

2. The rapid blood lipid detection method based on spectral analysis according to claim 1, characterized in that, Historical blood lipid data are categorized according to preset blood lipid diagnostic criteria, and appropriate testing methods are selected based on the categorization results. The blood lipid index category is determined according to the preset blood lipid diagnostic criteria, including the first index category and the second index category, and the corresponding test method is selected based on the classification result; When the classification result is the first indicator category, select the first detection method, which is venous blood sampling detection; When the classification result is the second indicator category, a multi-source interference factor screening step is performed on the historical blood lipid data.

3. The rapid blood lipid detection method based on spectral analysis according to claim 2, characterized in that, The steps for screening historical blood lipid data for multiple sources of confounding factors include: Obtain basic indicator information, and based on this information and historical blood lipid data, obtain blood lipid prediction data curves and healthy blood lipid data curves. Based on the deviation value and the trend of deviation value changes, determine whether the patient's blood lipid level changes deviate from the normal state. Obtain the deviation value change curve, fit a linear equation to the deviation value change curve, obtain the absolute value of the slope difference of the linear equation, and compare it with the standard slope value. When the absolute value of the slope difference is greater than the standard slope value, it is determined that the trend of blood lipid level changes deviates from the normal state, and non-spontaneous blood lipid index influencing factors are classified and eliminated. When the absolute value of the slope difference is less than or equal to the standard slope value, the trend of blood lipid level change is determined to be normal, and the current preset examination cycle is determined to be the re-examination cycle. Here, any deviation value is the difference between the data point on the blood lipid prediction data curve and the data point on the healthy blood lipid data curve corresponding to the same time series.

4. The rapid blood lipid detection method based on spectral analysis according to claim 3, characterized in that, The classification and elimination of factors influencing non-spontaneous lipid indicators includes an assessment of the impact of multidimensional confounding factors. A multidimensional interference factor impact assessment was performed based on basic indicator information, resulting in a multidimensional interference factor impact assessment score, which was then compared with the standard score of the multidimensional interference factor impact. When the multidimensional interference factor influence assessment score is greater than or equal to the multidimensional interference factor influence standard score, the influence of all relevant interference factors is reduced to the normal range before blood lipid testing is performed. When the multidimensional interference factor impact assessment score is lower than the multidimensional interference factor impact standard score, the impact of comorbidity parameters is assessed.

5. The rapid blood lipid detection method based on spectral analysis according to claim 3, characterized in that, The classification and elimination of factors influencing non-spontaneous lipid indicators also includes an assessment of the impact of comorbidity, among which... Several comorbidity parameters were obtained based on statistical analysis of past medical history data and clinical data in the database. If no prior medical history is determined, the second predictive testing method will be implemented; If a pre-existing medical history is determined, the impact of comorbidity on comorbidity parameters is further assessed to obtain the real-time comorbidity. The real-time comorbidity correlation is compared with the standard comorbidity correlation interval. When the real-time comorbidity correlation is greater than or equal to the minimum value of the standard comorbidity correlation interval, the first predictive detection method is executed, and the corresponding detection content is adjusted according to the abnormal comorbidity parameters. When the real-time comorbidity correlation is less than the minimum value of the standard comorbidity correlation interval, the second predictive detection method is executed.

6. The rapid blood lipid detection method based on spectral analysis according to claim 5, characterized in that, Predictive detection methods include, The predictive detection methods include a first predictive detection method and a second predictive detection method; The first predictive detection method is infrared spectroscopy for blood lipid detection, and the infrared spectroscopy detection content is adjusted based on the comorbidity parameters that have the greatest impact on the real-time comorbidity correlation. The second predictive detection method is metal Raman spectroscopy detection. The detection process involves acquiring several detection indicators and corresponding detection analysis results, and determining whether to adjust the preset inspection cycle based on the analysis results.

7. The rapid blood lipid detection method based on spectral analysis according to claim 6, characterized in that, Several detection indicators were obtained, including The detection indicators include peak area deviation and peak displacement deviation; Obtain the raw Raman spectral data and preprocess it to obtain the target Raman spectral data; Based on the required detection content, the corresponding features of the target Raman spectral data are extracted to obtain several target feature peaks; Perform peak area deviation analysis and peak displacement deviation analysis on any of the target characteristic peaks, and determine whether to adjust the preset inspection cycle based on the analysis results.

8. The rapid blood lipid detection method based on spectral analysis according to claim 7, characterized in that, The preprocessing yields the target Raman spectral data, including: Random noise in the original Raman spectrum was removed using a smoothing algorithm, and baseline drift was removed by polynomial fitting to obtain intermediate Raman spectral data; The intermediate Raman spectral data are normalized to obtain the target Raman spectral data.

9. The rapid blood lipid detection method based on spectral analysis according to claim 7, characterized in that, Performing peak area deviation analysis and peak displacement deviation analysis on any of the target characteristic peaks includes... Based on the peak area prediction model, the standard peak area value of any of the target characteristic peaks under the basic index information is obtained; Obtain the real-time peak area value and the real-time peak area difference of the target characteristic peak; Among them, the real-time peak area difference is the absolute value of the difference between the real-time peak area value and the standard peak area value; The real-time peak area difference is compared with the standard peak area difference. If the real-time peak area difference is greater than the standard peak area difference, a risk result of dyslipidemia is obtained. If the real-time peak area difference is less than or equal to the standard peak area difference, the standard peak displacement value corresponding to the target characteristic peak is obtained based on the peak displacement prediction model, and the real-time peak displacement value and real-time displacement offset of the target characteristic peak are obtained. Among them, the real-time displacement offset is the absolute value of the difference between the real-time peak displacement value and the standard peak displacement value; The real-time displacement offset is compared with the standard displacement offset. If the real-time displacement offset is greater than the standard displacement offset, a Class II dyslipidemia risk result is obtained; If the real-time displacement is less than or equal to the standard displacement, the risk level of dyslipidemia is considered normal.

10. The rapid blood lipid detection method based on spectral analysis according to claim 9, characterized in that, Determining whether to adjust the preset inspection cycle based on the analysis results includes: When blood lipid levels are abnormal, adjust the preset examination cycle to the corrected examination cycle; Among them, abnormal blood lipid indicators include Class I and Class II risk outcomes for dyslipidemia. When the abnormal result is a type of dyslipidemia, the corrected examination cycle is the product of the preset examination cycle and the first adjustment step size. The first adjustment step size is the difference between 1 and the first adjustment degree. The first adjustment degree is the ratio of the difference between the real-time peak area value and the standard peak area value to the real-time peak area value. When the abnormal result is a type II dyslipidemia result, the corrected examination cycle is the product of the preset examination cycle and the second adjustment step size. The second adjustment step size is the difference between 1 and the second adjustment degree. The second adjustment degree is the ratio of the difference between the real-time displacement offset and the standard displacement offset to the real-time displacement offset.

Images