Abnormality identification method and system in coagulation process based on big data analysis

Abstract

The application discloses a method and system for identifying abnormalities in the coagulation process based on big data analysis, comprising: combining standard coagulation detection and thromboelastography dynamic monitoring to realize the leap from static indicators to functional whole-process evaluation, while combining patient individual characteristics to screen similar historical cases, and accurately identifying key coagulation indicators; introducing correlation analysis to dynamically screen key indicators, and using clustering analysis to establish an adaptive normal range, and finally evaluating abnormalities by quantifying the degree of deviation, the application significantly improves the accuracy and timeliness of coagulation state monitoring, provides individualized early warning and intervention basis for clinical practice, effectively prevents bleeding or thrombosis risk, and promotes the intelligent and refined development of coagulation management.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of data processing, and relates to an abnormality identification method and system in a blood coagulation process based on big data analysis. BACKGROUND

[0002] The blood coagulation process is a key link in maintaining a dynamic balance between normal hemostasis and preventing thrombus formation. When the blood coagulation function is abnormal, it may cause bleeding or thrombotic diseases, and even life-threatening in severe cases. Therefore, the abnormality identification of the blood coagulation process plays an important role in preventing and controlling bleeding diseases, reducing surgical risks, and preventing thrombosis.

[0003] The commonly used blood coagulation process abnormality identification method is standard blood coagulation detection, that is, through a series of laboratory tests (such as prothrombin time (PT), activated partial thromboplastin time (APTT), thrombin time (TT), etc.), the blood coagulation function, bleeding risk and thrombosis tendency of the patient are evaluated. Although this method is an important means of clinical blood coagulation evaluation, it plays an important role in preoperative evaluation, bleeding disease screening, and anticoagulant drug monitoring, but the judgment of blood coagulation abnormalities depends on the experience of doctors or unified normal standards, and cannot provide objective and personalized evaluation results for different patient conditions.

[0004] Considering the differences in medical history, age, physiological state, etc., the same blood coagulation indicators may have different meanings for different patients, that is, even under a unified standard, if the blood coagulation indicators of a patient exceed the normal range, it does not necessarily mean that there is an abnormality, and another patient may have blood coagulation dysfunction even if the indicators are within the normal range; using a unified standard to identify patient blood coagulation process abnormalities often leads to inaccurate identification results. SUMMARY

[0005] The purpose of the present application is to solve the problem in the prior art that the same blood coagulation indicators have different normal ranges for different patients due to differences in age, medical history, etc., and using a unified standard to identify patient blood coagulation process abnormalities often leads to inaccurate identification results, and to provide an abnormality identification method and system in a blood coagulation process based on big data analysis.

[0006] To achieve the above object, the following technical solutions are adopted: an abnormality identification method in a blood coagulation process based on big data analysis, comprising: obtaining blood coagulation index data of a current patient through standard blood coagulation detection, and obtaining a thromboelastogram of the current patient by using a TEG device; obtaining the similarity of the age, disease background and drug use of the patient in the historical data to the characteristics of the current patient, and then screening out target data reflecting the blood coagulation condition of the current patient; obtaining the reflection degree value of each blood coagulation index data on the normal state of the blood coagulation process of the current patient according to the correlation of each blood coagulation index data in the target data and the change of the thromboelastogram, and sorting all the reflection degree data, and selecting a number of blood coagulation indexes with the highest reflection degree as the blood coagulation indexes that have a significant impact on the blood coagulation process of the patient; regarding the number of blood coagulation indexes with the highest reflection degree obtained as target blood coagulation indexes, and drawing a coordinate system according to the target blood coagulation indexes, putting all the target historical patient data into the coordinate system, clustering the data of the coordinate system, and obtaining the cluster data with the highest density as the normal patient data, and then determining the normal range of the target blood coagulation indexes; judging whether the blood coagulation index data of the current patient is within the normal range of the target blood coagulation indexes, if not, obtaining the deviation degree of the blood coagulation index data of the current patient from the normal range of the target blood coagulation indexes according to the maximum and minimum values of the blood coagulation index data of the current patient and the normal range of the target blood coagulation indexes, and then obtaining the blood coagulation abnormality degree of the current patient.

[0007] Further improvement of the application is that the parameters contained in the thromboelastogram are: R time, K time, alpha angle, MA value and Ly30; wherein, R time is the time from liquid blood to fibrin formation; K time is the time from blood coagulation to a certain intensity; alpha angle is the fibrin formation rate; MA value is the final strength of the blood clot; Ly30 is the thrombus degradation rate within 30 minutes; the blood coagulation index data includes prothrombin time PT, international normalized ratio INR, activated partial thromboplastin time APTT, thrombin time TT, fibrinogen FIB and D-dimer and platelet count PLT.

[0008] Further, the similarity of the age, disease background and drug use of the patient in the historical data to the characteristics of the current patient is obtained, specifically: the similarity of the patient age is obtained according to the age difference of the patient age in the historical data and the current patient age, specifically: wherein, represents the age similarity of the i-th historical data patient and the current patient, represents the age of the i-th patient in the historical data, represents the age of the current patient, represents the absolute value of the age difference between the i-th patient in the historical data and the current patient; It is explained that the smaller the absolute value of the age difference is, the higher the age similarity is; ensure that the sub-phrase is meaningful and ; the similarity of the disease background of the patient, according to the disease coincidence of the patient in the historical data and the current patient, if the patient in the historical data and the current patient have the same disease, the similarity is high, and if the patient in the historical data has the corresponding disease but the current patient does not, the similarity is reduced, if the patient in the historical data and the current patient have or do not have a certain disease, count, otherwise do not count, thereby calculating the disease background similarity of the patient in the historical data and the current patient, specifically: Wherein, represents the disease similarity of the i-th historical data patient and the current patient, represents the total number of diseases that affect the coagulation index, represents the count value of the i-th historical data patient corresponding to the d-th disease; the similarity of the patient's drug use is calculated in the same way as the disease similarity; if the patient in the historical data and the current patient take or do not take a certain drug at the same time, count 1, if the patient in the historical data takes a certain drug, and the current patient does not take the drug, otherwise do not count 0, thereby calculating the similarity of the drug use of the patient in the historical data and the current patient, specifically: Wherein, represents the drug use similarity of the i-th historical data patient and the current patient, represents the total number of drugs that affect the coagulation index, represents the count value of the i-th historical data patient corresponding to the b-th drug.

[0009] Further, the target data reflecting the coagulation condition of the current patient is screened out, specifically; the similarity value of each historical data patient and the current patient is obtained by combining the similarity of the patient in the historical data and the current patient in terms of age, disease and drug use; it is judged whether the obtained similarity value is greater than the set threshold value, if greater, it is considered that the patient corresponding to the historical data and the current patient have high similarity, and the coagulation process of the patient corresponding to the historical data can reflect the abnormal condition of the coagulation process of the current patient, all target data that can reflect the coagulation condition of the current patient are screened out; the similarity value of each historical data patient and the current patient is obtained, specifically: Wherein, represents the similarity of the i-th historical data patient and the current patient.

[0010] Further, the correlation between each coagulation index data in the target data and the thromboelastography change is specifically: step 1: selecting any coagulation index data in the target data and any parameter in the thromboelastography, and performing normalization processing on the selected coagulation index data and parameter; step 2: taking the normalized coagulation index data as the abscissa, the parameter as the ordinate, and the values corresponding to all target data as data into the coordinates; step 3: dividing the abscissa into several segments, taking the center point coordinate value of each segment as the abscissa value, and taking the average value of all ordinate data in the segment as the ordinate value corresponding to the center point, so that the abscissa data and the ordinate data are one-to-one corresponding, and a coagulation index data-parameter coordinate graph is obtained; step 4: fitting the data in the coordinate graph by using the least square method to obtain a fitting curve; step 5: obtaining the influence degree of the coagulation index data on the parameter according to the slope of the fitting curve; step 6: repeating steps 1 to 5 to obtain the influence degree of the coagulation index data on other parameters in the thromboelastography; step 7: obtaining the reflection degree value of the coagulation index data on the patient's coagulation process based on the influence degree of the coagulation index data on all parameters in the thromboelastography; and step 8: repeating steps 1 to 7 to obtain the reflection degree value of all coagulation index data in the target data on the patient's coagulation process.

[0011] Further, the influence degree of the coagulation index data on the parameter is obtained according to the slope of the fitting curve, and specifically, taking the coagulation index data INR in the target data and the parameter R time in the thromboelastography as an example, wherein, represents the influence degree of the coagulation index data INR on R time, represents the number of selected points on the fitting curve, represents the slope value of the oth point on the “INR-R time” fitting curve, represents the absolute value of the slope difference between two adjacent points on the fitting curve, The smaller the absolute value of the slope difference is, the more consistent the slopes at different positions on the curve are, and the higher the correlation degree between the INR coagulation index and R time is. The average slope of different points on the fitting curve represents the overall trend of the fitting curve, and the greater the absolute value of the average slope is, the greater the influence degree of INR on R time is. The reflection degree value of the coagulation index data on the patient's coagulation process is obtained based on the influence degree of the coagulation index data on all parameters in the thromboelastography, and specifically, the influence degree values of the coagulation index data on all parameters in the thromboelastography are added and averaged, and the obtained average value is taken as the reflection degree value of the coagulation index data on the patient's coagulation process.

[0012] Further, according to the target coagulation index, a coordinate system is drawn, all target historical patient data are put into the coordinate system, the data in the coordinate system are clustered, and the cluster data with the highest density are obtained as normal patient data, and then the normal range of the target coagulation index is determined. Specifically, the obtained target coagulation index is considered to be the index most closely related to the patient's coagulation process, so the target coagulation index of the patient with normal coagulation should be within the normal range; the normal data are concentrated in the same range, while the abnormal data of coagulation are usually different due to different influencing factors causing the abnormality, and the abnormality is also quite different; therefore, according to the target coagulation index, a coordinate system is drawn, and all target historical patient data are put into the coordinate system; wherein the target historical patient data are data with high similarity to the current patient characteristics screened out; the data with high similarity refers to the historical data of the patient similar to the current patient in age, disease background and medication, which can reflect the target data of the current patient's coagulation; the data in the coordinate system are clustered by using the DBSCAN clustering method, and then a plurality of clusters and single point data are obtained, the single point noise data are excluded, the cluster data with the highest density are obtained from all clusters, and are used as normal patient data; according to the maximum and minimum values of the target coagulation index corresponding to the coagulation normal patient data, the normal range of the target coagulation index corresponding to the patient is determined.

[0013] Further, according to the target coagulation index, a coordinate system is drawn, all target historical patient data are put into the coordinate system, the data in the coordinate system are clustered, and the cluster data with the highest density are obtained as normal patient data, and then the normal range of the target coagulation index is determined. Specifically, the obtained target coagulation index is considered to be the index most closely related to the patient's coagulation process, so the target coagulation index of the patient with normal coagulation should be within the normal range; the normal data are concentrated in the same range, while the abnormal data of coagulation are usually different due to different influencing factors causing the abnormality, and the abnormality is also quite different; therefore, according to the target coagulation index, a coordinate system is drawn, and all target historical patient data are put into the coordinate system; wherein the target historical patient data are data with high similarity to the current patient characteristics screened out; the data with high similarity refers to the historical data of the patient similar to the current patient in age, disease background and medication, which can reflect the target data of the current patient's coagulation; the data in the coordinate system are clustered by using the DBSCAN clustering method, and then a plurality of clusters and single point data are obtained, the single point noise data are excluded, the cluster data with the highest density are obtained from all clusters, and are used as normal patient data; according to the maximum and minimum values of the target coagulation index corresponding to the coagulation normal patient data, the normal range of the target coagulation index corresponding to the patient is determined. ; otherwise, according to the deviation of the current patient data relative to the normal range, the abnormality degree of the current patient is calculated as: wherein, represents the abnormality degree of the u-th target coagulation index of the current patient, represents the corresponding value of the u-th target coagulation index of the current patient, represents the minimum value of the u-th target coagulation index normal range, represents the maximum value of the u-th target coagulation index normal range; represents the minimum value function.

[0014] Further, the coagulation abnormality degree of the current patient is obtained, specifically: the coagulation abnormality degree of the current patient is calculated by comprehensively considering the abnormality degree of the patient under each target coagulation index, wherein, represents the coagulation abnormality degree of the current patient, represents the number of target coagulation indexes, represents the abnormality degree of the u-th target coagulation index of the current patient.

[0015] The abnormality identification system in the blood coagulation process based on big data analysis comprises: a first acquisition module, which obtains the blood coagulation index data of the current patient through standard blood coagulation detection, and obtains the thromboelastogram of the current patient by using a TEG device; a second acquisition module, which obtains the similarity of the age, disease background and drug use of the patient in the historical data to the characteristics of the current patient, and then screens out target data reflecting the blood coagulation condition of the current patient; a selection module, which obtains the reflection degree value of each blood coagulation index data on the normal state of the blood coagulation process of the current patient according to the correlation of each blood coagulation index data in the target data with the change of the thromboelastogram, sorts all the reflection degree data, and selects a plurality of blood coagulation indexes with the highest reflection degree as the blood coagulation indexes that have obvious influence on the blood coagulation process of the patient; a clustering module, which regards the plurality of blood coagulation indexes with the highest reflection degree as target blood coagulation indexes, draws a coordinate system according to the target blood coagulation indexes, puts all the target historical patient data into the coordinate system, clusters the data of the coordinate system, and obtains the cluster data with the highest density as the normal patient data, so as to determine the normal range of the target blood coagulation index; and a judgment module, which judges whether the blood coagulation index data of the current patient is in the normal range of the target blood coagulation index, and if not, obtains the deviation degree of the blood coagulation index data of the current patient from the normal range of the target blood coagulation index according to the maximum value and the minimum value of the blood coagulation index data of the current patient and the normal range of the target blood coagulation index, and then obtains the blood coagulation abnormality degree of the current patient.

[0016] Compared with the prior art, the present application has the following beneficial effects: the present application combines standard blood coagulation detection with thromboelastogram dynamic monitoring, realizes the leap from static indicators to functional full-process evaluation, simultaneously combines patient individualized characteristics to screen similar historical cases, and accurately identifies key blood coagulation indicators; the present application introduces correlation analysis to dynamically screen key indicators, adopts clustering analysis to establish an adaptive normal range, and finally evaluates the abnormality by quantifying the deviation degree, which significantly improves the accuracy and timeliness of blood coagulation state monitoring, provides individualized early warning and intervention basis for the clinic, effectively prevents the risk of bleeding or thrombosis, and promotes the intelligent and refined development of blood coagulation management. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0018] Figure 1 The flowchart of the blood coagulation process abnormality identification method based on big data analysis of the present application; Figure 2A structural schematic diagram of the abnormality identification system in the blood coagulation process based on big data analysis of the present application. DETAILED DESCRIPTION

[0019] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0021] It should be noted that: similar reference numerals and letters represent similar items in the following drawings, therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0022] In the description of the embodiments of the present application, it should be noted that, if the orientation or position relationship indicated by the terms "upper", "lower", "horizontal", "inner" and the like is based on the orientation or position relationship shown in the drawings, or is the orientation or position relationship when the product of the present application is usually placed, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the indicated device or element must have a particular orientation, be constructed and operated in a particular orientation, therefore, it cannot be understood as a limitation on the present application. In addition, the terms "first", "second" and the like are only used for differentiation in description, and cannot be understood as indicating or implying relative importance.

[0023] In addition, if the term "horizontal" appears, it does not mean that the component must be absolutely horizontal, but can be slightly inclined. For example, "horizontal" only means that its direction is relatively more horizontal than "vertical", and does not mean that the structure must be completely horizontal, but can be slightly inclined.

[0024] In the description of the embodiments of the present application, it should also be noted that, unless otherwise explicitly specified and limited, if the terms "arrangement", "installation", "connection", "connection" appear, they should be understood in a broad sense, for example, can be fixedly connected, can be detachably connected, or integrally connected; can be mechanically connected, can be electrically connected; can be directly connected, can be indirectly connected through an intermediate medium, can be the communication inside two elements. For those of ordinary skill in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0025] The application will be further described in detail below with reference to the accompanying drawings: Figure 1 The application discloses an abnormality identification method in a blood coagulation process based on big data analysis, which comprises the following steps: S101, obtaining blood coagulation index data of a current patient through standard blood coagulation detection, and obtaining a thromboelastogram of the current patient by using a TEG device; parameters contained in the thromboelastogram are R time, K time, an alpha angle, an MA value and Ly30; wherein the R time is the time from liquid blood to fibrin formation; the K time is the time from blood coagulation to a certain intensity; the alpha angle is the fibrin formation rate; the MA value is the final strength of the blood clot; and the Ly30 is the thrombus degradation rate within 30 minutes; the blood coagulation index data comprises prothrombin time PT, international normalized ratio INR, activated partial thromboplastin time APTT, thrombin time TT, fibrinogen FIB, D-dimer and platelet count PLT.

[0026] S102, obtaining the similarity of the age, disease background and drug use of the patient in the historical data to the characteristics of the current patient, and then screening out target data reflecting the blood coagulation condition of the current patient; the similarity of the patient age is obtained according to the age difference between the patient age in the historical data and the current patient age, and specifically is: Wherein, represents the age similarity of the i th historical data patient to the current patient, represents the age of the i th patient in the historical data, represents the age of the current patient, represents the absolute value of the age difference between the i th patient in the historical data and the current patient; It is explained that the smaller the absolute value of the age difference is, the higher the age similarity is; to ensure that the fraction is meaningful and ; the similarity of the disease background of the patient is obtained according to the disease coincidence degree between the patient in the historical data and the current patient, if the patient in the historical data has the same disease as the current patient, the similarity degree is higher; if the patient in the historical data has the corresponding disease but the current patient does not, the similarity degree is reduced; if the patient in the historical data and the current patient have or do not have a certain disease at the same time, the count is counted, otherwise, the count is not counted, thereby calculating the disease background similarity between the patient in the historical data and the current patient, and specifically is: Wherein, represents the disease similarity of the i th historical data patient to the current patient, represents the total number of diseases affecting the blood coagulation index, represents the count value of the i-th historical data patient corresponding to the d-th disease; the calculation process of the drug use similarity and the disease similarity of the patient is the same; if the patient in the historical data and the current patient take or do not take a certain drug at the same time, count 1, if the patient in the historical data takes a certain drug, and the current patient does not take the drug, otherwise, do not count 0, thereby counting the similarity of the drug use of the patient in the historical data and the current patient, specifically: wherein, represents the drug use similarity of the i-th historical data patient and the current patient, represents the total number of drugs affecting the coagulation index, represents the count value of the i-th historical data patient corresponding to the b-th drug.

[0027] The target data reflecting the coagulation condition of the current patient is screened out, specifically; the similarity value of the patient in each historical data and the current patient is obtained by combining the similarity of the patient in the historical data and the current patient in terms of age, disease and drug use; it is judged whether the obtained similarity value is greater than the set threshold value, if greater, it is considered that the similarity of the patient corresponding to the historical data and the current patient is high, and the coagulation process of the patient corresponding to the historical data can reflect the abnormal condition of the coagulation process of the current patient, and all target data capable of reflecting the coagulation condition of the current patient are screened out; the similarity value of the patient in each historical data and the current patient is obtained, specifically: wherein, represents the similarity of the i-th historical data patient and the current patient.

[0028] S103, obtaining a reflecting degree value of each coagulation index data in the target data on the normal state of the current patient's coagulation process according to the correlation between each coagulation index data in the target data and the change of the thromboelastogram, and sorting all the reflecting degree data, and selecting several coagulation indexes with the highest reflecting degree as the coagulation indexes that have a significant impact on the patient's coagulation process; the correlation between each coagulation index data in the target data and the change of the thromboelastogram is as follows: S103.1: selecting any coagulation index data in the target data and any parameter in the thromboelastogram, and performing normalization processing on the selected coagulation index data and parameter respectively; S103.2: taking the normalized coagulation index data as the abscissa and the parameter as the ordinate, and placing the data into the coordinates according to the values corresponding to all target data; S103.3: dividing the abscissa into several segments, taking the center point coordinate value of each segment as the abscissa value, taking the average value of all ordinate data corresponding to the segment as the ordinate value corresponding to the center point, and obtaining a coagulation index data-parameter coordinate graph by making the abscissa data and the ordinate data correspond to each other; S103.4: fitting the data in the coordinate graph by using the least square method to obtain a fitting curve; S103.5: obtaining the influence degree of the coagulation index data on the parameter according to the slope of the fitting curve; S103.6: repeating S103.1 to S103.5 to obtain the influence degree of the coagulation index data on other parameters in the thromboelastogram; S103.7: obtaining the reflecting degree value of the coagulation index data on the patient's coagulation process based on the influence degree of the coagulation index data on all parameters in the thromboelastogram; S103.8: repeating S103.1 to S103.7 to obtain the reflecting degree value of all coagulation index data in the target data on the patient's coagulation process.

[0029] According to the slope of the fitting curve, the influence degree of the coagulation index data on the parameter is obtained, for example, taking the coagulation index data INR in the target data and the parameter R time in the thromboelastogram as an example, wherein, represents the influence degree of the coagulation index data INR on R time, represents the number of selected points on the fitting curve, represents the slope value of the oth point on the "INR-R time" fitting curve, represents the absolute value of the slope difference of the adjacent two points on the fitting curve, The smaller the absolute value of the slope difference is, the more consistent the slopes at different positions on the curve are, and the higher the correlation degree of the INR coagulation index and R time is; The overall trend of the fitting curve is represented by the average slope of different points on the fitting curve, and the greater the absolute value of the average slope, the greater the degree of influence of INR on R time; the degree of influence of the coagulation index data on all parameters in the thrombelastogram is obtained to obtain the reflection degree value of the coagulation index data on the patient's coagulation process, specifically: the influence degree values of the coagulation index data on all parameters in the thrombelastogram are added and averaged, and the obtained average value is taken as the reflection degree value of the coagulation index data on the patient's coagulation process.

[0030] S104, based on the obtained several coagulation indexes with the highest reflection degree as the target coagulation index, and according to the target coagulation index, a coordinate system is drawn, all target historical patient data are put into the coordinate system, the data of the coordinate system are clustered, and the cluster data with the highest density are obtained as normal patient data, and then the normal range of the target coagulation index is determined; the obtained target coagulation index is considered as the index most closely related to the patient's coagulation process, so the target coagulation index of the patient with normal coagulation should be within the normal range; the normal data are concentrated in the same range, while the abnormal data of coagulation are usually different due to different influencing factors causing the abnormality, and the abnormality is also quite different; therefore, according to the target coagulation index, a coordinate system is drawn, and all target historical patient data are put into the coordinate system; the data in the coordinate system are clustered by using the DBSCAN clustering method, a plurality of clusters and single point data are obtained, the single point noise data are excluded, the cluster data with the highest density are obtained from all clusters, and are taken as normal patient data; according to the maximum and minimum values of the corresponding target coagulation index in the coagulation normal patient data, the normal range of the target coagulation index corresponding to the patient is determined.

[0031] S105, it is judged whether the coagulation index data of the current patient is in the normal range of the target coagulation index, if not, the deviation degree of the coagulation index data of the current patient from the normal range of the target coagulation index is obtained according to the maximum and minimum values of the coagulation index data of the current patient and the normal range of the target coagulation index, and then the coagulation abnormality degree of the current patient is obtained.

[0032] If the target coagulation index data of the current patient is in the normal range of the target coagulation index, the coagulation index abnormality degree is recorded as ; otherwise, according to the deviation degree of the current patient data relative to the normal range, the abnormality degree is calculated: wherein, represents the abnormality degree of the u-th target coagulation index of the current patient, represents the corresponding value of the u-th target coagulation index of the current patient, represents the minimum value of the u-th target coagulation index normal range, represents the maximum value of the u-th target coagulation index normal range. denotes a minimum value function.

[0033] The abnormality degree of the current patient is obtained, specifically: the abnormality degrees of the patient under each target coagulation index are integrated, and the abnormality degree of the current patient is calculated, Wherein, denotes the abnormality degree of the current patient, denotes the number of target coagulation indexes, denotes the abnormality degree of the u-th target coagulation index of the current patient.

[0034] Referring to Figure 2 The application discloses a coagulation process abnormality identification system based on big data analysis, which comprises a first acquisition module, a second acquisition module, a selection module and a clustering module. The first acquisition module obtains coagulation index data of a current patient through standard coagulation detection, and obtains a thromboelastogram of the current patient by using a TEG device. The second acquisition module obtains the feature similarity between the age, disease background and drug use of a patient in historical data and the current patient, and then screens out target data reflecting the coagulation condition of the current patient. The selection module obtains a reflection degree value of each coagulation index data on the normal state of the coagulation process of the current patient according to the correlation between each coagulation index data in the target data and the change of the thromboelastogram, sorts all the reflection degree data, and selects a plurality of coagulation indexes with the highest reflection degree as coagulation indexes that have a significant influence on the coagulation process of the patient. The clustering module regards the plurality of coagulation indexes with the highest reflection degree as target coagulation indexes, draws a coordinate system according to the target coagulation indexes, puts all the target historical patient data into the coordinate system, clusters the data of the coordinate system, and obtains the cluster data with the highest density as normal patient data, so as to determine the normal range of the target coagulation indexes. The judgment module judges whether the coagulation index data of the current patient is in the normal range of the target coagulation indexes, and if not, obtains the deviation degree of the coagulation index data of the current patient from the normal range of the target coagulation indexes according to the maximum value and the minimum value of the coagulation index data of the current patient and the normal range of the target coagulation indexes, and then obtains the abnormality degree of the coagulation of the current patient.

[0035] Embodiment: The application discloses a coagulation process abnormality identification method based on big data analysis, which comprises the following steps: S1, obtaining coagulation index data of a patient through standard coagulation detection, and obtaining a thromboelastogram (TEG) of the patient by using a TEG device.

[0036] Because the possibility of abnormality in the coagulation process of a patient mainly depends on the coagulation index data obtained by standard coagulation detection items, the blood sample of the patient is first collected by venous blood sampling. At least 2 ml of blood is collected to ensure the accuracy of the concentration of coagulation factors, and 3.2% sodium citrate anticoagulant is added. After blood sampling, mix gently to avoid clot formation, and send for examination within 30 minutes to prevent degradation of coagulation factors. After blood sampling is completed, the laboratory is detected by the automatic coagulation analyzer ACL TOP, and the data during the detection process are recorded, including PT (prothrombin time), INR (international normalized ratio (PT standardized ratio)), APTT (activated partial thromboplastin time), TT (thrombin time), FIB (fibrinogen), D-dimer and platelet count (PLT).

[0037] Because the standard coagulation detection cannot reflect the dynamic changes of the coagulation process, in order to analyze the correlation between the coagulation index of the patient obtained by the standard coagulation detection and the abnormality of the coagulation process of the patient, the thromboelastography (TEG) of the patient also needs to be obtained, so part of the collected blood sample (1-2 ml) is used to draw the coagulation process curve by the TEG device. According to the coagulation process curve, important parameters are recorded, including R time (from liquid blood to fibrin formation), K time (from coagulation to a certain intensity), alpha angle (fibrin generation rate (curve slope)), MA value (final strength of blood clot) and Ly30 (thrombus degradation rate within 30 minutes).

[0038] S2, according to the similarity of the historical patients and the current patient, and the degree of manifestation of different coagulation indexes to the abnormality of the coagulation process of the patient, the normal range of the coagulation index for the current patient is obtained.

[0039] Because the coagulation process of the patient is greatly affected by the age, disease and medication of the patient, etc., under the influence of these factors, the unified coagulation index normal range standard may not be suitable for different patients, therefore, according to the similarity of the patient and the current patient in the historical data in these characteristics, the target data reflecting the coagulation state of the current patient is screened, and the influence of these factors on the accuracy of the abnormality recognition result of the coagulation process of the patient is reduced. In addition, for different patients, each coagulation index reflects the degree of abnormality of the coagulation process differently, such as the coagulation process of the patient may have higher tolerance to some coagulation indexes, that is, when these indexes change greatly, the coagulation process of the patient is still normal, so according to the correlation between each coagulation index and the manifestation of the abnormality of the coagulation process of the patient, the coagulation index that best reflects the abnormality of the coagulation process of the current patient is obtained, so as to reduce the influence of the change of other indexes with smaller reaction degree on the final result.

[0040] S2.1, according to the similarity of the characteristics of the patient in the historical data and the current patient, the target data reflecting the coagulation condition of the current patient is screened.

[0041] Because the coagulation process of different patients will be affected by individual differences, disease state, age and other factors, the same coagulation indicators will have different normal ranges for different patients. For example, the coagulation factor of newborns has not yet matured, and PT and APTT are physiologically prolonged, but not abnormal; the fibrinogen and D-dimer of the elderly are physiologically elevated, which may increase the risk of thrombosis, but PT / APTT may still be normal; the PT, APTT and D-dimer of DIC patients will all be elevated. Therefore, the patient data in the historical data that is more similar to the current patient's characteristics can accurately reflect the normal range of the current patient's coagulation indicators. Therefore, by calculating the similarity of the patient in the historical data and the current patient's characteristics, the target data reflecting the current patient's coagulation condition is selected.

[0042] The main influencing factors of the normal range of the patient's coagulation indicators are the patient's age, disease background and drug use, so the similarity of the patient in the historical data and the current patient is calculated from these aspects.

[0043] The patient age similarity is obtained according to the age difference between the patient in the historical data and the current patient: wherein, represents the age similarity of the i-th patient in the historical data and the current patient, represents the age of the i-th patient in the historical data, represents the age of the current patient, represents the absolute value of the age difference between the i-th patient in the historical data and the current patient. It is explained that the smaller the absolute value of the age difference, the higher the age similarity. to ensure that the fraction is meaningful and .

[0044] The disease background similarity of the patient is calculated according to the disease coincidence degree between the patient in the historical data and the current patient. If the patient in the historical data and the current patient have the same disease, the similarity is high; if the patient in the historical data has the corresponding disease but the current patient does not, the similarity is reduced. The main diseases affecting the patient's coagulation indicators include liver disease, hemophilia, DIC (disseminated intravascular coagulation), APS (antiphospholipid syndrome), deep vein thrombosis (DVT), pulmonary embolism (PE) and pregnancy, so statistics are made from these diseases. If the patient in the historical data and the current patient have or do not have a certain disease at the same time, count, otherwise do not count, for example: if the patient in the historical data and the current patient have or do not have hemophilia at the same time, the disease is counted ; if the patient in the historical data has DIC disease, but the current patient does not, the count is . Thus, the disease background similarity of the patient in the historical data and the current patient is calculated: wherein, represents the disease similarity of the ith historical data patient to the current patient, represents the total number of diseases affecting the coagulation index (here ), represents the count value of the ith historical data patient corresponding to the dth disease.

[0045] The drug use similarity of the patient is calculated in the same way as the disease similarity. Since the main drugs affecting the coagulation process of the patient include warfarin, ordinary heparin (UFH), low molecular weight heparin (LMWH), Xa inhibitor (rivaroxaban, apixaban) and direct thrombin inhibitor (dabigatran), the similarity of the drug use of the patient in the historical data to the current patient is calculated as follows: wherein, represents the drug use similarity of the ith historical data patient to the current patient, represents the total number of drugs affecting the coagulation index, represents the count value of the ith historical data patient corresponding to the bth drug.

[0046] The similarity of each historical data patient to the current patient is calculated in combination with the similarities of the patient in the historical data to the current patient in terms of age, disease and drug use. wherein, represents the similarity of the ith historical data patient to the current patient, represents the age similarity of the ith historical data patient to the current patient, represents the disease similarity of the ith historical data patient to the current patient, represents the drug use similarity of the ith historical data patient to the current patient. represents the average similarity of the ith historical data patient to the current patient in these aspects. When , it is considered that the historical data corresponding to the patient has a high similarity to the current patient, i.e. can reflect the abnormal condition of the coagulation process of the current patient, and thus all target data that can reflect the coagulation condition of the current patient are selected.

[0047] S2.2, according to the relationship between the thromboelastogram and the coagulation index of the patient in the target data, the normal range of the coagulation index is obtained.

[0048] Because of the influence of patient's own characteristics, different coagulation indicators show different degrees of abnormal coagulation in patients, that is, for some patients, even if a coagulation indicator data is greatly deviated, the patient's coagulation process is still normal, for example, the INR of a newborn can reach 1.5-2.0, but it is still normal and does not need to be intervened. Therefore, first, according to the correlation of each coagulation indicator data in the target data and the thromboelastogram change, the degree of reflection of each coagulation indicator data on the normal state of the current patient's coagulation process is obtained, and the coagulation indicators that have a significant impact on the patient's coagulation process are screened out.

[0049] Because when the patient's coagulation process is normal, the patient's coagulation indicator data should be within the normal range, so the coagulation indicator data of the patient with normal coagulation process should be relatively concentrated, and there are many factors that cause the patient's coagulation abnormality and different ways of influence, so the coagulation indicator data of the patient with abnormal coagulation is relatively large. Therefore, by clustering the screened coagulation indicator data in the target data, the patient data set with normal coagulation process is obtained, and the normal range of the coagulation indicator for the current patient is obtained accordingly.

[0050] First, according to the thromboelastogram of the patient in the target data, the R time, K time, a angle, MA value and Ly30 data on the curve are recorded. By analyzing the correlation of these data and the coagulation indicators, the reaction degree of each coagulation indicator to the patient's coagulation process is calculated, and the specific process is as follows: (1) Draw the thromboelastogram parameters and coagulation indicator coordinate graph. Here, take the R time in the thromboelastogram and the INR in the coagulation indicator as an example: a) Normalize the R time data and INR data. Because the INR data itself is a standardized ratio data, it does not need to be processed here. Only the R time data needs to be normalized: Wherein, represents the value of the i-th R time data after normalization, represents the original data value of R time, represents the maximum value of R time, represents the minimum value of R time.

[0051] b) Take INR as the horizontal coordinate and R time as the vertical coordinate to draw the coordinate graph, and put the values corresponding to all target data into the coordinate.

[0052] c) Uniform the relationship between the horizontal coordinate and the vertical coordinate value in the coordinate graph. Because there can be some differences between different patient data, one horizontal coordinate can also correspond to multiple vertical coordinate values. In order to avoid the influence of this situation on subsequent analysis, it is necessary to uniform the relationship between the horizontal coordinate and the vertical coordinate value in the coordinate graph. The horizontal coordinate is evenly divided into 100 segments, and the center point coordinate value of each segment is taken as the horizontal coordinate value. The average value of all the vertical coordinate data corresponding to the horizontal coordinate segment is taken as the vertical coordinate value corresponding to the center point. In this way, the horizontal coordinate data and the vertical coordinate data are one-to-one corresponding, and the "INR-R time" coordinate graph is obtained.

[0053] (2) The data in the coordinate graph are fitted by using the least square method to obtain a fitting curve.

[0054] (3) According to the slope of the fitting curve, the influence degree of the INR coagulation index on the R time is calculated. First, if the INR coagulation index has a strong correlation with the R time, the slope value of each point on the obtained fitting curve should be consistent. At the same time, the greater the absolute value of the overall slope of the fitting curve, the greater the influence degree of the change of the INR value on the R time, that is, a small change in the INR value can cause a significant change in the R time. Therefore, 20 points (empirical value) are uniformly selected on the fitting curve, and according to the slope characteristics, the influence degree of the INR coagulation index on the R time is calculated: wherein, represents the influence degree of the INR coagulation index on the R time, represents the number of selected points on the fitting curve, and here , represents the slope value of the oth point on the "INR-R time" fitting curve, represents the absolute value of the slope difference between two adjacent points on the fitting curve, represents the smaller the absolute value of the slope difference, the more consistent the slope at different positions on the curve, and the higher the correlation degree between the INR coagulation index and the R time. The average slope of different points on the fitting curve represents the overall trend of the fitting curve. The greater the absolute value of the average slope, the greater the influence degree of the INR on the R time.

[0055] (4) Repeat the process of steps (1) to (3) to calculate the influence degree of the INR coagulation index on the K time, the a angle, the MA value and the Ly30 value, respectively , , , .

[0056] (5) Calculate the degree of INR reflecting the patient's coagulation process. If INR has a significant impact on multiple parameters in TEG, the degree of INR reflecting the patient's coagulation process is stronger, and the degree of INR reflecting the patient's coagulation process is calculated: wherein, represents the degree of INR reflecting the patient's coagulation process, represents the degree of INR coagulation index affecting R time, represents the degree of INR coagulation index affecting K time, represents the degree of INR coagulation index affecting the angle, represents the degree of INR coagulation index affecting MA value, represents the degree of INR coagulation index affecting Ly30 value. The average of the degrees of influence of all parameters in TEG is taken as the degree of INR reflecting the patient's coagulation process.

[0057] (6) According to the above method, the degree of reflection of all coagulation indexes in standard blood coagulation detection on the patient's coagulation process is obtained respectively. According to the corresponding degree of reflection from high to low, the three coagulation indexes with the highest degree of reflection are considered as the coagulation indexes with obvious influence on the patient's coagulation process.

[0058] After screening by the above steps, the three coagulation indexes obtained are the three indexes most closely related to the patient's coagulation process, and they are regarded as target coagulation indexes, so the data of three coagulation indexes of normal patients should be within the normal range. Normal data are concentrated in the same range, while abnormal data usually differ greatly due to different influencing factors leading to abnormality. Therefore, the three target coagulation indexes are respectively taken as x-axis, y-axis and z-axis, and a coordinate system is drawn, and all target historical patient data are put into the coordinate system. The data in the coordinate system are clustered by using DBSCAN clustering method, and multiple clusters and single point data are obtained, and single point noise data are excluded. The cluster data with the highest density among all clusters are obtained as normal patient data.

[0059] According to the maximum and minimum values of the corresponding target coagulation indexes in the coagulation normal patient data, the normal range of the corresponding target coagulation index of the patient is determined.

[0060] S3, according to the coagulation index normal range of the current patient, the coagulation abnormality degree of the current patient is calculated.

[0061] The coagulation index data of the current patient is obtained by standard coagulation detection. If the target coagulation index data of the current patient is within the normal range obtained in the above step, the coagulation index abnormality degree is Otherwise, the abnormality degree of the current patient is calculated according to the deviation of the current patient data from the normal range: wherein, represents the abnormality degree of the u-th target coagulation index of the current patient, represents the corresponding value of the u-th target coagulation index of the current patient, represents the minimum value of the normal range of the u-th target coagulation index, represents the maximum value of the normal range of the u-th target coagulation index. represents the minimum function.

[0062] The abnormality degree of the current patient is calculated by integrating the abnormality degrees of the patient under each target coagulation index: wherein, represents the coagulation abnormality degree of the current patient, represents the number of target coagulation indexes, here , represents the abnormality degree of the u-th target coagulation index of the current patient.

[0063] The above is only a preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A system for identifying abnormalities in coagulation process based on big data analytics, characterized in that, The method comprises the following steps: A first acquisition module obtains coagulation index data of a current patient through standard coagulation detection and obtains a thromboelastogram of the current patient by using a TEG device; A second acquisition module obtains the similarity of the age, disease background and drug use of a patient in historical data to the characteristics of the current patient, and then screens out target data reflecting the coagulation condition of the current patient; The similarity of each patient in the historical data to the current patient is obtained based on the similarity of the age, disease and drug use of the patient in the historical data to the current patient; It is judged whether the obtained similarity value is greater than the set threshold value. If it is greater, it is considered that the patient corresponding to the historical data has a high similarity to the current patient, and the coagulation process of the patient corresponding to the historical data can reflect the abnormal condition of the coagulation process of the current patient. All target data reflecting the coagulation condition of the current patient are screened out; A selection module obtains the reflection degree value of each coagulation index data on the normal state of the coagulation process of the current patient according to the correlation of each coagulation index data in the target data to the change of the thromboelastogram, and sorts all the reflection degree data. Select a number of coagulation indexes with the highest reflection degree as coagulation indexes that have a significant impact on the coagulation process of the patient; The correlation of each coagulation index data in the target data to the change of the thromboelastogram is as follows: Step 1: Select any coagulation index data in the target data and any parameter in the thromboelastogram, and normalize the selected coagulation index data and parameter respectively; Step 2: Take the normalized coagulation index data as the abscissa and the parameter as the ordinate, and put all the values corresponding to the target data into the coordinates as data; Step 3: Divide the abscissa into several segments, take the center point coordinate value of each segment as the abscissa value, take the average value of all ordinate data in the segment as the ordinate value corresponding to the center point, and make the abscissa data and the ordinate data correspond to each other to obtain a coagulation index data-parameter coordinate graph; Step 4: Use the least square method to fit the data in the coordinate graph to obtain a fitting curve; Step 5: Obtain the influence degree of the coagulation index data on the parameter according to the slope of the fitting curve; Step 6: Repeat steps 1 to 5 to obtain the influence degree of the coagulation index data on other parameters in the thromboelastogram; Step 7: Based on the influence degree of the coagulation index data on all parameters in the thromboelastogram, obtain the reflection degree value of the coagulation index data on the coagulation process of the patient; Step 8: Repeat steps 1 to 7 to obtain the reflection degree value of all coagulation index data in the target data on the coagulation process of the patient; A clustering module takes the several coagulation indexes with the highest reflection degree obtained as target coagulation indexes, draws a coordinate system according to the target coagulation indexes, puts all the target historical patient data into the coordinate system, clusters the data in the coordinate system, takes the cluster data with the highest density as normal patient data, and then determines the normal range of the target coagulation index. A judgment module judges whether the coagulation index data of the current patient is in the target coagulation index normal range. If not, the deviation degree of the coagulation index data of the current patient from the target coagulation index normal range is obtained according to the maximum and minimum values of the coagulation index data of the current patient and the target coagulation index normal range, and then the coagulation abnormality degree of the current patient is obtained.

2. The system for identifying abnormalities in coagulation process based on big data analytics as claimed in claim 1 wherein, The parameters contained in the thrombelastogram are R time, K time, alpha angle, MA value and Ly30; wherein, the R time is the time from liquid blood to fibrin formation; the K time is the time from coagulation to a certain intensity; the alpha angle is the fibrin production rate; the MA value is the final strength of the blood clot; the Ly30 is the thrombus degradation rate within 30 minutes; the coagulation index data includes prothrombin time PT, international normalized ratio INR, activated partial thromboplastin time APTT, thrombin time TT, fibrinogen FIB, D-dimer and platelet count PLT.

3. The system for identifying abnormalities in coagulation process based on big data analytics as claimed in claim 2 wherein, The similarity of the age, disease background and drug use of the patient in the historical data to the characteristics of the current patient is obtained, specifically as follows: The patient age similarity degree is obtained according to the difference between the patient age in the historical data and the current patient age, specifically as follows: ； wherein, represents the age similarity of the i-th historical data patient to the current patient, represents the age of the i-th patient in the historical data, represents the age of the current patient, represents the absolute value of the age difference of the i-th patient in the historical data to the current patient; it is explained that the smaller the absolute value of the age difference, the higher the age similarity; guarantees that the fraction is meaningful and ; The patient disease background similarity degree is obtained according to the disease coincidence degree between the patient in the historical data and the current patient. If the patient in the historical data has the same disease as the current patient, the similarity degree is high; if the patient in the historical data has the corresponding disease but the current patient does not, the similarity degree is reduced; if the patient in the historical data and the current patient have or do not have a certain disease, the count is counted, otherwise it is not counted, thereby calculating the disease background similarity between the patient in the historical data and the current patient, specifically as follows: ； wherein, represents the degree of similarity of the i-th historical data patient to the current patient's disease, represents the total number of diseases statistically affecting the coagulation index, represents the count value of the i-th historical data patient corresponding to the d-th disease; The similarity of the patient's drug use to the disease is calculated in the same way. If the patient in the historical data and the current patient take or do not take a certain drug at the same time, the count is 1, and if the patient in the historical data takes a certain drug but the current patient does not, the count is 0, thereby calculating the similarity of the drug use of the patient in the historical data to the current patient, specifically as follows: ； wherein, represents the degree of similarity of the i-th historical data patient to the current patient in drug use, represents the total number of drugs that statistically affect the coagulation index, represents the count value of the i-th historical data patient for the b-th drug.

4. The system for identifying abnormalities in coagulation process based on big data analytics as claimed in claim 3 wherein, The similarity degree value of the patient in each historical data to the current patient is obtained, specifically as follows: ； wherein, represents the degree of similarity of the i-th historical data to the current patient.

5. The big data analytics based coagulation process anomaly identification system as claimed in claim 4, wherein, The influence degree of the coagulation index data on the parameter is obtained according to the slope of the fitting curve, specifically as follows: Taking the coagulation index data INR in the target data and the parameter R time in the thrombelastogram as an example, ； wherein, represents the influence degree of the coagulation index data INR on R time, represents the number of selected points on the fitting curve, represents the slope value of the oth point on the "INR-R time" fitting curve, represents the absolute value of the slope difference of the adjacent two points on the fitting curve, The smaller the absolute value of the slope difference is, the more consistent the slopes at different positions on the curve are, and the higher the correlation degree of the INR coagulation index and the R time is; The average slope of different points on the fitting curve represents the overall trend of the fitting curve. The larger the absolute value of the average slope is, the greater the influence degree of INR on R time is. The reflection degree value of the coagulation index data on the patient's coagulation process is obtained based on the influence degree of the coagulation index data on all parameters in the thrombelastogram, specifically as follows: the influence degree values of the coagulation index data on all parameters in the thrombelastogram are added and averaged, and the obtained average value is taken as the reflection degree value of the coagulation index data on the patient's coagulation process.

6. The big data analytics based coagulation process anomaly identification system as claimed in claim 5, wherein, The target coagulation index normal range is determined by drawing a coordinate system according to the target coagulation index, putting all target historical patient data into the coordinate system, clustering the data of the coordinate system, and taking the cluster data with the highest density as the normal patient data, specifically as follows: The obtained target coagulation index is considered to be the most closely related to the coagulation process of the patient, so the target coagulation index of the patient with normal coagulation should be within the normal range; the normal data are concentrated in the same range, while the abnormal data of coagulation are usually different due to different influencing factors causing the abnormality, and the abnormality is also quite different; therefore, a coordinate system is drawn according to the target coagulation index, and all target historical patient data are put into the coordinate system; wherein the target historical patient data are data with high similarity to the current patient characteristics screened out; the data with high similarity refer to the data that can reflect the coagulation condition of the current patient, and the patient in the historical data is similar to the current patient in age, disease background and medication; the DBSCAN clustering method is used to cluster the data in the coordinate system, and then a plurality of clusters and single point data are obtained, the single point noise data is excluded, the cluster data with the highest density in all clusters is obtained as the normal patient data; and the maximum and minimum values of the corresponding target coagulation index in the coagulation normal patient data are used to determine the normal range of the target coagulation index corresponding to the patient.

7. The big data analytics based coagulation process anomaly identification system as claimed in claim 6, wherein, The deviation degree of the coagulation index data of the current patient from the normal range of the target coagulation index is obtained according to the coagulation index data of the current patient and the maximum and minimum values of the normal range of the target coagulation index, and specifically as follows: If the target coagulation index data of the current patient is in the normal range of the target coagulation index, the abnormality degree of the coagulation index is recorded as 0; otherwise, the abnormality degree of the current patient is calculated according to the deviation of the current patient data from the normal range. ; otherwise, the abnormality degree of the current patient is calculated according to the deviation of the current patient data from the normal range. ； wherein, represents the u-th target coagulation index abnormality degree of the current patient, represents the u-th target coagulation index corresponding value of the current patient, represents the u-th target coagulation index normal range minimum value, represents the u-th target coagulation index normal range maximum value; represents the minimum value function.

8. The big data analytics based coagulation process anomaly identification system as claimed in claim 7, wherein, The coagulation abnormality degree of the current patient is obtained, and specifically as follows: The coagulation abnormality degree of the current patient is obtained by comprehensively considering the abnormality degree of the patient under each target coagulation index, ； wherein, represents the abnormality degree of coagulation of the current patient, represents the number of target coagulation indexes, represents the abnormality degree of the u-th target coagulation index of the current patient.

Images