CN117954115A - Blood purification sample analysis method and blood filter - Google Patents
Blood purification sample analysis method and blood filter Download PDFInfo
- Publication number
- CN117954115A CN117954115A CN202410353443.2A CN202410353443A CN117954115A CN 117954115 A CN117954115 A CN 117954115A CN 202410353443 A CN202410353443 A CN 202410353443A CN 117954115 A CN117954115 A CN 117954115A
- Authority
- CN
- China
- Prior art keywords
- data
- blood
- minimum sample
- vector
- clustering
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 210000004369 blood Anatomy 0.000 title claims abstract description 136
- 239000008280 blood Substances 0.000 title claims abstract description 136
- 238000000746 purification Methods 0.000 title claims abstract description 35
- 238000012284 sample analysis method Methods 0.000 title claims abstract description 10
- 238000000034 method Methods 0.000 claims abstract description 67
- 239000013598 vector Substances 0.000 claims abstract description 58
- 230000008569 process Effects 0.000 claims abstract description 40
- 230000036541 health Effects 0.000 claims abstract description 35
- 230000008859 change Effects 0.000 claims abstract description 34
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000012804 iterative process Methods 0.000 claims description 17
- 230000000694 effects Effects 0.000 claims description 12
- 238000010606 normalization Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 230000000739 chaotic effect Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000012935 Averaging Methods 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 abstract description 10
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000012544 monitoring process Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 210000003743 erythrocyte Anatomy 0.000 description 2
- 208000021959 Abnormal metabolism Diseases 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 210000000265 leukocyte Anatomy 0.000 description 1
- 235000012054 meals Nutrition 0.000 description 1
- 230000006371 metabolic abnormality Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/40—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Epidemiology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention relates to the technical field of blood data processing, in particular to a blood purification sample analysis method and a blood filter. The method acquires multidimensional blood data, constructs a parameter space aiming at the blood data in each dimension, determines a reference clustering center and executes an ISODATA algorithm. In each iteration process in the ISODATA algorithm, the uniformity of health trends among blood data in different dimensions is considered, the overall health trend is obtained, and the minimum sample number change rate is further obtained by combining the confusion degree of data vectors in the corresponding clusters of the reference cluster center points. And adjusting the minimum sample point number in each iteration process based on the minimum sample number change rate until the iteration is finished to obtain an accurate blood data clustering result. According to the invention, through a dynamic minimum sample point number adjustment process, an accurate blood data clustering result is obtained, and analysis of a blood purification sample is facilitated.
Description
Technical Field
The invention relates to the technical field of blood data processing, in particular to a blood purification sample analysis method and a blood filter.
Background
Blood purification is a common medical procedure by which harmful substances in blood are removed by certain medical means. In the process, a blood filter is required to be used for processing a blood purification sample, so that blood data in various dimensions such as red blood cell content, white blood cell content, platelet content and the like are detected, and the blood purification method is beneficial to assisting medical staff in further knowing the condition of a patient and adjusting a purification scheme.
In order to assist medical staff in understanding and judging the condition of a patient, the acquired multidimensional blood data can be clustered in the prior art, and the purification effect is determined according to the change of the blood data of the patient in the clustering result analysis process. In the prior art, an ISODATA algorithm is adopted to cluster blood data on different time sequences, but because the clustering algorithm is an iterative processing algorithm, the minimum sample number in each iterative process in the traditional ISODATA algorithm is fixed, and the improper selection of the minimum sample number can lead to the fact that data belonging to the same class in a clustering result are divided into different class clusters or data not belonging to the same class in the clustering result are divided into the same class clusters, the real attribute of the class clusters in the blood data clustering result can not be determined, and further the analysis of blood purification samples is influenced.
Disclosure of Invention
In order to solve the technical problems that in the prior art, the minimum sample number is set improperly, so that the clustering effect is poor and the analysis of blood purification samples is affected, the invention aims to provide a blood purification sample analysis method and a blood filter, and the adopted technical scheme is as follows:
The invention provides a blood purification sample analysis method, which comprises the following steps:
acquiring multidimensional blood data in a blood purification sample within a preset time period;
mapping the corresponding blood data in each dimension into a value parameter space to obtain data points in the corresponding dimension at each moment; determining a reference clustering center according to overall data characteristics in a parameter space under each dimension, performing iterative clustering of an ISODATA algorithm according to the reference clustering center and the preset minimum sample point number, updating the reference clustering center in each iterative process, and obtaining a data vector of each data point relative to the reference clustering center;
In each iteration process, according to the similarity between the data vectors in the corresponding clusters of the reference cluster center and the standard blood data vectors in the corresponding parameter space under all dimensions, obtaining the overall health trend characteristics under the corresponding time period; obtaining the minimum sample number change rate according to the integral health trend characteristics and the chaotic degree of the data vector in the corresponding cluster of the reference cluster center; adjusting the minimum sample point number according to the minimum sample number change rate; and (5) until the iteration is finished, obtaining a blood data clustering result.
Further, the abscissa of the parameter space represents time information, and the ordinate represents blood data in a corresponding dimension.
Further, the method for acquiring the reference clustering center comprises the following steps:
The average value of the ordinate of all data points and the polar difference of the ordinate are counted in the parameter space, half of the polar difference of the ordinate is taken as an ordinate intermediate value, and the average value of the ordinate intermediate value are taken as the ordinate of the reference clustering center;
normalizing the ordinate of each data point to obtain the coordinate weight of the corresponding abscissa, and carrying out weighted averaging on the abscissa of all the data points according to the corresponding coordinate weight to obtain the abscissa of the reference cluster center.
Further, the obtaining a data vector for each data point relative to the reference cluster center comprises:
And taking the vector of which the reference clustering center points to each data point as the data vector of the corresponding data point.
Further, the method for acquiring the overall health trend characteristics comprises the following steps:
Obtaining standard blood data in each dimension under a corresponding time period, wherein the standard blood data comprises a group of standard data points in the parameter space, and standard data vectors of the standard data points relative to a clustering center are obtained; and taking the average vector of the standard data vectors as a first vector, taking the average vector of the data vectors in the corresponding cluster of the reference cluster center as a second vector, calculating cosine similarity of the first vector and the second vector, obtaining dimension health trend characteristics in the corresponding iteration process under the corresponding dimension, and taking the average value of the dimension health trend characteristics under all the dimensions as the integral health trend characteristics in the corresponding iteration process.
Further, the obtaining formula of the minimum sample number change rate includes:
; wherein/> For/>First/>, in the individual dimensionsMinimum sample number rate of change in the course of a second iteration,/>For presetting and adjusting weight,/>To at/>The modulus of the data vector of the data points in the cluster corresponding to the reference cluster center in the iterative process,For variance calculation function,/>For/>The overall health trend feature in the secondary iterative process,/>Is a normalization function.
Further, the adjusting the minimum number of sample points according to the minimum number of sample change rate includes:
taking the product of the minimum sample number of each iterative process and the corresponding minimum sample number change rate as the minimum sample point number of the next iterative process.
Further, the adjustment weight is set to 0.4.
Further, after obtaining the blood data clustering result, the method further comprises:
And obtaining blood data clustering results in different time periods, and evaluating the treatment effect according to the blood data clustering results in different time periods.
The invention also provides a blood filter, which comprises a filter body and a blood data processing module, wherein the blood data processing module executes the steps of the blood purification sample analysis method.
The invention has the following beneficial effects:
According to the embodiment of the invention, independent cluster analysis is respectively carried out on blood data in each dimension, and in consideration of the fact that the ISODATA algorithm process is a cluster iterative merging process, a plurality of clusters can be possibly generated in the iterative process, in order to facilitate data analysis, a reference cluster center is selected, and whether the minimum sample number needs to be changed can be judged by monitoring data information in the cluster corresponding to the reference cluster center in the subsequent process. If the human body is in a healthy state, blood data in different dimensions should be biased towards a healthy result together, because the reference cluster center is obtained according to the overall data characteristics, the corresponding class clusters also represent the overall blood data characteristics of the human body in the dimension, if the minimum sample number is reasonable, the class clusters corresponding to the reference cluster center in each dimension should be a healthy trend, so that the overall health trend characteristics are obtained based on the data in all dimensions, the minimum sample number change rate is obtained by further combining the confusion degree of the data vectors in the clusters, the minimum sample number in each iteration process is changed based on the minimum sample number change rate, and the accurate blood data clustering result is obtained until the iteration is finished.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for analyzing a blood purification sample according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following description refers to the specific implementation, structure, characteristics and effects of a blood purification sample analysis method and a blood filter according to the invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a method for analyzing a blood purification sample and a specific scheme of a blood filter provided by the present invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a method for analyzing a blood purification sample according to an embodiment of the present invention is shown, where the method includes:
step S1: and acquiring multidimensional blood data in the blood purification sample within a preset time period.
In the embodiment of the invention, the blood purification sample is separated and filtered by the blood filter, so that different types of blood samples are obtained, and multidimensional blood data such as erythrocyte content, leucocyte content, platelet content and the like can be obtained by further content detection technology in the prior art.
It should be noted that, because there are multiple treatment periods, that is, multiple time periods, in a blood purification process, a blood sample of a patient should be collected according to a certain time period in an analysis process of the blood purification sample, so as to obtain corresponding blood data for analysis, a preset time period in the embodiment of the invention can be regarded as a monitoring period, and after the blood data analysis result is obtained, a medical staff can be assisted to judge the blood purification effect according to the difference of analysis results among different monitoring periods. In one embodiment of the invention, the patient's daily pre-meal blood data is collected with one week as a time period, i.e., there are 21 sets of multidimensional blood data. In another embodiment of the present invention, in order to expand the number of data points, blood data of a patient is collected multiple times in a day, and the blood data at each time in the day is fitted by a least square method or other data fitting method, so as to further expand the data points, and specific technical means are known to those skilled in the art, and are not described and limited herein.
Step S2: mapping the corresponding blood data in each dimension into a value parameter space to obtain data points in the corresponding dimension at each moment; and determining a reference clustering center in a parameter space under each dimension according to the overall data characteristics, performing iterative clustering of an ISODATA algorithm according to the reference clustering center and the preset minimum sample point number, and updating the reference clustering center in each iterative process to obtain a data vector of each data point relative to the reference clustering center.
The collected blood data is a group of time sequence data, and each collection time corresponds to one dimension of blood data, so that in order to realize iterative clustering of the ISODATA algorithm, the collected blood data needs to be mapped into a parameter space first, and then corresponding data points are obtained.
Preferably, in one embodiment of the invention, the abscissa of the parameter space represents time information and the ordinate represents blood data in the corresponding dimension. Wherein the parameter space is selectableParameter space, which transforms the time information of the abscissa to the range by parameter transformationTransforming the blood data on the ordinate to the range/>, by parametric transformationOther forms of parameter space may be selected in other embodiments of the present invention, and are not described and limited herein.
The ISODATA algorithm is an iterative merging algorithm of class clusters, and the expected class number, the initial cluster center number, the minimum sample number, the upper limit of the relative standard deviation of the distribution of each characteristic component in the class, the lower limit of the minimum distance between the two classes of centers, the maximum number of times of merging operation and the maximum allowed iteration number in each iteration are required to be preset before the algorithm starts. Wherein, the ISODATA algorithm is a cluster merging and splitting algorithm, so the initial cluster center number may not be the expected class number; the minimum sample number limits the minimum sample number in the class cluster, and if the minimum sample number is not satisfied, the class cluster cannot be used as a class; once the upper limit of the relative standard deviation of the distribution of each characteristic component in the class is reached in the class cluster, splitting can be started; the minimum distance lower limits between the two classes of centers are met. Therefore, a plurality of class clusters exist in each iteration process after the ISODATA algorithm is executed, two classes, namely a normal class and an abnormal class, are needed to be distinguished for blood data, the excessive class clusters can influence subsequent analysis, so that a reference cluster center is determined according to the overall data characteristics in a parameter space under each dimension, the reference cluster center can be updated along with updating of the class clusters in the whole iteration process, but because the initial stage of the reference cluster center is determined based on the overall data characteristics in the parameter space, for blood data of a person, the health states of different periods are different, the health states of the blood data of a person are approximate to an overall value in one monitoring period, the determined reference cluster center can be used as the initial cluster center in the ISODATA algorithm for iterative clustering, the reference cluster center can be updated in the iteration process, the class clusters in the whole parameter space can be combined and split, and the analysis can be carried out only for the class clusters where the reference cluster center is in the reference cluster in the subsequent data analysis process.
Preferably, in one embodiment of the present invention, the method for acquiring the reference cluster center includes:
The mean value of all data points on the ordinate and the polar difference on the ordinate are counted in the parameter space, and half of the polar difference on the ordinate is taken as the median value on the ordinate. The mean value of the ordinate and the mean value of the ordinate can jointly represent the central data characteristic of the data points in the parameter space under the ordinate data, so that the mean value of the ordinate and the mean value of the ordinate are taken as the ordinate of the reference clustering center. For one cluster, the size of the blood data value has a larger influence on the clustering result, and the type of the cluster can be determined according to the size of the blood data value, so that in order to enable the reference cluster center to be more approximate to the region with a larger ordinate data value, the ordinate of each data point is normalized to obtain the coordinate weight of the corresponding abscissa, and the abscissas of all data points are weighted and averaged according to the corresponding coordinate weights to obtain the abscissas of the reference cluster center. In one embodiment of the invention, the coordinate information of the reference cluster center is formulated as:
Wherein, For reference to the abscissa of the cluster center,/>For reference to the ordinate of the cluster center,/>Is the maximum ordinate value of the data point in the parameter space,/>Is the minimum ordinate value of the data point in the parameter space,/>For the/>, in parameter spaceOrdinate value of data point,/>Number of data points,/>For the/>, in parameter spaceAbscissa values of data points.
To facilitate the quantification of each data point, a data vector for each data point relative to a reference cluster center may be obtained, with the vector of reference cluster centers pointing to each data point being the data vector for the corresponding data point in one embodiment of the invention.
It should be noted that, in the embodiment of the present invention, data points in different dimensions participate in clustering iteration respectively and simultaneously, each dimension in each iteration process described in a subsequent process has an iteration result, if data in a certain dimension reaches an iteration stop condition and other dimensions continue to iterate, the last iteration result of the dimension stopping iteration in the subsequent process, that is, the final clustering result participates in calculation in the iteration process of other dimensions.
Step S3: in each iteration process, according to the similarity between the data vectors in the corresponding clusters of the reference cluster center and the standard blood data vectors in the corresponding parameter space in all dimensions, obtaining the overall health trend characteristics in the corresponding time period; obtaining the minimum sample number change rate according to the integral health trend characteristics and the chaotic degree of the data vector in the corresponding cluster of the reference cluster center; and adjusting the minimum sample point number according to the minimum sample number change rate until the iteration is finished, and obtaining a blood data clustering result.
In order to ensure that data points belonging to different classes in a clustering result can be obviously distinguished, the embodiment of the invention dynamically adjusts the minimum sample number in each iteration process, so that the result of each iteration has the capability of separating abnormal blood data. To be able to do this, it is necessary to analyze in combination with the data characteristics of the blood data. For a patient, if they are in a healthy state, blood data in different dimensions should be commonly biased towards a healthy outcome. That is, although there are a plurality of data in different dimensions, the blood data belong to one patient, so that in each iteration process, the overall health trend characteristic in the corresponding time period needs to be obtained according to the similarity between the data vector in the corresponding cluster of the reference cluster center in the corresponding parameter space in all dimensions and the standard blood data vector. If the overall health trend feature is larger, the change relation between the data in different dimensions is developed towards the health proportion, and the corresponding minimum sample number is closer to the trend, which means that the minimum sample number is more reasonable.
Preferably, in one embodiment of the present invention, the method for acquiring the overall health trend feature includes:
Standard blood data in each dimension under a corresponding time period is obtained, wherein a group of standard data points exist in a parameter space in the standard blood data, and a standard data vector of the standard data points relative to a clustering center is obtained. And taking the average vector of the standard data vectors as a first vector, taking the average vector of the data vectors in the corresponding cluster of the reference cluster center as a second vector, and calculating the cosine similarity of the first vector and the second vector to obtain the dimension health trend characteristics in the corresponding iteration process under the corresponding dimension, wherein the larger the cosine similarity is, the closer the clustering result of the corresponding cluster of the reference cluster center under the dimension is to the standard health index trend. And taking the average value of the dimension health trend characteristics in all the dimensions as the integral health trend characteristic of the corresponding iterative process. It should be noted that, the standard blood data may be directly determined through a priori knowledge such as medical papers, and the present invention is not limited thereto.
Further obtaining the confusion degree of the data vector in the corresponding cluster of the reference cluster center, wherein the greater the confusion degree is, the more unreasonable the minimum sample number exists in the current cluster, so that the minimum sample number change rate can be obtained by combining the confusion degree and the overall health trend characteristics, the minimum sample number change rate can be used for evaluating whether the minimum sample number in the current iteration process is reasonable or not, and the minimum sample number can be adjusted according to the minimum sample number change rate.
Preferably, the obtaining formula of the minimum sample number change rate in one embodiment of the present invention includes:
; wherein/> For/>First/>, in the individual dimensionsMinimum sample number rate of change in the course of a second iteration,/>For presetting and adjusting weight,/>To at/>In the process of iteration, referring to the modulus of the data vector of the data points in the cluster corresponding to the cluster center,For variance calculation function,/>For/>Integral health trend characteristics in the iterative process,/>Is a normalization function. In the embodiment of the present invention, the normalization functions are all sigmoid functions, and in other embodiments of the present invention, normalization methods such as maximum and minimum normalization may be used, and the specific implementation process is a technical means well known to those skilled in the art, and will not be described herein.
In the formula of the minimum sample number change rate,The degree of confusion of the data vector is represented, the larger the degree of confusion is, the more unsuitable the current minimum sample number is, and the condition that the data point category in the cluster is disordered possibly exists due to the fact that the condition is too large, so that the minimum sample number needs to be adjusted to be small in the subsequent adjustment process, the larger the degree of confusion is after inverse negative correlation processing in the formula, the smaller the change rate of the minimum sample number is, and the smaller the minimum sample number needs to be adjusted in the subsequent adjustment process. The larger the overall health trend feature is, the closer the clustering result under the current minimum sample number is to the health distribution trend, namely, the cluster only contains data points of normal categories, the more suitable the current minimum sample number is, the more suitable the change rate of the minimum sample number is, and the less the minimum sample number is required to be adjusted in the subsequent adjustment process.
In one embodiment of the invention, the adjustment weight is set to 0.4.
Preferably, adjusting the minimum sample point number according to the minimum sample number change rate in one embodiment of the present invention includes:
taking the product of the minimum sample number of each iterative process and the corresponding minimum sample number change rate as the minimum sample point number of the next iterative process. That is, the larger the minimum sample number change rate is, the less the current minimum sample number needs to be changed, and the smaller the minimum sample number change rate is, the further the current minimum sample number needs to be reduced.
And a dynamic minimum sample number adjusting process exists in each iteration process until the ISODATA algorithm iteration is finished, so that a blood data clustering result can be obtained. It should be noted that, the condition of ending the iteration includes satisfying two cases of iteration times and iteration convergence, and the specific isadata algorithm is a technical means well known to those skilled in the art, and will not be described herein.
Preferably, after obtaining the blood data clustering result in one embodiment of the present invention, the method further includes:
And obtaining blood data clustering results in different time periods, and evaluating the treatment effect according to the blood data clustering results in different time periods. In one embodiment of the present invention, two types of data exist in the clustering result, one type is normal blood data, and the other type is abnormal blood data, wherein the abnormal blood data is generated in relation to the blood lesion or abnormal metabolism of the patient, because the time period and the data acquisition frequency in the embodiment of the present invention are the same, that is, the number of data points acquired in one time period is the same, the treatment effect can be measured by comparing the number difference of the normal blood data type or the abnormal blood data type in different time periods, that is, if the number of data points in the normal blood data type is in an increasing trend, the current treatment effect is better, and the greater the increasing trend degree indicates the greater the treatment effect degree; if the number of data points in the abnormal blood data category is in an increasing trend or unchanged, the current treatment effect is poor, further deterioration of the blood condition of the patient may occur, and the medical staff can adjust the treatment scheme in time.
In summary, the embodiment of the present invention acquires multidimensional blood data, constructs a parameter space for the blood data in each dimension, determines a reference cluster center, and executes the ISODATA algorithm. In each iteration process in the ISODATA algorithm, the uniformity of health trends among blood data in different dimensions is considered, the overall health trend is obtained, and the minimum sample number change rate is further obtained by combining the confusion degree of data vectors in the corresponding clusters of the reference cluster center points. And adjusting the minimum sample point number in each iteration process based on the minimum sample number change rate until the iteration is finished to obtain an accurate blood data clustering result. According to the invention, through a dynamic minimum sample point number adjustment process, an accurate blood data clustering result is obtained, and analysis of a blood purification sample is facilitated.
The invention also provides a blood filter, which comprises a filter body and a blood data processing module, wherein the filter body is used for separating each component in a blood purification sample and acquiring the content of each component, and the blood data processing module executes the steps of any blood purification sample analysis method.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (10)
1. A method of analyzing a blood purification sample, the method comprising:
acquiring multidimensional blood data in a blood purification sample within a preset time period;
mapping the corresponding blood data in each dimension into a value parameter space to obtain data points in the corresponding dimension at each moment; determining a reference clustering center according to overall data characteristics in a parameter space under each dimension, performing iterative clustering of an ISODATA algorithm according to the reference clustering center and the preset minimum sample point number, updating the reference clustering center in each iterative process, and obtaining a data vector of each data point relative to the reference clustering center;
In each iteration process, according to the similarity between the data vectors in the corresponding clusters of the reference cluster center and the standard blood data vectors in the corresponding parameter space under all dimensions, obtaining the overall health trend characteristics under the corresponding time period; obtaining the minimum sample number change rate according to the integral health trend characteristics and the chaotic degree of the data vector in the corresponding cluster of the reference cluster center; adjusting the minimum sample point number according to the minimum sample number change rate; and (5) until the iteration is finished, obtaining a blood data clustering result.
2. A method of analyzing a blood purification sample according to claim 1, wherein the abscissa of the parameter space represents time information and the ordinate represents blood data in the corresponding dimension.
3. The method for analyzing a blood purification sample according to claim 2, wherein the method for acquiring the reference cluster center comprises:
The average value of the ordinate of all data points and the polar difference of the ordinate are counted in the parameter space, half of the polar difference of the ordinate is taken as an ordinate intermediate value, and the average value of the ordinate intermediate value are taken as the ordinate of the reference clustering center;
normalizing the ordinate of each data point to obtain the coordinate weight of the corresponding abscissa, and carrying out weighted averaging on the abscissa of all the data points according to the corresponding coordinate weight to obtain the abscissa of the reference cluster center.
4. The method of claim 1, wherein said obtaining a data vector for each data point relative to the reference cluster center comprises:
And taking the vector of which the reference clustering center points to each data point as the data vector of the corresponding data point.
5. The method for analyzing a blood purification sample according to claim 1, wherein the method for acquiring the overall health trend feature comprises:
Obtaining standard blood data in each dimension under a corresponding time period, wherein the standard blood data comprises a group of standard data points in the parameter space, and standard data vectors of the standard data points relative to a clustering center are obtained; and taking the average vector of the standard data vectors as a first vector, taking the average vector of the data vectors in the corresponding cluster of the reference cluster center as a second vector, calculating cosine similarity of the first vector and the second vector, obtaining dimension health trend characteristics in the corresponding iteration process under the corresponding dimension, and taking the average value of the dimension health trend characteristics under all the dimensions as the integral health trend characteristics in the corresponding iteration process.
6. The method according to claim 1, wherein the obtaining formula of the minimum sample number change rate includes:
; wherein/> For/>First/>, in the individual dimensionsMinimum sample number rate of change in the course of a second iteration,/>For presetting and adjusting weight,/>To at/>The modulus of the data vector of the data points in the cluster corresponding to the reference cluster center in the iterative process,For variance calculation function,/>For/>The overall health trend feature in the secondary iterative process,/>Is a normalization function.
7. The method of claim 6, wherein said adjusting said minimum sample point number according to a minimum sample number change rate comprises:
taking the product of the minimum sample number of each iterative process and the corresponding minimum sample number change rate as the minimum sample point number of the next iterative process.
8. The method according to claim 6, wherein the adjustment weight is set to 0.4.
9. The method for analyzing a blood purification sample according to claim 1, further comprising, after obtaining the blood data clustering result:
And obtaining blood data clustering results in different time periods, and evaluating the treatment effect according to the blood data clustering results in different time periods.
10. A blood filter comprising a filter body and a blood data processing module that performs the steps of a blood purification sample analysis method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410353443.2A CN117954115B (en) | 2024-03-27 | 2024-03-27 | Blood purification sample analysis method and blood filter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410353443.2A CN117954115B (en) | 2024-03-27 | 2024-03-27 | Blood purification sample analysis method and blood filter |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117954115A true CN117954115A (en) | 2024-04-30 |
| CN117954115B CN117954115B (en) | 2024-06-04 |
Family
ID=90793066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410353443.2A Active CN117954115B (en) | 2024-03-27 | 2024-03-27 | Blood purification sample analysis method and blood filter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117954115B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017190542A1 (en) * | 2016-05-04 | 2017-11-09 | 山东大学 | Partition-based vca end-member extraction method |
| CN112101374A (en) * | 2020-08-01 | 2020-12-18 | 西南交通大学 | Unmanned aerial vehicle obstacle detection method based on SURF feature detection and ISODATA clustering algorithm |
| CN117454671A (en) * | 2023-12-22 | 2024-01-26 | 广东力宏微电子有限公司 | Artificial intelligence-based field effect transistor life assessment method |
| CN117473351A (en) * | 2023-12-26 | 2024-01-30 | 深圳市胜威南方科技有限公司 | Power supply information remote transmission system based on Internet of things |
| CN117493921A (en) * | 2024-01-03 | 2024-02-02 | 智洁云服(大连)信息技术有限公司 | Artificial intelligence energy-saving management method and system based on big data |
| CN117556279A (en) * | 2024-01-12 | 2024-02-13 | 江苏雷博微电子设备有限公司 | Method and system for monitoring running state of spin coater based on electrical parameter analysis |
-
2024
- 2024-03-27 CN CN202410353443.2A patent/CN117954115B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017190542A1 (en) * | 2016-05-04 | 2017-11-09 | 山东大学 | Partition-based vca end-member extraction method |
| CN112101374A (en) * | 2020-08-01 | 2020-12-18 | 西南交通大学 | Unmanned aerial vehicle obstacle detection method based on SURF feature detection and ISODATA clustering algorithm |
| CN117454671A (en) * | 2023-12-22 | 2024-01-26 | 广东力宏微电子有限公司 | Artificial intelligence-based field effect transistor life assessment method |
| CN117473351A (en) * | 2023-12-26 | 2024-01-30 | 深圳市胜威南方科技有限公司 | Power supply information remote transmission system based on Internet of things |
| CN117493921A (en) * | 2024-01-03 | 2024-02-02 | 智洁云服(大连)信息技术有限公司 | Artificial intelligence energy-saving management method and system based on big data |
| CN117556279A (en) * | 2024-01-12 | 2024-02-13 | 江苏雷博微电子设备有限公司 | Method and system for monitoring running state of spin coater based on electrical parameter analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117954115B (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109952614B (en) | Biological particle classification system and method | |
| CN117609813B (en) | Intelligent management method for intensive patient monitoring data | |
| CN110364259B (en) | Method, system, medium and electronic device for predicting altitude disease | |
| JPH07105166A (en) | Belonging degree decision device | |
| CN111739634A (en) | Method, device and equipment for intelligently grouping similar patients and storage medium | |
| CN117216131B (en) | Traditional Chinese medicine data management system based on data retrieval | |
| CN107545133A (en) | A kind of Gaussian Blur cluster calculation method for antidiastole chronic bronchitis | |
| CN117912712B (en) | Thyroid disease data intelligent management method and system based on big data | |
| CN113889267A (en) | Method for constructing diabetes diagnosis model based on eye image recognition and electronic equipment | |
| CN113749619A (en) | Mental fatigue assessment method based on K-TRCA | |
| CN117954115B (en) | Blood purification sample analysis method and blood filter | |
| CN117958835B (en) | Electrocardiogram analysis method for postoperative rehabilitation of patient | |
| CN114707608A (en) | Medical quality control data processing method, apparatus, device, medium, and program product | |
| WO2019042097A1 (en) | Optimization method and device for system parameter design space | |
| Georgopoulos | Advanced time-frequency analysis and machine learning for pathological voice detection | |
| CN112233742B (en) | Medical record document classification system, equipment and storage medium based on clustering | |
| CN113270191A (en) | Data correction and classification method and storage medium | |
| CN110010246A (en) | A kind of disease Intelligent Diagnosis Technology based on neural network and confidence interval | |
| CN110583573A (en) | Construction and evaluation method of blood deficiency mouse model | |
| CN116933166A (en) | Cerebral apoplexy-oriented unbalanced data set classification method and system | |
| CN111637610B (en) | Indoor environment health degree adjusting method and system based on machine vision | |
| CN112733727B (en) | Electroencephalogram consciousness dynamic classification method based on linear analysis and feature decision fusion | |
| CN114862851A (en) | Processing method based on tongue picture analysis | |
| Jha et al. | Automated Acute Lymphoblastic Leukemia Detection Using Blood Smear Image Analysis | |
| Linder et al. | Computer‐Aided Diagnosis of Multiple Sclerosis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |