Anesthesia information intelligent management method and platform based on big data

Abstract

The invention discloses an anesthesia information intelligent management method and platform based on big data, and relates to the field of medical information intelligent management, and the method comprises the steps: obtaining anesthesia big data information, carrying out the anesthesia big data processing based on an improved K-MEANS combined DBSCAN algorithm, and carrying out the intelligent generation of an anesthesia strategy based on an improved random forest algorithm. According to the invention, intelligent management of anesthesia big data is realized, the anesthesia strategy is predicted based on the anesthesia big data information, subjective errors of anesthetists are reduced, and the accuracy and objectivity of the intelligently generated anesthesia strategy are improved. Moreover, the anesthesia big data is classified based on an improved clustering algorithm, and different decision trees of a random forest algorithm are trained based on different types of anesthesia data samples, so that more accurate intelligent selection of anesthesia strategies is realized.

Description

Technical Field

[0001] The present invention relates to the field of intelligent management of medical information, and in particular to an intelligent management method and platform for anesthesia information based on big data. Background Art

[0002] The amount of data generated in the medical field has surged with the development of medical technology, including electronic health records, medical imaging, genetic information, patient behavior data, etc. The analysis of these big data is of great significance for improving perioperative diagnosis, treatment and complication prevention, including the development of anesthesia.

[0003] Intelligent management of anesthesia big data helps to discover potential problems and risks in anesthesia practice, thereby providing anesthesiologists with more accurate and safe anesthesia plans. By analyzing and summarizing a large number of cases, we can find the regularity and differences in the anesthesia process, providing a scientific basis for individualized anesthesia.

[0004] The existing technology lacks a corresponding big data platform and professional management methods for the management of anesthesia information. The choice of anesthesia strategy is based on the doctor's experience and judgment, and intelligent management of anesthesia data has not been achieved. Summary of the invention

[0005] In order to solve the technical problem of intelligent management of anesthesia information in the prior art, the present invention provides an intelligent management method and platform for anesthesia information based on big data.

[0006] The present invention is achieved through the following technical solutions:

[0007] A method for intelligent management of anesthesia information based on big data, characterized by comprising:

[0008] S1: Acquire anesthesia big data information, wherein the anesthesia big data information includes patient personal information, physiological information, disease information, and intraoperative anesthesia information;

[0009] S2: Anesthesia big data processing, based on the improved K-MEANS combined with DBSCAN algorithm, the anesthesia data of special patients are screened out and deleted, and the remaining anesthesia data of general patients are classified to obtain several types of anesthesia populations;

[0010] S3: Intelligently generate anesthesia strategies based on an improved random forest algorithm. The random forest algorithm inputs patient personal information, physiological information, and disease information, and outputs intraoperative anesthesia information.

[0011] Furthermore, the patient's personal information includes the patient's gender, age, height, and weight; the physiological information includes blood pressure, blood routine test results, respiratory system conditions, liver function conditions, and kidney function conditions; the disease information includes the patient's disease type, illness duration, and previous surgical history; the intraoperative anesthesia information includes the type of surgery, surgery time, and anesthesia parameters.

[0012] Furthermore, the anesthesia parameters include anesthetic drugs, anesthetic doses, and anesthesia methods, and the anesthesia methods include local anesthesia, intravenous injection, inhalation, spinal anesthesia, nerve block anesthesia, and combined anesthesia.

[0013] Furthermore, the step S2 further includes:

[0014] S21: Screening and removing big anesthesia data of special patients based on DBSCAN clustering algorithm;

[0015] S22: Classifying the anesthesia big data information of general patients into several categories based on an improved dynamic K-Means clustering algorithm, wherein the improved dynamic K-Means clustering algorithm determines the cluster center based on a sample density weight method.

[0016] Furthermore, the step S22 further includes:

[0017] S221: Calculate the samples in the data set Density:

[0018]

[0019] Where n is the number of samples, is the average distance, is the Euclidean distance between two samples, and the function f(x) is expressed as follows:

[0020]

[0021]

[0022] in, is the Euclidean distance between two samples;

[0023] S222: Select the element with the largest density value as the first cluster center;

[0024] S223: Will satisfy Sample elements that are smaller than the preset condition are added to the current cluster, and these sample elements are deleted from the set;

[0025] S224: Calculate the remaining element weight product according to the following formula , find the maximum value and select the corresponding sample element as the next cluster center:

[0026]

[0027] in, is the distance between clusters, representing the sample elements in the data set With samples with higher local density The distance between The larger it is, the farther the distance between the two clusters is, and the greater the difference is; is the product density of the data set. The larger it is, the more sample points The more element points there are around, the more concentrated the element is; is the average distance of samples within the cluster, The smaller the cluster, the greater the density of elements. The expression is as follows:

[0028]

[0029] S225: Repeat steps S223-S224 until the set is an empty set.

[0030] Furthermore, the preset conditions of step S223 are dynamically selected. When the amount of sample data of a certain category of clustering is greater than a certain threshold, the preset conditions are adjusted to reduce the number of samples of the current cluster center and dynamically increase the number of cluster centers.

[0031] Furthermore, the newly added cluster center is generated within a certain range around the current cluster center.

[0032] Furthermore, the improved random forest algorithm trains each decision tree based on different training sets; the different training sets are composed of H sub-training sets, where H is the number of anesthesia population classifications obtained in step S2;

[0033] If the number of sub-training sets is greater than the number of decision trees, K of the H sub-training sets are randomly selected to train K decision trees respectively. If the number of sub-training sets is less than the number of decision trees, 2~H-1 subsets are randomly selected to form a training set.

[0034] The present invention also provides an intelligent management platform for anesthesia information based on big data, and an intelligent management method for anesthesia information based on big data as described above, which includes:

[0035] An anesthesia big data information acquisition module, which is used to acquire anesthesia big data information, wherein the anesthesia big data information includes patient personal information, physiological information, disease information, and intraoperative anesthesia information;

[0036] Anesthesia big data processing module, which is used to process data based on the improved K-MEANS combined with DBSCAN algorithm, filter out and delete the anesthesia data of special patients and classify the anesthesia data to obtain several types of anesthesia populations; it includes filtering out the anesthesia big data information of special patients and classifying the anesthesia big data information of general patients based on the improved dynamic K-Means clustering algorithm;

[0037] An intelligent generation module for anesthesia strategies is used to intelligently generate anesthesia strategies based on an improved random forest algorithm, wherein the input of the random forest algorithm is patient personal information, physiological information, and disease information, and the output is intraoperative anesthesia information; wherein each decision tree is trained based on different training sets; and the different training sets are composed of a plurality of sub-training sets.

[0038] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, on which program instructions of the intelligent management method of anesthesia information based on big data are stored. The program instructions of the intelligent management method of anesthesia information based on big data can be executed by one or more processors to implement the steps of the intelligent management method of anesthesia information based on big data as described above.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The present invention realizes the intelligent management of anesthesia big data, predicts anesthesia strategies based on anesthesia big data information, reduces the subjective errors of anesthesiologists, and improves the accuracy and objectivity of intelligently generated anesthesia strategies. In addition, the present invention classifies anesthesia big data based on an improved clustering algorithm, and trains different decision trees of the random forest algorithm based on different categories of anesthesia data samples, thereby realizing a more accurate intelligent selection of anesthesia strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0042] Figure 1 It is a flowchart of an intelligent management method for anesthesia information based on big data according to an embodiment of the present application;

[0043] Figure 2 It is a schematic diagram of an improved random forest algorithm model according to an embodiment of the present application. DETAILED DESCRIPTION

[0044] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0045] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0046] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention. The illustrations only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the form, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0047] See also Figure 1 , an intelligent management method for anesthesia information based on big data, comprising the following steps:

[0048] S1: Obtain anesthesia big data information, wherein the anesthesia big data information includes patient personal information, physiological information, disease information and intraoperative anesthesia information; wherein the patient personal information includes the patient's gender, age, height, weight, etc.; physiological information includes blood pressure, blood routine test results, respiratory system status, liver function status, kidney function status; disease information includes the patient's disease type, illness time, previous surgical history, etc.; intraoperative anesthesia information includes surgery type, surgery time, anesthesia parameters; wherein the anesthesia parameters include anesthetic drugs, anesthetic doses, and anesthesia methods, and the anesthesia methods include local anesthesia, intravenous injection, inhalation, spinal anesthesia, nerve block anesthesia, and combined anesthesia.

[0049] S2: Anesthesia Big Data Processing

[0050] The present invention processes data based on the improved K-MEANS combined with DBSCAN algorithm, filters out and deletes the anesthesia data of special patients and classifies the remaining anesthesia data of general patients to obtain H types of anesthesia populations. Combining the two clustering algorithms can achieve effective data processing and classification. It includes the following steps:

[0051] S21: Screening and removing anesthesia big data information of special patients

[0052] For some special patients, due to their special diseases or special physiological conditions, the anesthesia strategies used for them are quite different from those for ordinary people. The anesthesia information of these patients will affect the intelligent selection of the overall anesthesia strategy. Therefore, the anesthesia data information of these special patients needs to be screened and deleted.

[0053] Based on the DBSCAN clustering algorithm, the acquired anesthesia big data information is classified into outliers, and the outlier special patient information is deleted to obtain the anesthesia big data information of general patients. The specific steps are as follows:

[0054] a. Input sample set C, neighborhood parameters , sample distance measurement method; among them, is the neighborhood radius, To set the threshold;

[0055] b. Select an object p that has not been visited from the sample set C if it satisfies , then go to step c, otherwise classify the object as outlier data and repeat step b;

[0056] in, For the sample object p in The total number of samples in the neighborhood;

[0057] c. Build a cluster based on object p, and add objects that meet density reachability or density connection to the cluster;

[0058] The density can be reached: if point q is at point p In the neighborhood, and ; then point q satisfies the density reachability with p;

[0059] If point At the point of In the neighborhood (t=1,2), and the middle point satisfy , then point Satisfaction with Density can reach;

[0060] in, For sample objects exist The total number of samples in the neighborhood;

[0061] If the point r satisfies , point p is reachable by density o, and point q is reachable by density o, then the densities p and q are connected.

[0062] in, For sample objects exist The total number of samples in the neighborhood;

[0063] d. Continue to repeat the above steps b and c until all sample objects in the sample set are visited.

[0064] e. After deleting the outlier data, we can obtain the anesthesia big data information of general patients.

[0065] S22: Classify the anesthesia big data information of general patients based on the improved dynamic K-Means clustering algorithm to obtain H categories;

[0066] In the prior art, the selection of cluster center objects p is usually done by random selection. This randomness will have a great impact on the clustering results. The present invention improves the prior art and improves the K-Means clustering algorithm based on a sample density weight method to reduce the instability of clustering results caused by randomness and improve the accuracy. The present invention specifically includes the following steps:

[0067] S221: Calculate the samples in the data set Density:

[0068]

[0069] Where n is the number of samples, is the average distance, is the Euclidean distance between two samples, and the function f(x) is expressed as follows:

[0070]

[0071]

[0072] in, is the Euclidean distance between two samples.

[0073] S222: Select the element with the largest density value as the first cluster center.

[0074] S223: Will satisfy Sample elements that are smaller than the preset condition are added to the current cluster, and these sample elements are deleted from the set;

[0075] S224: Calculate the remaining element weight product according to the following formula , find the maximum value and select the corresponding sample element as the next cluster center:

[0076]

[0077] in, is the distance between clusters, representing the sample elements in the data set With samples with higher local density The distance between The larger it is, the farther the distance between the two clusters is, and the greater the difference is; is the product density of the data set. The larger it is, the more sample points The more element points there are around, the more concentrated the element is; is the average distance of samples within the cluster, The smaller the cluster, the greater the density of elements. The expression is as follows:

[0078]

[0079] S225: Repeat steps S223-S224 until the set is an empty set.

[0080] Among them, the preset conditions of step S223 are dynamically selected. In the prior art, in the clustering process, fixed preset conditions are usually selected. For the situation where the amount of data in different categories differs greatly, clustering cannot be achieved well. The present invention is based on dynamically selecting preset conditions. When the amount of cluster sample data in a certain category is greater than a certain threshold, the preset conditions are adjusted to reduce the number of samples in the current cluster center and dynamically increase the number of cluster centers so that cluster centers with more samples can be subdivided. Optionally, the newly added cluster centers are generated within a certain range around the current cluster center.

[0081] S3: Intelligent generation of anesthesia strategies

[0082] Based on Figure 2 The improved random forest algorithm shown in the figure generates anesthesia strategies intelligently. The random forest algorithm inputs patient personal information, physiological information, and disease information, and outputs intraoperative anesthesia information, including the following steps:

[0083] S31: Preprocessing the dataset

[0084] Based on the H classification results obtained in step S2, the data is divided into H data sets, a certain proportion of data is extracted from each data set as a training set, and H sub-training sets are generated;

[0085] S32: Model building

[0086] Determine the number and level of decision trees for the random forest;

[0087] S33: Decision Tree Training

[0088] The present invention improves the training method and trains each decision tree based on different training sets; the different training sets are composed of H sub-training sets;

[0089] Optionally, if the number of sub-training sets is greater than the number of decision trees, K of the H sub-training sets are randomly selected to train K decision trees respectively; if the number of sub-training sets is less than the number of decision trees, 2~H-1 subsets are randomly selected to combine to form a training set.

[0090] S34: Merge all the trained decision trees together to complete the construction of the random forest.

[0091] S35: Input the patient information into the trained model to intelligently generate anesthesia strategies and output anesthesia information during surgery, including anesthetic drugs, anesthetic dosages, and anesthesia methods.

[0092] In this embodiment, intelligent management of anesthesia information is achieved, and more objective and accurate intelligent generation of anesthesia strategies is achieved based on improved random forest and clustering algorithms.

[0093] The embodiment of the present invention further proposes an intelligent management platform for anesthesia information based on big data, based on the intelligent management method for anesthesia information based on big data as described above, comprising:

[0094] Anesthesia big data information acquisition module, which is used to acquire anesthesia big data information, which includes patient personal information, physiological information, disease information and intraoperative anesthesia information; wherein patient personal information includes patient gender, age, height, weight, etc.; physiological information includes blood pressure, blood routine test results, respiratory system status, liver function status, kidney function status; disease information includes patient disease type, illness time, previous surgical history, etc.; intraoperative anesthesia information includes surgery type, surgery time, anesthesia parameters;

[0095] Anesthesia big data processing module, which is used to process data based on the improved K-MEANS combined with DBSCAN algorithm, filter out and delete the anesthesia data of special patients and classify the anesthesia data to obtain several types of anesthesia populations; it includes filtering out the anesthesia big data information of special patients and classifying the anesthesia big data information of general patients based on the improved dynamic K-Means clustering algorithm;

[0096] An intelligent generation module for anesthesia strategies, which is used to intelligently generate anesthesia strategies based on an improved random forest algorithm, wherein the input of the random forest algorithm is patient personal information, physiological information, and disease information, and the output is intraoperative anesthesia information; wherein each decision tree is trained based on a different training set; and the different training sets are composed of a plurality of sub-training sets;

[0097] Optionally, if the number of sub-training sets is greater than the number of decision trees, K of the H sub-training sets are randomly selected to train K decision trees respectively; if the number of sub-training sets is less than the number of decision trees, 2~H-1 subsets are randomly selected to combine to form a training set.

[0098] In addition, an embodiment of the present invention also proposes a computer-readable storage medium, on which program instructions of an intelligent management method for anesthesia information based on big data are stored. The program instructions of the intelligent management method for anesthesia information based on big data can be executed by one or more processors to implement the steps of the intelligent management method for anesthesia information based on big data as described above.

[0099] The embodiments described above are only descriptions of the preferred implementation modes of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for intelligent management of anesthesia information based on big data, characterized in that: include: S1: Acquire anesthesia big data information, wherein the anesthesia big data information includes patient personal information, physiological information, disease information, and intraoperative anesthesia information; S2: Anesthesia big data processing, based on the improved K-MEANS combined with DBSCAN algorithm, the anesthesia data of special patients are screened out and deleted, and the remaining anesthesia data of general patients are classified to obtain several types of anesthesia populations; S3: Intelligently generate anesthesia strategies based on an improved random forest algorithm. The random forest algorithm inputs patient personal information, physiological information, and disease information, and outputs intraoperative anesthesia information.

2. The method for intelligent management of anesthesia information based on big data according to claim 1, characterized in that: The patient's personal information includes the patient's gender, age, height, and weight; the physiological information includes blood pressure, blood routine test results, respiratory system conditions, liver function conditions, and kidney function conditions; the disease information includes the patient's disease type, illness duration, and previous surgical history; the intraoperative anesthesia information includes the type of surgery, surgery time, and anesthesia parameters.

3. The method for intelligent management of anesthesia information based on big data according to claim 2, characterized in that: The anesthesia parameters include anesthetic drugs, anesthetic dosage, and anesthesia methods. The anesthesia methods include local anesthesia, intravenous injection, inhalation, spinal anesthesia, nerve block anesthesia, and combined anesthesia.

4. The method for intelligent management of anesthesia information based on big data according to claim 1, characterized in that: The step S2 further comprises: S21: Screening and removing big anesthesia data of special patients based on DBSCAN clustering algorithm; S22: Classifying the anesthesia big data information of general patients into several categories based on an improved dynamic K-Means clustering algorithm, wherein the improved dynamic K-Means clustering algorithm determines the cluster center based on a sample density weight method.

5. The method for intelligent management of anesthesia information based on big data according to claim 4, characterized in that: The step S22 further includes: S221: Calculate the samples in the data set Density: ； Where n is the number of samples, is the average distance, is the Euclidean distance between two samples, and the function f(x) is expressed as follows: ； ； in, is the Euclidean distance between two samples; S222: Select the element with the largest density value as the first cluster center; S223: Will satisfy Sample elements that are smaller than the preset condition are added to the current cluster, and these sample elements are deleted from the set; S224: Calculate the remaining element weight product according to the following formula , find the maximum value and select the corresponding sample element as the next cluster center: ； in, is the distance between clusters, representing the sample elements in the data set With samples with higher local density The distance between The larger it is, the farther the distance between the two clusters is, and the greater the difference is; is the product density of the data set. The larger it is, the more sample points The more element points there are around, the more concentrated the element is; is the average distance of samples within the cluster, The smaller the cluster, the greater the density of elements. The expression is as follows: ； S225: Repeat steps S223-S224 until the set is an empty set.

6. The method for intelligent management of anesthesia information based on big data according to claim 5, characterized in that: The preset conditions of step S223 are dynamically selected. When the amount of sample data of a certain category of clustering is greater than a certain threshold, the preset conditions are adjusted to reduce the number of samples of the current cluster center and dynamically increase the number of cluster centers.

7. The method for intelligent management of anesthesia information based on big data according to claim 6, characterized in that: The newly added cluster center is generated within a certain range around the current cluster center.

8. The method for intelligent management of anesthesia information based on big data according to claim 1, characterized in that: The improved random forest algorithm trains each decision tree based on different training sets; the different training sets are composed of H sub-training sets, where H is the number of anesthesia population classifications obtained in step S2; If the number of sub-training sets is greater than the number of decision trees, K of the H sub-training sets are randomly selected to train K decision trees respectively. If the number of sub-training sets is less than the number of decision trees, 2~H-1 subsets are randomly selected to form a training set.

9. An intelligent management platform for anesthesia information based on big data, based on the intelligent management method for anesthesia information based on big data as claimed in any one of claims 1 to 8, comprising: An anesthesia big data information acquisition module, which is used to acquire anesthesia big data information, wherein the anesthesia big data information includes patient personal information, physiological information, disease information, and intraoperative anesthesia information; Anesthesia big data processing module, which is used to process data based on the improved K-MEANS combined with DBSCAN algorithm, filter out and delete the anesthesia data of special patients and classify the anesthesia data to obtain several types of anesthesia populations; it includes filtering out the anesthesia big data information of special patients and classifying the anesthesia big data information of general patients based on the improved dynamic K-Means clustering algorithm; An intelligent generation module for anesthesia strategies is used to intelligently generate anesthesia strategies based on an improved random forest algorithm, wherein the input of the random forest algorithm is patient personal information, physiological information, and disease information, and the output is intraoperative anesthesia information; wherein each decision tree is trained based on different training sets; and the different training sets are composed of a plurality of sub-training sets.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program instructions of an intelligent management method for anesthesia information based on big data, and the program instructions of the intelligent management method for anesthesia information based on big data can be executed by one or more processors to implement the steps of the intelligent management method for anesthesia information based on big data as described in any one of claims 1 to 8.

Images