Decision tree-based consultation allocation method, device, equipment and storage medium

Abstract

The present invention relates to the field of artificial intelligence, and is applied to the field of smart medical care. A method, device, equipment and storage medium for consultation allocation based on a decision tree are disclosed, which are used to improve the efficiency of consultation allocation. The consultation allocation method based on a decision tree includes: obtaining initial medical data, preprocessing the initial medical data, and obtaining preprocessed medical data; calling a preset clustering algorithm, clustering the preprocessed medical data, and obtaining cluster samples; calling a preset decision tree-based vector machine classification algorithm, classifying the cluster samples based on the number of symptom phases, and obtaining symptom classification results; calling a preset similarity algorithm, matching the symptom classification results with a preset consultation doctor information database, and obtaining a consultation allocation result. In addition, the present invention also relates to blockchain technology, and the consultation allocation result can be stored in a blockchain node.

Description

Technical Field

[0001] The present invention relates to the field of decision trees, and in particular to a method, device, equipment and storage medium for allocating medical consultations based on a decision tree. Background Art

[0002] Online consultation will be the general trend of Internet medical care, and how to do a good job in symptom analysis and rapid patient diversion is particularly important. In recent years, with the rapid development of artificial intelligence technology, artificial intelligence has been applied to symptom analysis and patient diversion. Symptom analysis conducted by artificial intelligence can support disease auxiliary diagnosis, health management, remote consultation and other functions.

[0003] Traditional medical guidance is based on simple and rough division in a manual manner, which cannot guarantee a full understanding of the condition. The current online medical guidance is more of a random allocation. At present, due to the imperfect mechanism for allocating doctors for medical guidance, some patients cannot be promptly assigned to professional doctors due to their special symptoms. At the same time, when the number of patients is large, the existing allocation mechanism cannot respond in time, resulting in low efficiency in consultation allocation. Summary of the invention

[0004] The present invention provides a method, device, equipment and storage medium for assigning consultations based on a decision tree, which are used to call a preset clustering algorithm, perform clustering processing on preprocessed medical data, obtain cluster samples, call a preset decision tree-based vector machine classification algorithm, classify the cluster samples based on the number of symptom phases, obtain symptom classification results, and finally assign consulting doctors based on the symptom classification results, thereby improving the efficiency of consultation assignment.

[0005] The first aspect of the present invention provides a consultation allocation method based on a decision tree, comprising: obtaining initial medical data, preprocessing the initial medical data to obtain preprocessed medical data; calling a preset clustering algorithm to cluster the preprocessed medical data to obtain cluster samples; calling a preset decision tree-based vector machine classification algorithm to classify the cluster samples based on the number of symptom phases to obtain symptom classification results; calling a preset similarity algorithm to match the symptom classification results with a preset consulting doctor information database to obtain a consultation allocation result.

[0006] Optionally, in a first implementation method of the first aspect of the present invention, the acquiring of initial medical data and preprocessing of the initial medical data to obtain preprocessed medical data include: receiving medical record data sent by the patient side and consultation result data sent by the doctor side, and determining the medical record data and the consultation result data as initial medical data; calling a preset software simulation package to perform simulation processing on the initial medical data to obtain medical simulation data, and performing missing value completion, outlier filtering and duplicate value filtering on the medical simulation data to obtain preprocessed medical data.

[0007] Optionally, in a second implementation method of the first aspect of the present invention, calling a preset clustering algorithm to cluster the preprocessed medical data to obtain cluster samples includes: based on a preset inter-class overlap formula in the preset clustering algorithm, calculating the spatial overlap corresponding to each type of disease in the preprocessed medical data to obtain a spatial overlap matrix; based on the spatial overlap matrix, merging and screening the preprocessed medical data to obtain cluster samples.

[0008] Optionally, in a third implementation of the first aspect of the present invention, the calling of a preset decision tree-based vector machine classification algorithm to classify the cluster samples based on the number of symptom phases to obtain a symptom classification result includes: determining the inter-class separability of the cluster samples, constructing a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm to obtain an initial decision tree; pruning the initial decision tree to obtain a target decision tree, determining the target decision tree as the symptom classification result, and the target decision tree contains multiple leaf nodes.

[0009] Optionally, in a fourth implementation manner of the first aspect of the present invention, determining the inter-class separability of the cluster samples, constructing a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm, and obtaining an initial decision tree includes: calling a preset separability measurement formula, calculating the inter-class separability of the cluster samples, sorting the inter-class separability in order from large to small, and obtaining a separation ranking result; extracting a preset number of inter-class separability from the separation ranking result, labeling the sample sets corresponding to the preset number of inter-class separability to obtain a labeled sample set, calling a preset decision tree-based vector machine classification algorithm to construct a binary classifier, and determining the labeled sample sets as child nodes of the binary classifier respectively, to obtain an initial decision tree.

[0010] Optionally, in a fifth implementation manner of the first aspect of the present invention, calling a preset similarity algorithm to match the symptom classification result with a preset consulting doctor information database to obtain a consultation assignment result includes: calling a preset similarity algorithm to perform similarity calculation on the symptom classification result and a preset consulting doctor information database to obtain a similarity calculation result, wherein the similarity calculation result includes multiple similarity values ​​and the symptom classification result and consulting doctor information corresponding to each similarity value; sorting the consulting doctor information in the similarity calculation result in descending order of the multiple similarity values ​​to obtain a similarity sorting result, and determining the consulting doctor information ranked first in the similarity sorting result as the consultation assignment result.

[0011] Optionally, in a sixth implementation manner of the first aspect of the present invention, before obtaining the initial medical data, preprocessing the initial medical data, and obtaining the preprocessed medical data, the decision tree-based consultation allocation method also includes: obtaining historical medical data, using a preset anonymization algorithm to desensitize the historical medical data to obtain desensitized data, wherein the historical medical data includes historical consultation data and consulting doctor data corresponding to the historical consultation data; loading the desensitized data into a preset database construction template to obtain a consulting doctor information database.

[0012] The second aspect of the present invention provides a consultation allocation device based on a decision tree, comprising: an acquisition module, used to acquire initial medical data, pre-process the initial medical data, and obtain pre-processed medical data; a clustering module, used to call a preset clustering algorithm, cluster the pre-processed medical data, and obtain cluster samples; a classification module, used to call a preset decision tree-based vector machine classification algorithm, classify the cluster samples based on the number of symptom phases, and obtain symptom classification results; a matching module, used to call a preset similarity algorithm, match the symptom classification results with a preset consulting doctor information database, and obtain a consultation allocation result.

[0013] Optionally, in a first implementation method of the second aspect of the present invention, the acquisition module includes: a receiving unit, used to receive medical record data sent by the patient side and consultation result data sent by the doctor side, and determine the medical record data and the consultation result data as initial medical data; a preprocessing unit, used to call a preset software simulation package, perform simulation processing on the initial medical data to obtain medical simulation data, and perform missing value completion, outlier filtering and duplicate value filtering on the medical simulation data to obtain preprocessed medical data.

[0014] Optionally, in a second implementation of the second aspect of the present invention, the clustering module includes: a calculation unit, used to calculate the spatial overlap corresponding to each type of disease in the pre-processed medical data based on the inter-class overlap formula preset in the preset clustering algorithm, and obtain a spatial overlap matrix; a screening unit, used to merge and screen the pre-processed medical data based on the spatial overlap matrix to obtain cluster samples.

[0015] Optionally, in a third implementation of the second aspect of the present invention, the classification module includes: a classification unit, used to determine the inter-class separability of the cluster samples, and to construct a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm to obtain an initial decision tree; a pruning unit, used to prune the initial decision tree to obtain a target decision tree, and determine the target decision tree as a disease classification result, wherein the target decision tree includes multiple leaf nodes.

[0016] Optionally, in a fourth implementation manner of the second aspect of the present invention, the classification unit can be specifically used to: call a preset separability measurement formula, calculate the inter-class separability of the cluster samples, sort the inter-class separability in order from large to small, and obtain a separation ranking result; extract a preset number of inter-class separability in the separation ranking result, label the sample sets corresponding to the preset number of inter-class separability to obtain a labeled sample set, call a preset decision tree-based vector machine classification algorithm to construct a binary classifier, and determine the labeled sample sets as child nodes of the binary classifier respectively to obtain an initial decision tree.

[0017] Optionally, in a fifth implementation of the second aspect of the present invention, the matching module includes: a similarity calculation unit, used to call a preset similarity algorithm, perform similarity calculation on the disease classification result and a preset consulting doctor information database, to obtain a similarity calculation result, wherein the similarity calculation result includes multiple similarity values ​​and the disease classification result and consulting doctor information corresponding to each similarity value; a sorting unit, used to sort the consulting doctor information in the similarity calculation result in a descending order of multiple similarity values, to obtain a similarity sorting result, and determine the consulting doctor information ranked first in the similarity sorting result as the consultation assignment result.

[0018] Optionally, in a sixth implementation method of the second aspect of the present invention, before the acquisition module, the decision tree-based consultation allocation device also includes a construction module, and the construction module includes: a desensitizing unit, used to acquire historical medical data, and use a preset anonymization algorithm to desensitize the historical medical data to obtain desensitized data, wherein the historical medical data includes historical consultation data and consulting doctor data corresponding to the historical consultation data; a loading unit, used to load the desensitized data into a preset database construction template to obtain a consulting doctor information database.

[0019] The third aspect of the present invention provides a decision tree-based consultation assignment device, comprising: a memory and at least one processor, wherein the memory stores a computer program; the at least one processor calls the computer program in the memory so that the decision tree-based consultation assignment device executes the above-mentioned decision tree-based consultation assignment method.

[0020] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, which, when executed on a computer, enables the computer to execute the above-mentioned decision tree-based consultation allocation method.

[0021] In the technical solution provided by the present invention, initial medical data is obtained, and the initial medical data is preprocessed to obtain preprocessed medical data; a preset clustering algorithm is called to cluster the preprocessed medical data to obtain cluster samples; a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain symptom classification results; a preset similarity algorithm is called to match the symptom classification results with a preset consulting doctor information database to obtain a consultation allocation result. In an embodiment of the present invention, a preset clustering algorithm is called to cluster the preprocessed medical data to obtain cluster samples, a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain a symptom classification result, and finally the consulting doctor is allocated based on the symptom classification result, thereby improving the efficiency of consultation allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 A schematic diagram of an embodiment of a method for allocating medical consultations based on a decision tree in an embodiment of the present invention;

[0023] Figure 2 It is a schematic diagram of another embodiment of the method for allocating medical consultations based on a decision tree in an embodiment of the present invention;

[0024] Figure 3 It is a schematic diagram of an embodiment of a device for allocating medical consultations based on a decision tree in an embodiment of the present invention;

[0025] Figure 4 It is a schematic diagram of another embodiment of a medical consultation allocation device based on a decision tree in an embodiment of the present invention;

[0026] Figure 5 The figure is a schematic diagram of an embodiment of a medical consultation allocation device based on a decision tree in an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The embodiments of the present invention provide a method, apparatus, device and storage medium for assigning consultations based on a decision tree, which are used to call a preset clustering algorithm, perform clustering processing on preprocessed medical data, obtain cluster samples, call a preset decision tree-based vector machine classification algorithm, classify the cluster samples based on the number of symptom phases, obtain symptom classification results, and finally assign consulting doctors based on the symptom classification results, thereby improving the efficiency of consultation assignment.

[0028] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the method for allocating medical consultations based on a decision tree in an embodiment of the present invention includes:

[0030] 101. Acquire initial medical data, and preprocess the initial medical data to obtain preprocessed medical data.

[0031] It is understandable that the execution subject of the present invention can be a consultation allocation device based on a decision tree, or a terminal or a server, which is not limited here. The embodiment of the present invention is described by taking the server as the execution subject as an example. The server can be an independent server, or it can be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0032] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0033] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0034] The server obtains initial medical data, preprocesses the initial medical data, and obtains preprocessed medical data. The patient is tested and medical record data is generated through a preset medical device. The medical device uploads the medical record data to the medical cloud. The server extracts the corresponding medical record data from the medical cloud according to the patient information. At the same time, the medical consultation result data is obtained from the doctor. The medical record data and consultation result data in this embodiment are authorized by the user. The obtained case data and consultation result data are the initial medical data. The process of preprocessing the initial medical data mainly includes missing value completion, outlier filtering and duplicate value filtering to obtain preprocessed medical data.

[0035] 102. Call a preset clustering algorithm to perform clustering processing on the preprocessed medical data to obtain cluster samples.

[0036] The server calls the preset clustering algorithm to cluster the pre-processed medical data and obtain cluster samples. The spatial overlap matrix is ​​calculated by calling the preset inter-class overlap formula in the clustering algorithm, and then the pre-processed medical data is screened to finally obtain cluster samples. The decision tree-based vector machine classification algorithm used in this embodiment uses a general measurement method to determine the degree of difficulty of class separability, and there is still a certain deviation. Therefore, this solution first clusters the pre-processed medical data, and the obtained cluster samples can effectively realize the classification of symptoms in the downstream classification process. Clustering algorithms can be applied to a variety of different fields, such as: speech recognition, medical diagnosis, application testing, etc.

[0037] 103. Call the preset decision tree-based vector machine classification algorithm to classify the clustered samples based on the number of symptom phases to obtain the symptom classification results.

[0038] The server calls the preset decision tree-based vector machine classification algorithm to classify the cluster samples based on the number of symptom phases and obtain the symptom classification results. The number of symptom phases is used to refer to the number of symptom features of the patient. For example, the initial medical data corresponding to patient A, after processing, the cluster samples obtained include "osteoporosis / arthritis / bone trauma", and the corresponding number of symptom phases is 3. The server constructs a binary SVM classifier through the decision tree support vector machine classification algorithm (decision tree support vector machines, DTSVM). The classified positive and negative sample sets (are the two child nodes of the decision tree. Support vector machine (SVM) is a machine learning algorithm based on the principle of structural risk minimization of statistical learning theory. It has shown great advantages in solving small sample, nonlinear and high-dimensional pattern recognition, and can be extended to other machine learning such as function fitting.

[0039] 104. Call a preset similarity algorithm to match the disease classification result with a preset consulting doctor information database to obtain a consultation allocation result.

[0040] The server calls a preset similarity algorithm to match the symptom classification result with the preset consulting doctor information database to obtain the consultation allocation result. The preset similarity algorithm can be a Euclidean metric algorithm, a Pearson correlation coefficient algorithm, or a cosine similarity algorithm. By calling the similarity algorithm to perform similarity calculation on the symptom classification result and the consulting doctor information database, a similarity calculation result can be obtained. The similarity calculation result includes multiple similarity values ​​and the symptom classification result and consulting doctor information corresponding to each similarity value. The server extracts the corresponding consulting doctor information in the similarity calculation result, thereby determining the consultation allocation result. Among them, the consulting doctor information database is obtained by obtaining historical medical data, using an anonymization algorithm to desensitize the historical medical data, and then loading it into a preset database construction template.

[0041] In the embodiment of the present invention, a preset clustering algorithm is called to perform clustering processing on the preprocessed medical data to obtain cluster samples, and a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain symptom classification results, and finally the allocation of consulting doctors is performed based on the symptom classification results, thereby improving the efficiency of consultation allocation. This solution can be applied to the field of smart medical care, thereby promoting the construction of smart cities.

[0042] See also Figure 2 Another embodiment of the decision tree-based consultation allocation method in the embodiment of the present invention includes:

[0043] 201. Acquire initial medical data, and preprocess the initial medical data to obtain preprocessed medical data.

[0044] The server obtains initial medical data, pre-processes the initial medical data, and obtains pre-processed medical data. Specifically, the server receives the medical record data sent by the patient and the consultation result data sent by the doctor, and determines the medical record data and the consultation result data as the initial medical data; the server calls a preset software simulation package, simulates the initial medical data, obtains medical simulation data, and performs missing value completion, outlier filtering, and duplicate value filtering on the medical simulation data to obtain pre-processed medical data.

[0045] The medical record data includes multiple project names and patient examination data corresponding to each project name. The medical record data can be stored in the patient's corresponding health record. The consultation result data is the data obtained by the doctor after diagnosis based on the medical record data. The server determines the medical record data and the consultation result data as the initial medical data. After generating the initial medical data, the software simulation package is used to simulate typical symptoms of different phases to obtain medical simulation data, and then the medical simulation data is preprocessed. The preprocessing process mainly includes missing value filling, outlier filtering and duplicate value filtering. Among them, missing value filling can be multiple interpolation or hot card filling. Outlier filtering mainly uses the outlier detection algorithm z-score to identify and delete outliers. Duplicate value filtering is to deduplicate duplicate values ​​to obtain preprocessed medical data. The preprocessed medical data can be expressed as Data={(x i ,y i ), i = 1, ..., N}, N represents the total number of sample sets, the preprocessed data contains data corresponding to different disease phases, the disease phase number is used to refer to the number of disease characteristics of patients, and the disease data of different phases are recorded as Where K means there are K phase diseases, M means there are no more than M diseases, n k Indicates the number of samples of K-phase disease.

[0046] 202. Call a preset clustering algorithm to perform clustering processing on the preprocessed medical data to obtain cluster samples.

[0047] The server calls a preset clustering algorithm to perform clustering processing on the pre-processed medical data to obtain cluster samples. Specifically, the server calculates the spatial overlap corresponding to each type of disease in the pre-processed medical data based on the preset inter-class overlap formula in the preset clustering algorithm to obtain a spatial overlap matrix; the server merges and filters the pre-processed medical data based on the spatial overlap matrix to obtain cluster samples.

[0048] The overlap between classes is calculated based on the distribution of the eigenvalues ​​of the two classes. The specific formula (i.e., the overlap between classes formula) is defined as follows: Assume that the preprocessed medical data contains two categories: class c1 and class c2, where i is the dimension label and m is the total dimension of class c1 and class c2. f i represents the eigenvalue of the i-th dimension in the preprocessed medical data of types c1 and c2, represents the maximum value of the eigenvalue of the i-th dimension in the c1 class training samples, represents the maximum value of the eigenvalue of the i-th dimension in the c2 class training samples, represents the minimum value of the eigenvalue of the i-th dimension in the c1 class training samples, Represents the minimum value of the eigenvalue of the i-th dimension in the c2 class training samples; Suppose there are a total of n types of disease data O in the preprocessed medical data. ij Represents the overlap between class i and class j, and the spatial overlap matrix is: For the overlap matrix O, such as O ij ≠0, then merge all categories in the preprocessed medical data, denoted as D i If the i-th dimension or the j-th dimension (i and j are dimension labels) overlaps with other dimensions, the overlapping parts will be merged until all overlapping classes are merged into one class. For the non-overlapping classes, they will be retained separately and recorded as new classes D j , and finally get the cluster samples.

[0049] 203. Determine the inter-class separability of the cluster samples, build a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm, and obtain an initial decision tree.

[0050] The server determines the inter-class separability of the clustered samples, constructs a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm, and obtains an initial decision tree. Specifically, the server calls a preset separability measurement formula, calculates the inter-class separability of the clustered samples, sorts the inter-class separability in descending order, and obtains a separation ranking result; the server extracts a preset number of inter-class separability from the separation ranking result, labels the sample sets corresponding to the preset number of inter-class separability, obtains the labeled sample sets, calls a preset decision tree-based vector machine classification algorithm to construct a binary classifier, and determines the labeled sample sets as child nodes of the binary classifier, and obtains an initial decision tree.

[0051] The preset separability measurement formula is as follows: The inter-class separation measure of class a and class b in the cluster sample is defined as: where d ab =||c a -c b || represents the center distance between class a and class b, c a With cb are the class centers of class a and class b respectively, n a is the number of samples, k is the number of sample categories, δ a is the class variance of class a. γ ab The larger the value, the better the separation between class a and class b. Then define the inter-class separation measure matrix γ as The calculation formula for the degree of separability between classes is: S a represents the degree of separability between classes, and S a Sort them in descending order to get the separation degree sorting result. Take the first m / 2 sample sets in the separation degree sorting result and mark them as positive classes, and mark the remaining classes as negative classes to get the labeled sample set. Construct a binary vector machine classifier based on the decision tree support vector machine classification algorithm (DTSVM). The classified positive and negative sample sets (i.e., the labeled sample sets) are the two child nodes of the decision tree. If the sample set of the child node contains more than two categories, then construct a binary vector machine classifier according to the above division scheme until all D i are separated and the initial decision tree is obtained.

[0052] 204. Prune the initial decision tree to obtain a target decision tree, and determine the target decision tree as a disease classification result. The target decision tree includes multiple leaf nodes.

[0053] The server prunes the initial decision tree to obtain a target decision tree, and determines the target decision tree as the disease classification result, and the target decision tree contains multiple leaf nodes. After the initial decision tree is generated, it is necessary to prune the initial decision tree to obtain a target decision tree, which is the disease classification result. The pruning process used in this embodiment is mainly post-pruning. Post-pruning is bottom-up pruning, which refers to estimating non-leaf nodes from the bottom up for a complete decision tree that has been generated. If replacing the subtree corresponding to the node with a leaf node can improve the generalization performance of the decision tree, the subtree is replaced with a leaf node. Post-pruning mainly includes: reduced-error pruning (REP), pessimistic-error pruning (PEP), cost-complexity pruning (CCP) and error-based pruning (EBP).

[0054] 205. Call a preset similarity algorithm to match the symptom classification result with a preset consulting doctor information database to obtain a consultation allocation result.

[0055] The server calls a preset similarity algorithm to match the symptom classification result with the preset consulting doctor information database to obtain the consultation allocation result. Specifically, the server calls a preset similarity algorithm to perform similarity calculation on the symptom classification result and the preset consulting doctor information database to obtain a similarity calculation result, and the similarity calculation result includes multiple similarity values ​​and the symptom classification result and consulting doctor information corresponding to each similarity value; the server sorts the consulting doctor information in the similarity calculation result in descending order of the multiple similarity values ​​to obtain a similarity sorting result, and determines the consulting doctor information ranked first in the similarity sorting result as the consultation allocation result.

[0056] For example, the disease classification results include information such as "osteoporosis, arthritis, and bone trauma". By performing similarity calculation with the consulting doctor information database, the similarity calculation results are: Dr. Zhang, who specializes in joint replacement, limb fractures, tumor surgery, and joint revision surgery, with a similarity value of 30%; Dr. Wang, who specializes in the diagnosis and treatment of various bone degenerative diseases, the treatment of multiple traumas, and the treatment of various complex intra-articular comminuted fractures, with a similarity value of 70%; Dr. Li, who specializes in standardized treatment of osteoporosis and high-risk screening, diagnosis and treatment of common bone and joint diseases, and trauma treatment, with a similarity value of 80%. The similarity values ​​are then sorted to obtain the similarity sorting results, and the consulting doctor information ranked first in the similarity sorting results (i.e., Dr. Li corresponding to a similarity value of 80%) is determined as the consultation allocation result.

[0057] In the embodiment of the present invention, a preset clustering algorithm is called to perform clustering processing on the preprocessed medical data to obtain cluster samples, and a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain symptom classification results, and finally the allocation of consulting doctors is performed based on the symptom classification results, thereby improving the efficiency of consultation allocation. This solution can be applied to the field of smart medical care, thereby promoting the construction of smart cities.

[0058] The above describes the method for allocating medical consultations based on a decision tree in an embodiment of the present invention. The following describes the device for allocating medical consultations based on a decision tree in an embodiment of the present invention. Figure 3 In one embodiment of the present invention, a decision tree-based consultation allocation device includes:

[0059] An acquisition module 301 is used to acquire initial medical data and pre-process the initial medical data to obtain pre-processed medical data;

[0060] The clustering module 302 is used to call a preset clustering algorithm to perform clustering processing on the pre-processed medical data to obtain cluster samples;

[0061] The classification module 303 is used to call a preset decision tree-based vector machine classification algorithm to classify the clustered samples based on the number of symptom phases to obtain a symptom classification result;

[0062] The matching module 304 is used to call a preset similarity algorithm to match the symptom classification result with a preset consulting doctor information database to obtain a consultation allocation result.

[0063] In the embodiment of the present invention, a preset clustering algorithm is called to perform clustering processing on the preprocessed medical data to obtain cluster samples, and a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain symptom classification results, and finally the allocation of consulting doctors is performed based on the symptom classification results, thereby improving the efficiency of consultation allocation. This solution can be applied to the field of smart medical care, thereby promoting the construction of smart cities.

[0064] See also Figure 4 Another embodiment of the decision tree-based consultation allocation device in the embodiment of the present invention includes:

[0065] An acquisition module 301 is used to acquire initial medical data and pre-process the initial medical data to obtain pre-processed medical data;

[0066] The clustering module 302 is used to call a preset clustering algorithm to perform clustering processing on the pre-processed medical data to obtain cluster samples;

[0067] The classification module 303 is used to call a preset decision tree-based vector machine classification algorithm to classify the clustered samples based on the number of symptom phases to obtain a symptom classification result;

[0068] The classification module 303 specifically includes:

[0069] The classification unit 3031 is used to determine the inter-class separability of the cluster samples, and to construct a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm to obtain an initial decision tree;

[0070] A pruning unit 3032 is used to prune the initial decision tree to obtain a target decision tree, and determine the target decision tree as the disease classification result. The target decision tree includes multiple leaf nodes;

[0071] The matching module 304 is used to call a preset similarity algorithm to match the symptom classification result with a preset consulting doctor information database to obtain a consultation allocation result.

[0072] Optionally, the acquisition module 301 includes:

[0073] The receiving unit 3011 is used to receive the medical record data sent by the patient and the consultation result data sent by the doctor, and determine the medical record data and the consultation result data as initial medical data;

[0074] The preprocessing unit 3012 is used to call a preset software simulation package to perform simulation processing on the initial medical data to obtain medical simulation data, and to complete missing values, filter abnormal values ​​and filter duplicate values ​​on the medical simulation data to obtain preprocessed medical data.

[0075] Optionally, the clustering module 302 includes:

[0076] The calculation unit 3021 is used to calculate the spatial overlap corresponding to each type of disease in the pre-processed medical data based on the preset inter-class overlap formula in the preset clustering algorithm to obtain a spatial overlap matrix;

[0077] The screening unit 3022 is used to merge and screen the pre-processed medical data based on the spatial overlap matrix to obtain cluster samples.

[0078] Optionally, the classification unit 3031 may be specifically used for:

[0079] Call the preset separability measurement formula, calculate the inter-class separability of cluster samples, sort the inter-class separability in descending order, and obtain the separation ranking result; extract a preset number of inter-class separability in the separation ranking result, label the sample sets corresponding to the preset number of inter-class separability, and obtain the labeled sample sets, call the preset decision tree-based vector machine classification algorithm to build a binary classifier, determine the labeled sample sets as child nodes of the binary classifier, and obtain an initial decision tree.

[0080] Optionally, the matching module 304 includes:

[0081] The similarity calculation unit 3041 is used to call a preset similarity algorithm to perform similarity calculation on the disease classification result and the preset consulting doctor information database to obtain a similarity calculation result, wherein the similarity calculation result includes multiple similarity values ​​and the disease classification result and consulting doctor information corresponding to each similarity value;

[0082] The sorting unit 3042 is used to sort the consulting doctor information in the similarity calculation result in descending order of multiple similarity values ​​to obtain a similarity sorting result, and determine the consulting doctor information ranked first in the similarity sorting result as the consultation allocation result.

[0083] Optionally, before the acquisition module 301, the decision tree-based consultation allocation device further includes a construction module 305, including:

[0084] The desensitization unit 3051 is used to obtain historical medical data, and perform desensitization processing on the historical medical data using a preset anonymization algorithm to obtain desensitized data, wherein the historical medical data includes historical consultation data and consultation doctor data corresponding to the historical consultation data;

[0085] The loading unit 3052 is used to load the desensitized data into a preset database construction template to obtain a consulting doctor information database.

[0086] In the embodiment of the present invention, a preset clustering algorithm is called to perform clustering processing on the preprocessed medical data to obtain cluster samples, and a preset decision tree-based vector machine classification algorithm is called to classify the cluster samples based on the number of symptom phases to obtain symptom classification results, and finally the allocation of consulting doctors is performed based on the symptom classification results, thereby improving the efficiency of consultation allocation. This solution can be applied to the field of smart medical care, thereby promoting the construction of smart cities.

[0087] above Figure 3 and Figure 4 The decision tree-based consultation allocation device in the embodiment of the present invention is described in detail from the perspective of modular functional entities. The decision tree-based consultation allocation device in the embodiment of the present invention is described in detail from the perspective of hardware processing.

[0088] Figure 5 1 is a schematic diagram of the structure of a decision tree-based consultation allocation device provided by an embodiment of the present invention. The decision tree-based consultation allocation device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of computer program operations in the decision tree-based consultation allocation device 500. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of computer program operations in the storage medium 530 on the decision tree-based consultation allocation device 500.

[0089] The decision tree-based consultation allocation device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that Figure 5 The structure of the decision tree-based consultation assignment device shown does not constitute a limitation of the decision tree-based consultation assignment device, and may include more or less components than shown in the figure, or combine some components, or arrange the components differently.

[0090] The present application also provides a decision tree-based consultation allocation device, comprising: a memory and at least one processor, wherein a computer program is stored in the memory, and the memory and the at least one processor are interconnected via a line; the at least one processor calls the computer program in the memory so that the decision tree-based consultation allocation device performs the steps in the above-mentioned decision tree-based consultation allocation method.

[0091] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein a computer program is stored in the computer, and when the computer program runs on a computer, the computer executes the steps of the decision tree-based consultation assignment method.

[0092] Furthermore, the computer-readable storage medium may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the blockchain node, etc.

[0093] The blockchain referred to in this invention is a new application mode of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0094] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0095] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several computer programs to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0096] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for allocating medical consultations based on a decision tree, characterized in that: The decision tree-based consultation allocation method includes: Acquiring initial medical data, and preprocessing the initial medical data to obtain preprocessed medical data; Based on the preset inter-class overlap formula in the preset clustering algorithm, the spatial overlap corresponding to each type of disease in the pre-processed medical data is calculated to obtain a spatial overlap matrix; Based on the spatial overlap matrix, the pre-processed medical data are merged and screened to obtain cluster samples; Calling a preset decision tree-based vector machine classification algorithm to classify the clustered samples based on the number of symptom phases to obtain a symptom classification result, where the number of symptom phases is used to refer to the number of symptom characteristics of the patient; The preset similarity algorithm is called to match the disease classification result with the preset consulting doctor information database to obtain the consultation allocation result.

2. The decision tree-based consultation allocation method according to claim 1, characterized in that: The obtaining of initial medical data and preprocessing the initial medical data to obtain preprocessed medical data comprises: Receiving medical record data sent by the patient and consultation result data sent by the doctor, and determining the medical record data and the consultation result data as initial medical data; A preset software simulation package is called to perform simulation processing on the initial medical data to obtain medical simulation data, and missing value completion, abnormal value filtering and duplicate value filtering are performed on the medical simulation data to obtain preprocessed medical data.

3. The decision tree-based consultation allocation method according to claim 1, characterized in that: The calling of the preset decision tree-based vector machine classification algorithm to classify the cluster samples based on the number of symptom phases, and obtaining the symptom classification results includes: Determine the inter-class separability of the cluster samples, construct a decision tree through the inter-class separability and a preset decision tree-based vector machine classification algorithm, and obtain an initial decision tree; The initial decision tree is pruned to obtain a target decision tree, and the target decision tree is determined as a disease classification result, wherein the target decision tree includes a plurality of leaf nodes.

4. The decision tree-based consultation allocation method according to claim 3, characterized in that: The step of determining the inter-class separability of the cluster samples, constructing a decision tree by using the inter-class separability and a preset decision tree-based vector machine classification algorithm, and obtaining an initial decision tree comprises: Calling a preset separability measurement formula to calculate the inter-class separability of the cluster samples, sorting the inter-class separability in descending order, and obtaining a separation degree sorting result; Extract a preset number of inter-class separability degrees from the separation degree sorting result, annotate the sample sets corresponding to the preset number of inter-class separability degrees to obtain annotated sample sets, call a preset decision tree-based vector machine classification algorithm to build a binary classifier, and determine the annotated sample sets as child nodes of the binary classifier respectively to obtain an initial decision tree.

5. The decision tree-based consultation allocation method according to claim 1, characterized in that: The calling of a preset similarity algorithm to match the symptom classification result with a preset consulting doctor information database to obtain a consultation allocation result includes: Calling a preset similarity algorithm, performing similarity calculation on the disease classification result and a preset consulting doctor information database to obtain a similarity calculation result, wherein the similarity calculation result includes a plurality of similarity values ​​and the disease classification result and consulting doctor information corresponding to each similarity value; The consulting doctor information in the similarity calculation result is sorted in descending order of multiple similarity values ​​to obtain a similarity sorting result, and the consulting doctor information ranked first in the similarity sorting result is determined as the consultation allocation result.

6. The method for allocating medical consultations based on a decision tree according to any one of claims 1 to 5, characterized in that: Before obtaining the initial medical data and preprocessing the initial medical data to obtain the preprocessed medical data, the decision tree-based consultation allocation method further includes: Acquire historical medical data, and perform desensitization processing on the historical medical data using a preset anonymization algorithm to obtain desensitized data, wherein the historical medical data includes historical consultation data and consultation doctor data corresponding to the historical consultation data; The desensitized data is loaded into a preset database construction template to obtain a consulting doctor information database.

7. A medical consultation allocation device based on a decision tree, characterized in that: The decision tree-based consultation allocation device comprises: An acquisition module, used for acquiring initial medical data, and preprocessing the initial medical data to obtain preprocessed medical data; A clustering module, for calculating the spatial overlap corresponding to each type of disease in the pre-processed medical data based on a preset inter-class overlap formula in a preset clustering algorithm, and obtaining a spatial overlap matrix; Based on the spatial overlap matrix, the pre-processed medical data are merged and screened to obtain cluster samples; A classification module, used for calling a preset decision tree-based vector machine classification algorithm to classify the clustered samples based on the number of symptom phases to obtain a symptom classification result, where the number of symptom phases is used to refer to the number of symptom characteristics of the patient; The matching module is used to call a preset similarity algorithm to match the disease classification result with a preset consulting doctor information database to obtain a consultation allocation result.

8. A consultation allocation device based on a decision tree, characterized in that: The decision tree-based consultation allocation device comprises: a memory and at least one processor, wherein the memory stores a computer program; The at least one processor calls the computer program in the memory to enable the decision tree-based consultation assignment device to perform the decision tree-based consultation assignment method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the decision tree-based consultation allocation method according to any one of claims 1 to 6 is implemented.

Images