CN113539460A - Intelligent diagnosis guiding method and device for remote medical platform - Google Patents
Intelligent diagnosis guiding method and device for remote medical platform Download PDFInfo
- Publication number
- CN113539460A CN113539460A CN202110865124.6A CN202110865124A CN113539460A CN 113539460 A CN113539460 A CN 113539460A CN 202110865124 A CN202110865124 A CN 202110865124A CN 113539460 A CN113539460 A CN 113539460A
- Authority
- CN
- China
- Prior art keywords
- patient
- doctor
- data
- candidate
- determining
- 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.)
- Pending
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 66
- 238000000034 method Methods 0.000 title claims abstract description 45
- 239000013598 vector Substances 0.000 claims description 88
- 238000012549 training Methods 0.000 claims description 54
- 238000012545 processing Methods 0.000 claims description 17
- 239000013604 expression vector Substances 0.000 claims description 16
- 230000011218 segmentation Effects 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 6
- 238000004140 cleaning Methods 0.000 claims description 5
- 239000002131 composite material Substances 0.000 claims description 4
- 230000002349 favourable effect Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 description 9
- 238000012706 support-vector machine Methods 0.000 description 6
- 238000012795 verification Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000012417 linear regression Methods 0.000 description 4
- 238000009223 counseling Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000001419 dependent effect Effects 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Images
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
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/20—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
-
- 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
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Data Mining & Analysis (AREA)
- Primary Health Care (AREA)
- Pathology (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The invention relates to an intelligent diagnosis guiding method and device for a remote medical platform, and belongs to the technical field of intelligent diagnosis guiding. The method comprises the following steps: acquiring patient data; determining a target department of the patient according to the patient data and a preset department prediction model; determining candidate doctors of the patient according to the target department; acquiring doctor data of each candidate doctor; determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and doctor data of the candidate doctors; determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor; and determining the recommended doctor of the patient according to the comprehensive score. This application can match administrative or technical offices and doctor for the patient automatically, matches the rate of accuracy height, has guaranteed that the patient selects correct administrative or technical offices and doctor, more satisfies the patient demand, has avoided high-quality medical resource to be taken up excessively, and ordinary medical resource is idle, and medical resource utilization is higher.
Description
Technical Field
The invention relates to the technical field of intelligent diagnosis guide, in particular to an intelligent diagnosis guide method and device for a remote medical platform.
Background
The medical platform is used for realizing the purpose of remote communication between a patient and a doctor. Before the patient communicates with the doctor, the medical platform needs to match the patient with the appropriate doctor. In the related art, the patient is required to autonomously select a clinic and a doctor according to the understanding of the condition of the patient.
However, since the medical knowledge itself is complex and highly professional, it is often difficult for the patient to make a correct judgment on his or her condition, and thus to correctly select a department and a doctor. Moreover, patients often select doctors according to the medical experience and the famous atmosphere of the doctors, so that high-quality medical resources are excessively occupied, common medical resources are idle, and the utilization rate of the medical resources is low.
Disclosure of Invention
In view of this, an intelligent diagnosis guiding method and apparatus for a remote medical platform are provided to solve the problems in the related art that it is difficult for a patient to correctly select a department and a doctor, and that high-quality medical resources are excessively occupied, common medical resources are idle, and the utilization rate of the medical resources is low.
The invention adopts the following technical scheme:
in a first aspect, the present application provides an intelligent referral method for a telemedicine platform, comprising:
acquiring patient data;
determining a target department of the patient according to the patient data and a preset department prediction model;
determining candidate doctors of the patient according to the target department;
acquiring doctor data of each candidate doctor;
determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a score of the candidate physician for a physician feature relevant to the patient's decision to seek medical attention;
determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor;
and determining the recommended doctor of the patient according to the comprehensive score.
Preferably, the patient data comprises patient characteristic data;
the determining the target department of the patient according to the patient data and a preset department prediction model comprises:
determining a patient characteristic representation vector according to the patient characteristic data and a preset patient characteristic representation model;
and determining the target department according to the patient feature representation vector and the preset department prediction model.
Preferably, the patient characteristic data comprises patient personal information data and patient counseling text; the patient personal information data comprises the age and the sex of the patient;
the determining a patient feature representation vector according to the patient data and a preset patient feature representation model comprises:
performing preset word segmentation processing on the patient consultation text to obtain a word segmentation paragraph patient consultation text;
determining a patient consultation text representation vector according to the word segmentation paragraph patient consultation text;
carrying out dummy variable processing on the gender of the patient to obtain a gender expression vector of the patient;
performing box-dividing normalization processing on the patient age to obtain a patient age representation vector;
determining the patient characteristic representation vector based on the patient consultation text representation vector, the patient gender representation vector, and the patient age representation vector.
Preferably, the doctor data comprises diagnosis and treatment characteristic data of a doctor and doctor characteristic data related to the hospitalization decision of the patient;
determining doctor-patient similarity and decision factor scores of the candidate doctors according to the patient data and the doctor data of the candidate doctors, wherein the steps comprise:
determining the doctor-patient similarity of the candidate doctors according to the patient feature representation vector and the diagnosis and treatment feature data of the doctors;
judging whether the patient data contains doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient or not to obtain a preset result;
when the preset result is yes, determining the decision factor score of the candidate doctor according to the doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient and the doctor feature data related to the patient hospitalizing decision of the candidate doctor;
when the preset result is negative, acquiring training patient data from a preset database; the training patient data comprises training patient characteristic data and doctor characteristic data related to patient medical decision of a target historical medical doctor of the training patient;
determining the patient feature similarity of each training patient and the patient according to the training patient feature data and the patient feature data;
determining the target patient characteristic similarity with the maximum value from all the patient characteristic similarities;
defining doctor feature data related to patient medical decision of a target history doctor for training the patient corresponding to the target patient feature similarity as doctor feature data related to patient medical decision of the target history doctor for the patient;
determining a decision factor score for the candidate physician based on the patient hospitalization decision-related physician characteristic data for the patient historical attending physician and the patient hospitalization decision-related physician characteristic data for the candidate physician.
Preferably, the determining the doctor-patient similarity of the candidate doctor according to the patient feature representation vector and the diagnosis and treatment feature data of the doctor includes:
determining a doctor diagnosis and treatment characteristic representation vector according to the diagnosis and treatment characteristic data of the doctor and a preset doctor diagnosis and treatment characteristic representation model;
and determining the doctor-patient similarity of the candidate doctors according to the patient characteristic representation vector and the doctor diagnosis and treatment characteristic representation vector.
Preferably, the determining the decision factor score of the candidate doctor according to the doctor feature data related to the patient hospitalization decision of the target historical hospitalization doctor of the patient and the doctor feature data related to the patient hospitalization decision of the candidate doctor comprises:
determining a doctor decision factor characteristic representation vector according to doctor characteristic data related to the patient hospitalizing decision of the candidate doctor and a preset doctor decision factor characteristic representation model;
and determining the decision factor score of the candidate doctor according to the doctor characteristic data relevant to the patient medical decision of the target historical doctor of the patient and the doctor decision factor characteristic representation vector.
Preferably, before acquiring the patient data, the intelligent diagnosis guiding method for the remote medical platform of the present application further includes:
physician characteristics relevant to patient decision-making are determined.
Preferably, the determining doctor characteristics related to the patient hospitalization decision comprises:
determining candidate variables for medical decision-making;
cleaning preset data of the medical decision candidate variables to obtain medical decision candidate variable feature expression vectors;
determining doctor features related to the hospitalizing decision of the patient according to the doctor diagnosis and treatment feature representation vector and the hospitalizing decision candidate variable feature representation vector; the doctor characteristics related to the patient hospitalizing decision comprise inquiry price, favorable rating rate, mind number, peer acceptance and fan number.
Preferably, after determining the recommended doctor of the patient according to the composite score, the intelligent diagnosis guiding method for the remote medical platform of the present application further includes:
acquiring the service time of all the recommended doctors;
and determining the target doctor of the patient according to the service time of the recommended doctor and the comprehensive score.
In a second aspect, the present application provides an intelligent referral device for a telemedicine platform, comprising:
a patient data acquisition module for acquiring patient data;
the department prediction module is used for determining a target department of the patient according to the patient data and a preset department prediction model;
a candidate doctor determining module for determining candidate doctors of the patient according to the target department;
a doctor data acquisition module for acquiring doctor data of each candidate doctor;
the similarity matching module is used for determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and the doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a score of the candidate physician for a physician feature relevant to the patient's decision to seek medical attention;
the score calculation module is used for determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor;
and the recommended doctor determining module is used for determining the recommended doctor of the patient according to the comprehensive score.
By adopting the technical scheme, the invention provides an intelligent diagnosis guiding method for a remote medical platform, which comprises the following steps: acquiring patient data; determining a target department of the patient according to the patient data and a preset department prediction model; determining candidate doctors of the patient according to the target department; acquiring doctor data of each candidate doctor; determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is the score of the candidate doctor about doctor characteristics relevant to the patient hospitalization decision; determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor; and determining the recommended doctor of the patient according to the comprehensive score. Based on this, this application matches out the target department for the patient automatically according to patient data for the patient only needs to input relevant disease in the system can learn the department of seeing a doctor that corresponds, and the system matches the department for the patient automatically, matches the rate of accuracy height, has guaranteed that the patient selects the correct department. Secondly, the doctor-patient similarity and the decision factor score of each candidate doctor are determined according to the patient data and the doctor data, then the doctor-patient similarity and the decision factor score are matched with the doctor for the patient, the matching result is more accurate, the requirements of the patient are better met, high-quality medical resources are prevented from being excessively occupied, common medical resources are left unused, and the medical resource utilization rate is higher.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of an intelligent diagnosis guiding method for a remote medical platform according to an embodiment of the present invention.
Fig. 2 is a schematic structural diagram of an intelligent diagnosis guiding apparatus for a remote medical platform according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be described in detail below. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the examples given herein without any inventive step, are within the scope of the present invention.
Fig. 1 is a schematic flow chart of an intelligent diagnosis guiding method for a remote medical platform according to an embodiment of the present invention. As shown in fig. 1, the intelligent diagnosis guiding method for a remote medical platform of the present embodiment includes:
and S101, acquiring patient data.
In particular, the patient data includes patient characteristic data. The patient characteristic data includes patient age, gender and counseling text data. The advisory text data is data sent by the patient about his own condition.
S102, determining a target department of the patient according to the patient data and a preset department prediction model.
In detail, the preset department prediction model may be a multi-classifier constructed by using a linear support vector machine. The training data for which training is performed includes patient characteristics and corresponding department labels for the patient. The patient features are patient features in a vector format that are converted from patient feature data. When the model is trained, the training data is randomly divided into a training set and a verification set, the training set is used for training parameters of the model, and the verification set is used for evaluating the effect of the model, so that the parameters of the model are optimized. In a specific application process, training data is divided into K sub-samples, one sub-sample is sequentially selected from the K sub-samples as a verification set in each round of training, and the rest samples are used as a training set. Then, the patient characteristics of the training set and the department labels corresponding to the patients are input into the classifier, and various parameters of the classifier are obtained through training. The classifier is a two-classifier of the prior art. And finally, inputting the patient characteristics of the verification set into the trained classifier to obtain a corresponding department prediction result, and comparing the prediction result with the true value to obtain a classification error. The accuracy calculation formula of the classifier is as follows:
acc is the number of accurate classification/total number of samples
And repeating the training process, and after K rounds of training, selecting a group of models with the minimum classification errors as optimal classification models, namely, presetting a department prediction model for the department prediction of the patient.
The classifier adopts a multi-classifier based on a linear support vector machine. The linear support vector machine is a two-class classifier, and can divide two types of linear inseparable samples in a sample space into two types as accurately as possible. The formula for dividing the hyperplane is as follows:
WTX+b＝0
wherein, WTA normal vector representing a hyperplane; b denotes the intercept.
The classification criterion of the linear support vector machine is to find a partition hyperplane, so that the interval between heterogeneous points close to the hyperplane is as large as possible, the number of misclassified sample points is as small as possible, and the applied mathematical formula is as follows:
such that: y isi(wTxi+b)≥1-ξi(i＝1、2....m)
Wherein ξiIs a relaxation variable; c is a penalty coefficient; w is a weight vector.
In the actual classification problem, the samples are usually linearly inseparable, that is, a hyperplane cannot be found so that the samples are completely accurately classified into two types, so that the linear support vector machine introduces a slack variable to reduce the number of misclassified sample points. A larger C indicates a lower tolerance for misclassification.
Wherein, the calculation formula between the heterogeneous points close to the hyperplane is as follows:
the larger the value, the higher the accuracy of classification.
Furthermore, a plurality of linear support vector machines are combined to obtain the preset department prediction model of the application. In this embodiment, a one-to-one multi-classifier combination method is adopted, that is, a two-classifier is constructed between any two classes of samples, and K (K-1)/2 two-classifiers are required for K classes of samples. And voting for the samples of the unknown classes through a plurality of two classifiers, and dividing the samples into the classes with the largest votes.
S103, candidate doctors of the patient are determined according to the target department.
Specifically, all doctors in the target department may be determined as candidate doctors, or the candidate doctors may be determined according to the actual situation and the preset screening condition, for example, the candidate doctors of the patient may be screened according to the service time of the doctors and the visit time of the patient.
And S104, acquiring doctor data of each candidate doctor.
In detail, the doctor data includes doctor diagnosis and treatment characteristic data and data related to doctor's decision about patient's hospitalization. The physician encounter feature data is used to determine whether the physician is adept at diagnosing the condition of the patient. The data related to the doctor decision of the patient includes the degree of awareness of the hospital to which the doctor belongs, the job title of the doctor, the historical evaluation of the patient to the doctor, and the like, and is used for judging whether the doctor meets other medical needs of the patient.
S105, determining the doctor-patient similarity and the decision factor score of each candidate doctor according to the patient data and the doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a candidate physician's score for a physician characteristic relevant to the patient's decision to seek medical attention.
Specifically, a higher degree of similarity between doctors and patients indicates that doctors are better at diagnosing the patient's condition. A higher decision factor score indicates a better fit of the physician to the patient's other needs for the visit. The doctor-patient similarity and the decision factor score are integrated into the calculation of a patient matching doctor, so that the matching result is more accurate, and the requirements of the patient are better met.
And S106, determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor.
In detail, grid search is used for selecting weight factors of doctor-patient similarity and decision factor score, a final weight value is determined through parameter optimization, and the comprehensive score of the candidate doctor is obtained through calculation according to the weight value, the doctor-patient similarity and the decision factor score.
And S107, determining the recommended doctor of the patient according to the comprehensive score.
Specifically, first, all candidate doctors are ranked in the order of scores from high to low. Then, the doctor with the previous target ranking is selected as the recommended doctor of the patient, so that the system can select the target doctor of the patient from all the recommended doctors according to preset conditions.
Preferably, after determining the recommended doctor of the patient according to the composite score, the intelligent diagnosis guiding method for the remote medical platform of the embodiment further includes:
and acquiring the service time of all recommended doctors, and determining the target doctor of the patient according to the service time and the comprehensive score of the recommended doctors.
The embodiment adopts the technical scheme that the intelligent diagnosis guiding method for the remote medical platform comprises the following steps: acquiring patient data; determining a target department of the patient according to the patient data and a preset department prediction model; determining candidate doctors of the patient according to the target department; acquiring doctor data of each candidate doctor; determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is the score of the candidate doctor about doctor characteristics relevant to the patient hospitalization decision; determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor; and determining the recommended doctor of the patient according to the comprehensive score. Based on this, this application matches out the target department for the patient automatically according to patient data for the patient only needs to input relevant disease in the system can learn the department of seeing a doctor that corresponds, and the system matches the department for the patient automatically, matches the rate of accuracy height, has guaranteed that the patient selects the correct department. Secondly, the doctor-patient similarity and the decision factor score of each candidate doctor are determined according to the patient data and the doctor data, then the doctor-patient similarity and the decision factor score are matched with the doctor for the patient, the matching result is more accurate, the requirements of the patient are better met, high-quality medical resources are prevented from being excessively occupied, common medical resources are left unused, and the medical resource utilization rate is higher.
Preferably, the patient data comprises patient characteristic data. The patient characteristic data comprises patient personal information data and patient consultation texts; the patient personal information data includes the patient age and the patient sex.
Determining a target department of the patient according to the patient data and a preset department prediction model, comprising: firstly, a patient feature representation vector is determined according to patient feature data and a preset patient feature representation model. And then, determining a target department according to the patient characteristic expression vector and a preset department prediction model.
Specifically, the consultation text data is converted into an inquiry text vector through a text vector representation technology, and the symptom characteristics of the patient are extracted. And then combining the demographic characteristic data of the patient, mainly comprising two basic attributes of gender and age, distributing a certain weight to the gender, age and inquiry text vector, synthesizing, adjusting the weight parameter in the training process of the model, realizing the characteristic selection of the patient, and constructing the inquiry characteristic model of the patient.
Specifically, in this embodiment, the Doc2Vec (Word2Vec) model is used to convert the consulting text data into the consulting text vector, and the training method of the Doc2Vec model is as follows:
acquiring training data of a preset patient characteristic representation model; the training data comprises a plurality of patient data which are counted; patient data includes patient age, patient gender, and patient counseling text;
performing word segmentation processing on the consultation text;
marking paragraph labels on the word list of the word segmentation to obtain a training paragraph document;
and training the Doc2Vec model based on the training paragraph document to obtain the trained Doc2Vec model.
Specifically, after the patient characteristic data is obtained, the consulting text data is subjected to word segmentation processing to obtain word segmentation paragraph texts. Then, a vector representation of the word segmentation paragraph text, namely a query text vector, is obtained according to the Doc2Vec model. And finally, carrying out dummy variable processing on the gender data of the patient to obtain a gender expression vector. And performing box normalization processing on the ages of the patients to obtain age expression vectors. In one particular example, the patient age groups are primarily centered between 25-50 years of age, with one segment every 5 years. The specific age bins are those of < 25, [26,30), [30,35), [35,40), [40,45), [45,50, > 50. The user ages are converted into numerical values representing age groups corresponding to the numerical values of 1-8 respectively, then dimension differences are eliminated in the future, and the values in the boxes are normalized into numerical values between (0, 1).
The patient's age and gender have been converted into numbers in the range of 0-1, but the inquiry text vector of the patient is a high-dimensional vector, and in the subsequent classification process, if the inquiry text vector, the age representing vector and the gender representing vector are directly input, the high-dimensional inquiry text vector will play a main role, so that the role of gender and age is masked. Therefore, to get a better classification effect, the gender and age of the patient are expanded to a vector with the same dimension as the text features. For example, sex 1, expanded to a 200-dimensional vector [1,1, …,1], which is summed by linear weighting:
W＝k1*Wgender+k2*Wage+K3*Wtext
wherein W is a patient feature representation vector; k1, k2, and k3 are coefficients; wgender is a gender expression vector; wage is an age representation vector; wtext is a query text vector.
Wherein k1, k2, and k3 have a default value of 1. The patient feature representation vector is embedded in the department prediction and doctor-patient matching module as a preprocessing step. Therefore, the weights k1, k2, and k3 can be optimized using the objective function of the link concerned, so that the error between the model output and the real data is minimized, and the optimal results of the parameters k1, k2, and k3 are obtained.
Specifically, optimization adjustment is performed on k1, k2 and k3 in the formula through grid search and cross validation. Firstly, a group of candidate values are set for k1, k2 and k3, the candidate values are in a range of [0,1] by taking 0.1 as a step, and then various parameter combinations are exhausted by using a network search method to carry out model training. The training process of the model combines the department prediction module, namely, selects the values of k1, k2 and k3 by taking the fitting effect of the department prediction module as an objective function.
In the specific application process, firstly, training data of a preset department prediction model are divided into K sub-sample sets, one sub-sample set is selected as a verification set in each turn, in addition, K-1 sub-sample sets are used as training sets, the K-1 sub-sample sets are used as input, various parameters K1, K2 and K3 are combined to obtain a patient characteristic representation vector, then, the preset department prediction model is used for training, and then the verification data are sequentially input into the preset patient characteristic representation model and the preset department prediction model to obtain a prediction result. The set of parameters that predicts best in the round is summed up into one score. Repeating the K rounds, and obtaining the parameter combination with the most scores as the optimal parameter value. And storing the optimal parameters obtained by training to obtain a trained preset patient characteristic representation model. When a new patient needs to be subjected to department prediction and doctor recommendation, the trained values of k1, k2 and k3 are directly utilized to calculate a patient feature representation vector, and then the patient feature representation vector is input into a correlation module.
Preferably, the doctor data comprises medical characteristic data of a doctor and doctor characteristic data related to a medical decision of a patient. Determining doctor-patient similarity and decision factor scores of the candidate doctors according to the patient data and the doctor data of the candidate doctors, wherein the doctor-patient similarity and decision factor scores comprise the following steps:
determining the doctor-patient similarity of the candidate doctors according to the patient feature representation vector and the diagnosis and treatment feature data of the doctors;
judging whether the patient data contains doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient or not to obtain a preset result;
when the preset result is yes, determining the decision factor score of the candidate doctor according to the doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient and the doctor feature data related to the patient hospitalizing decision of the candidate doctor;
when the preset result is negative, acquiring training patient data from a preset database; the training patient data comprises training patient characteristic data and doctor characteristic data related to patient medical decision of a target historical medical doctor of the training patient;
determining the patient feature similarity of each training patient and the patient according to the training patient feature data and the patient feature data;
determining the target patient characteristic similarity with the maximum value from all the patient characteristic similarities;
defining doctor feature data related to patient medical decision of a target history doctor for training the patient corresponding to the target patient feature similarity as doctor feature data related to patient medical decision of the target history doctor for the patient;
determining a decision factor score for the candidate physician based on the patient hospitalization decision-related physician characteristic data for the patient historical attending physician and the patient hospitalization decision-related physician characteristic data for the candidate physician.
Preferably, determining the doctor-patient similarity of the candidate doctor according to the patient feature representation vector and the diagnosis and treatment feature data of the doctor comprises:
determining a doctor diagnosis and treatment characteristic representation vector according to diagnosis and treatment characteristic data of a doctor and a preset doctor diagnosis and treatment characteristic representation model;
and determining the similarity of the doctors and patients of the candidate doctors according to the patient characteristic representation vector and the doctor diagnosis and treatment characteristic representation vector.
Specifically, the clinical characteristic data includes patient data of patients that the target of the doctor has visited, for example, the patient data of 5 patients that the doctor has visited recently. From these patient data, it is possible to infer the condition that the physician is skilled in the treatment. And after the diagnosis and treatment characteristic data of the doctor are obtained, converting the diagnosis and treatment characteristic data of the doctor into diagnosis and treatment characteristic expression vectors of sub-doctors according to the preset patient characteristic expression model. And then forming a matrix by using all the sub-doctor diagnosis and treatment characteristic representation vectors to obtain a doctor diagnosis and treatment special diagnosis representation vector. And finally, calculating the cosine similarity between the diagnosis and treatment characteristic representation vector of the doctor and the characteristic representation vector of the patient to obtain the doctor-patient similarity.
Preferably, determining the decision factor score of the candidate doctor according to the doctor characteristic data related to the patient medical decision of the patient historical visit doctor and the doctor characteristic data related to the patient medical decision of the candidate doctor comprises:
determining a doctor decision factor characteristic representation vector according to doctor characteristic data related to the patient hospitalizing decision of the candidate doctor and a preset doctor decision factor characteristic representation model;
and determining the decision factor score of the candidate doctor according to the doctor characteristic data and the doctor decision factor characteristic expression vector related to the patient medical decision of the target historical doctor of the patient.
Specifically, after doctor feature data related to the decision of a candidate doctor for a patient to see a doctor is obtained, the integer variables larger than 0 are subjected to binning processing, and all variables are subjected to normalization processing to obtain a doctor decision factor feature expression vector. And then calculating cosine similarity between the doctor decision factor characteristic expression vector and doctor characteristic data related to the patient medical decision of the target historical doctor of the patient to obtain the decision factor score of the candidate doctor.
It should be noted that, the patient medical decision-related physician characteristic data of the target historical visit physicians of the patients may be patient medical decision-related physician characteristic data of 5 physicians visited recently by the patients, and is vector data.
Preferably, before acquiring the patient data, the intelligent diagnosis guiding method for the remote medical platform of the present embodiment further includes:
physician characteristics relevant to patient decision-making are determined.
Specifically, first, candidate variables for medical decision making are determined, including doctor's job title, hospital belonging, hospital class, inquiry price, rating rate, peer acceptance, mind number (feedback reward score of patient's sensibility to doctor's service expression, positive integer of 0-10000) and fan number. And then, cleaning preset data of the candidate variables of the medical decision to obtain the feature expression vectors of the candidate variables of the medical decision. The specific process of preset data cleaning comprises the following steps: and carrying out dummy variable processing on the doctor titles and the affiliated hospitals, and carrying out normalization processing on hospital levels. And finally, constructing a multiple linear regression model of the patient selected doctor by taking the doctor diagnosis and treatment feature expression vector as a target variable and taking the hospitalizing decision candidate variable as an independent variable. And analyzing by a multiple linear regression model to obtain the P value of each independent variable. The P value represents the significance level of the correlation between the independent and dependent variables. In the multiple linear regression model with the confidence coefficient of 95%, when the P value is greater than 0.05, the correlation between the independent variable and the dependent variable is not significant. And removing variables with low correlation, and reconstructing a multiple linear regression model. By the method, the doctor characteristics related to the hospitalizing decision of the patient are determined, wherein the doctor characteristics comprise inquiry price, favorable evaluation rate, mind number, peer acceptance and vermicelli number.
Fig. 2 is a schematic structural diagram of an intelligent diagnosis guiding apparatus for a remote medical platform according to an embodiment of the present invention. As shown in fig. 2, the intelligent diagnosis guiding apparatus for remote medical platform of the present embodiment includes: a patient data acquisition module 21, a department prediction module 22, a candidate doctor determination module 23, a doctor data acquisition module 24, a similarity matching module 25, a score calculation module 26, and a recommended doctor determination module 27.
The patient data acquisition module 21 is configured to acquire patient data; a department prediction module 22, configured to determine a target department of the patient according to the patient data and a preset department prediction model; a candidate doctor determining module 23, configured to determine candidate doctors of the patient according to the target department; a doctor data acquiring module 24, configured to acquire doctor data of each candidate doctor; a similarity matching module 25, configured to determine, according to the patient data and the doctor data of the candidate doctors, a doctor-patient similarity and a decision factor score of each candidate doctor; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a score of the candidate physician for a physician feature relevant to the patient's decision to seek medical attention; the score calculation module 26 is used for determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor; and a recommended doctor determining module 27, configured to determine a recommended doctor of the patient according to the composite score.
Preferably, the department prediction module 22 is specifically configured to determine a patient feature representation vector according to the patient feature data and a preset patient feature representation model, and determine the target department according to the patient feature representation vector and the preset department prediction model.
Preferably, the department prediction module 22 is further specifically configured to, first, perform preset word segmentation on the patient consultation text to obtain a word segmentation paragraph of the patient consultation text. Then, determining a patient consultation text representation vector according to the segmented paragraph patient consultation text; carrying out dummy variable processing on the sex of the patient to obtain a sex expression vector of the patient; and performing box normalization processing on the patient ages to obtain a patient age representation vector. Finally, a patient characteristic representation vector is determined based on the patient consultation text representation vector, the patient gender representation vector, and the patient age representation vector.
Preferably, the similarity matching module 25 is specifically configured to implement the following method:
determining the doctor-patient similarity of the candidate doctors according to the patient feature representation vector and the diagnosis and treatment feature data of the doctors;
judging whether the patient data contains doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient or not to obtain a preset result;
when the preset result is yes, determining the decision factor score of the candidate doctor according to the doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient and the doctor feature data related to the patient hospitalizing decision of the candidate doctor;
when the preset result is negative, acquiring training patient data from a preset database; the training patient data comprises training patient characteristic data and doctor characteristic data related to patient medical decision of a target historical medical doctor of the training patient;
determining the patient feature similarity of each training patient and the patient according to the training patient feature data and the patient feature data;
determining the target patient characteristic similarity with the maximum value from all the patient characteristic similarities;
defining doctor feature data related to patient medical decision of a target history doctor for training the patient corresponding to the target patient feature similarity as doctor feature data related to patient medical decision of the target history doctor for the patient;
determining a decision factor score for the candidate physician based on the patient hospitalization decision-related physician characteristic data for the patient historical attending physician and the patient hospitalization decision-related physician characteristic data for the candidate physician.
Preferably, the similarity matching module 25 is further configured to, first, determine a doctor diagnosis and treatment feature representation vector according to the doctor diagnosis and treatment feature data and a preset doctor diagnosis and treatment feature representation model. Then, the doctor-patient similarity of the candidate doctors is determined according to the patient characteristic representation vector and the doctor diagnosis and treatment characteristic representation vector.
Preferably, the similarity matching module 25 is further configured to, first, determine a physician decision factor feature representation vector according to the physician feature data related to the patient hospitalization decision of the candidate physician and a preset physician decision factor feature representation model. Then, the decision factor score of the candidate doctor is determined according to the doctor characteristic data related to the patient medical decision of the target historical doctor of the patient and the doctor decision factor characteristic representation vector.
Preferably, the intelligent diagnosis guiding apparatus for a remote medical platform of the present embodiment further includes a doctor characteristic determination module related to the patient medical decision, for determining doctor characteristics related to the patient medical decision.
Preferably, the doctor characteristic determination module related to the medical decision of the patient is used for determining the candidate variable of the medical decision firstly. And then, cleaning preset data of the candidate medical decision variables to obtain the candidate medical decision variable feature expression vectors. Finally, determining doctor characteristics related to the hospitalizing decision of the patient according to the doctor diagnosis and treatment characteristic representation vector and the hospitalizing decision candidate variable characteristic representation vector; the doctor characteristics related to the patient hospitalizing decision comprise inquiry price, favorable rating rate, mind number, peer acceptance and fan number.
Preferably, the intelligent diagnosis guiding apparatus for a remote medical platform of the present embodiment further includes a target doctor determining module, configured to obtain service times of all the recommended doctors, and then determine a target doctor of the patient according to the service time of the recommended doctor and the comprehensive score.
The present embodiment and the above embodiments are based on a general inventive concept, and have the same or corresponding implementation procedures and beneficial effects, which are not described herein again.
It is understood that the same or similar parts in the above embodiments may be mutually referred to, and the same or similar parts in other embodiments may be referred to for the content which is not described in detail in some embodiments.
It should be noted that the terms "first," "second," and the like in the description of the present invention are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. Further, in the description of the present invention, the meaning of "a plurality" means at least two unless otherwise specified.
Any process or method descriptions in flow diagrams or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing specific logical functions or steps of the process, and the scope of the preferred embodiments of the present invention includes additional implementations in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present invention.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
Claims (10)
1. An intelligent referral method for a telemedicine platform, comprising:
acquiring patient data;
determining a target department of the patient according to the patient data and a preset department prediction model;
determining candidate doctors of the patient according to the target department;
acquiring doctor data of each candidate doctor;
determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a score of the candidate physician for a physician feature relevant to the patient's decision to seek medical attention;
determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor;
and determining the recommended doctor of the patient according to the comprehensive score.
2. The intelligent referral method for a telemedicine platform of claim 1 wherein the patient data includes patient characteristic data;
the determining the target department of the patient according to the patient data and a preset department prediction model comprises:
determining a patient characteristic representation vector according to the patient characteristic data and a preset patient characteristic representation model;
and determining the target department according to the patient feature representation vector and the preset department prediction model.
3. The intelligent referral method for a telemedicine platform of claim 2 wherein the patient characteristic data includes patient personal information data and patient advisory text; the patient personal information data comprises the age and the sex of the patient;
the determining a patient feature representation vector according to the patient data and a preset patient feature representation model comprises:
performing preset word segmentation processing on the patient consultation text to obtain a word segmentation paragraph patient consultation text;
determining a patient consultation text representation vector according to the word segmentation paragraph patient consultation text;
carrying out dummy variable processing on the gender of the patient to obtain a gender expression vector of the patient;
performing box-dividing normalization processing on the patient age to obtain a patient age representation vector;
determining the patient characteristic representation vector based on the patient consultation text representation vector, the patient gender representation vector, and the patient age representation vector.
4. The intelligent referral method for a telemedicine platform of claim 2 wherein the physician data includes physician clinical characteristic data of a physician and physician characteristic data relating to patient decision-making;
determining doctor-patient similarity and decision factor scores of the candidate doctors according to the patient data and the doctor data of the candidate doctors, wherein the steps comprise:
determining the doctor-patient similarity of the candidate doctors according to the patient feature representation vector and the diagnosis and treatment feature data of the doctors;
judging whether the patient data contains doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient or not to obtain a preset result;
when the preset result is yes, determining the decision factor score of the candidate doctor according to the doctor feature data related to the patient hospitalizing decision of the target historical hospitalizing doctor of the patient and the doctor feature data related to the patient hospitalizing decision of the candidate doctor;
when the preset result is negative, acquiring training patient data from a preset database; the training patient data comprises training patient characteristic data and doctor characteristic data related to patient medical decision of a target historical medical doctor of the training patient;
determining the patient feature similarity of each training patient and the patient according to the training patient feature data and the patient feature data;
determining the target patient characteristic similarity with the maximum value from all the patient characteristic similarities;
defining doctor feature data related to patient medical decision of a target history doctor for training the patient corresponding to the target patient feature similarity as doctor feature data related to patient medical decision of the target history doctor for the patient;
determining a decision factor score for the candidate physician based on the patient hospitalization decision-related physician characteristic data for the patient historical attending physician and the patient hospitalization decision-related physician characteristic data for the candidate physician.
5. The intelligent diagnosis guiding method for the remote medical platform according to claim 4, wherein the determining the doctor-patient similarity of the candidate doctor according to the patient feature representation vector and the doctor diagnosis and treatment feature data comprises:
determining a doctor diagnosis and treatment characteristic representation vector according to the diagnosis and treatment characteristic data of the doctor and a preset doctor diagnosis and treatment characteristic representation model;
and determining the doctor-patient similarity of the candidate doctors according to the patient characteristic representation vector and the doctor diagnosis and treatment characteristic representation vector.
6. The intelligent referral method for a telemedicine platform of claim 4 wherein determining the candidate physician's decision factor score based on the patient encounter decision-related physician feature data of the patient's target history physician and the candidate physician's patient encounter decision-related physician feature data of the candidate physician comprises:
determining a doctor decision factor characteristic representation vector according to doctor characteristic data related to the patient hospitalizing decision of the candidate doctor and a preset doctor decision factor characteristic representation model;
and determining the decision factor score of the candidate doctor according to the doctor characteristic data relevant to the patient medical decision of the target historical doctor of the patient and the doctor decision factor characteristic representation vector.
7. The intelligent referral method for a telemedicine platform of claim 5, wherein prior to the acquiring patient data, further comprising:
physician characteristics relevant to patient decision-making are determined.
8. The intelligent referral method for a telemedicine platform of claim 7 wherein the determining patient hospitalization decision-related physician features comprises:
determining candidate variables for medical decision-making;
cleaning preset data of the medical decision candidate variables to obtain medical decision candidate variable feature expression vectors;
determining doctor features related to the hospitalizing decision of the patient according to the doctor diagnosis and treatment feature representation vector and the hospitalizing decision candidate variable feature representation vector; the doctor characteristics related to the patient hospitalizing decision comprise inquiry price, favorable rating rate, mind number, peer acceptance and fan number.
9. The intelligent referral method for a telemedicine platform of claim 7, further comprising, after determining the patient's recommended physician based on the composite score:
acquiring the service time of all the recommended doctors;
and determining the target doctor of the patient according to the service time of the recommended doctor and the comprehensive score.
10. An intelligent referral device for a telemedicine platform, comprising:
a patient data acquisition module for acquiring patient data;
the department prediction module is used for determining a target department of the patient according to the patient data and a preset department prediction model;
a candidate doctor determining module for determining candidate doctors of the patient according to the target department;
a doctor data acquisition module for acquiring doctor data of each candidate doctor;
the similarity matching module is used for determining doctor-patient similarity and decision factor score of each candidate doctor according to the patient data and the doctor data of the candidate doctors; the doctor-patient similarity is the similarity between the patient characteristics of the patient and the diagnosis and treatment characteristics of the candidate doctor; the decision factor score is a score of the candidate physician for a physician feature relevant to the patient's decision to seek medical attention;
the score calculation module is used for determining the comprehensive score of each candidate doctor according to the doctor-patient similarity and the decision factor score of the candidate doctor;
and the recommended doctor determining module is used for determining the recommended doctor of the patient according to the comprehensive score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110865124.6A CN113539460A (en) | 2021-07-29 | 2021-07-29 | Intelligent diagnosis guiding method and device for remote medical platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110865124.6A CN113539460A (en) | 2021-07-29 | 2021-07-29 | Intelligent diagnosis guiding method and device for remote medical platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113539460A true CN113539460A (en) | 2021-10-22 |
Family
ID=78121535
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110865124.6A Pending CN113539460A (en) | 2021-07-29 | 2021-07-29 | Intelligent diagnosis guiding method and device for remote medical platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113539460A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114220537A (en) * | 2022-02-18 | 2022-03-22 | 橙意家人科技（天津）有限公司 | AI intelligent online diagnosis method based on Internet hospital and cloud system |
| CN115019948A (en) * | 2022-06-28 | 2022-09-06 | 苏州景昱医疗器械有限公司 | Control device, remote consultation system, and computer-readable storage medium |
| CN116737940A (en) * | 2023-08-14 | 2023-09-12 | 成都飞航智云科技有限公司 | Intelligent decision method and decision system |
| CN117133448A (en) * | 2023-10-23 | 2023-11-28 | 长春中医药大学 | Gynecological disease consultation system based on big data |
| CN118098535A (en) * | 2024-04-16 | 2024-05-28 | 旭辉卓越健康信息科技有限公司 | Visual medical procedure active recommendation method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108039198A (en) * | 2017-12-11 | 2018-05-15 | 重庆邮电大学 | A kind of doctor towards portable medical recommends method and system |
| CN108922608A (en) * | 2018-06-13 | 2018-11-30 | 平安医疗科技有限公司 | Intelligent hospital guide's method, apparatus, computer equipment and storage medium |
| CN109411082A (en) * | 2018-11-08 | 2019-03-01 | 西华大学 | A kind of Evaluation of Medical Quality and medical recommended method |
| CN110910998A (en) * | 2019-12-05 | 2020-03-24 | 北京明略软件系统有限公司 | Method and device for recommending doctor according to patient condition information |
| CN110993081A (en) * | 2019-12-03 | 2020-04-10 | 济南大学 | Doctor online recommendation method and system |
| CN112614578A (en) * | 2020-12-29 | 2021-04-06 | 平安国际智慧城市科技股份有限公司 | Doctor intelligent recommendation method and device, electronic equipment and storage medium |
-
2021
- 2021-07-29 CN CN202110865124.6A patent/CN113539460A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108039198A (en) * | 2017-12-11 | 2018-05-15 | 重庆邮电大学 | A kind of doctor towards portable medical recommends method and system |
| CN108922608A (en) * | 2018-06-13 | 2018-11-30 | 平安医疗科技有限公司 | Intelligent hospital guide's method, apparatus, computer equipment and storage medium |
| CN109411082A (en) * | 2018-11-08 | 2019-03-01 | 西华大学 | A kind of Evaluation of Medical Quality and medical recommended method |
| CN110993081A (en) * | 2019-12-03 | 2020-04-10 | 济南大学 | Doctor online recommendation method and system |
| CN110910998A (en) * | 2019-12-05 | 2020-03-24 | 北京明略软件系统有限公司 | Method and device for recommending doctor according to patient condition information |
| CN112614578A (en) * | 2020-12-29 | 2021-04-06 | 平安国际智慧城市科技股份有限公司 | Doctor intelligent recommendation method and device, electronic equipment and storage medium |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114220537A (en) * | 2022-02-18 | 2022-03-22 | 橙意家人科技（天津）有限公司 | AI intelligent online diagnosis method based on Internet hospital and cloud system |
| CN115019948A (en) * | 2022-06-28 | 2022-09-06 | 苏州景昱医疗器械有限公司 | Control device, remote consultation system, and computer-readable storage medium |
| CN115019948B (en) * | 2022-06-28 | 2023-06-16 | 苏州景昱医疗器械有限公司 | Control device, remote consultation system, and computer-readable storage medium |
| CN116737940A (en) * | 2023-08-14 | 2023-09-12 | 成都飞航智云科技有限公司 | Intelligent decision method and decision system |
| CN116737940B (en) * | 2023-08-14 | 2023-11-07 | 成都飞航智云科技有限公司 | Intelligent decision method and decision system |
| CN117133448A (en) * | 2023-10-23 | 2023-11-28 | 长春中医药大学 | Gynecological disease consultation system based on big data |
| CN117133448B (en) * | 2023-10-23 | 2024-01-23 | 长春中医药大学 | Gynecological disease consultation system based on big data |
| CN118098535A (en) * | 2024-04-16 | 2024-05-28 | 旭辉卓越健康信息科技有限公司 | Visual medical procedure active recommendation method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113539460A (en) | Intelligent diagnosis guiding method and device for remote medical platform | |
| CN109411082B (en) | Medical quality evaluation and treatment recommendation method | |
| Wang et al. | Risk assessment of coronary heart disease based on cloud-random forest | |
| Fernandes et al. | Predicting Intensive Care Unit admission among patients presenting to the emergency department using machine learning and natural language processing | |
| Gadekallu et al. | Cuckoo search optimized reduction and fuzzy logic classifier for heart disease and diabetes prediction | |
| US20180150609A1 (en) | Server and method for predicting future health trends through similar case cluster based prediction models | |
| US10593431B1 (en) | Methods and systems for causative chaining of prognostic label classifications | |
| WO2020102175A1 (en) | Medical treatment metric modelling based on machine learning | |
| Mokeddem | A fuzzy classification model for myocardial infarction risk assessment | |
| CN101443780A (en) | Methods, system, and computer program products for developing and using predictive models for predicting a plurality of medical outcomes, for evaluating intervention strategies, and for simultaneously | |
| US12020818B1 (en) | Cross care matrix based care giving intelligence | |
| CN111091907A (en) | Health medical knowledge retrieval method and system based on similar case library | |
| US11710572B2 (en) | Experience engine-method and apparatus of learning from similar patients | |
| Dhar | An adaptive intelligent diagnostic system to predict early stage of parkinson's disease using two-stage dimension reduction with genetically optimized lightgbm algorithm | |
| CA3218611A1 (en) | System and method for text-based conversation with a user, using machine learning | |
| Molla et al. | A predictive analysis framework of heart disease using machine learning approaches | |
| Yu et al. | A general framework of nonparametric feature selection in high-dimensional data | |
| Songram et al. | A study of features affecting on stroke prediction using machine learning | |
| WO2023217737A1 (en) | Health data enrichment for improved medical diagnostics | |
| Khaleghi et al. | A tree based approach for multi-class classification of surgical procedures using structured and unstructured data | |
| Mansouri et al. | A hybrid machine learning approach for early mortality prediction of ICU patients | |
| Khademolqorani et al. | Development of a decision support system for handling health insurance deduction | |
| AU2021102593A4 (en) | A Method for Detection of a Disease | |
| Choi et al. | Explainable machine learning to bring database to the bedside: development and validation of the TROUT (Trauma fRailty OUTcomes) Index, a point-of-care tool to prognosticate outcomes after traumatic injury based on frailty | |
| Bahanshal et al. | Hybrid fuzzy weighted K-Nearest neighbor to predict hospital readmission for diabetic patients |
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 |