CN112017784A - Coronary heart disease risk prediction method based on multi-modal data and related equipment - Google Patents
Coronary heart disease risk prediction method based on multi-modal data and related equipment Download PDFInfo
- Publication number
- CN112017784A CN112017784A CN202011142479.4A CN202011142479A CN112017784A CN 112017784 A CN112017784 A CN 112017784A CN 202011142479 A CN202011142479 A CN 202011142479A CN 112017784 A CN112017784 A CN 112017784A
- Authority
- CN
- China
- Prior art keywords
- data
- patient
- intervention
- waveform data
- heart disease
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
- 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
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
- G06F16/353—Clustering; Classification into predefined classes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
Abstract
The embodiment of the invention relates to the technical field of medical treatment, and discloses a coronary heart disease risk prediction method based on multi-modal data and related equipment, wherein a processor of the equipment is used for executing the following steps: acquiring cardiac ultrasound reports, waveform data and intervention data for each patient in the sample set prior to a first target time; fusing the determined first characterization vector of the cardiac ultrasonic report and the second characterization vector of the waveform data to obtain a first target characterization vector; inputting the first target characterization vector and a second target characterization vector of the intervention data into a specified classification model for training to obtain a coronary heart disease risk prediction model; and inputting the text data, the waveform data and the intervention data of the patient to be tested at the second target moment into the coronary heart disease risk prediction model to obtain the third moment and the third risk probability of the patient to be tested suffering from the coronary heart disease and the fourth moment and the fourth risk probability of death, so that the prediction effect of the coronary heart disease is improved. The present invention relates to a block chain technique, and the data can be stored in the block chain.
Description
Technical Field
The invention relates to the technical field of medical treatment, in particular to a coronary heart disease risk prediction method based on multi-modal data and related equipment.
Background
Currently, Coronary heart disease can be determined by cardiac ultrasound reports acquired by a cardiac monitoring device in a Coronary heart disease monitoring Unit (CCU), which is a monitoring ward monitoring Coronary heart disease. Currently, the risk of coronary heart disease in patients is mainly determined by medical personnel analyzing cardiac ultrasound reports, or the risk of death in patients with coronary heart disease is analyzed. At present, an intelligent coronary heart disease prediction system exists, however, the system can only process a single type and a small amount of data, and a good prediction result is difficult to obtain for large-scale multi-modal data. Therefore, how to better predict coronary heart disease is very important.
Disclosure of Invention
The embodiment of the invention provides a coronary heart disease risk prediction method based on multi-mode data and related equipment, and by analyzing the multi-mode data of three dimensions of a cardiac ultrasonic report, waveform data and intervention data, the time and risk probability of coronary heart disease suffering of a patient and the time and risk probability of death of the patient can be predicted more efficiently and quickly, and the prediction effect of the coronary heart disease is improved.
In a first aspect, an embodiment of the present invention provides a coronary heart disease risk prediction apparatus based on multi-modal data, where the apparatus includes: a memory and a processor;
the memory to store program instructions;
the processor, configured to invoke the program instructions, and when the program instructions are executed, configured to:
acquiring a cardiac ultrasound report of each patient in a sample set before a first target time, extracting text data from the cardiac ultrasound report, and determining a first characterization vector of the text data;
obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data;
performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data;
inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a coronary heart disease risk prediction model, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient;
and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained coronary heart disease risk prediction model to obtain a third moment and a third risk probability of coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
Further, the text data carries a first tag and a second tag; when the processor determines the first characterization vector of the text data, the processor is specifically configured to:
inputting the text data into a first convolution neural network model to obtain a word vector corresponding to the text data;
and determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, wherein the first label is used for indicating whether the patient suffers from coronary heart disease, and the second label is used for indicating whether the patient dies.
Further, when determining, by the processor, a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label, and the second label, the processor is specifically configured to:
inputting the word vector and the first label into a first classification model to obtain a first vector corresponding to the first label in the text data;
inputting the word vector and the second label into a second classification model to obtain a second vector corresponding to the second label in the text data;
determining a first characterization vector corresponding to the first label and the second label of the text data according to the first vector and the second vector.
Further, the waveform data includes high frequency waveform data and low frequency waveform data; when the processor determines the second characterization vector of the waveform data, the processor is specifically configured to:
acquiring high-frequency waveform data and low-frequency waveform data from the waveform data according to the sampling frequency;
carrying out dimensionality reduction processing on the high-frequency waveform data to obtain dimensionality reduction waveform data taking hours as a unit;
and determining a second characterization vector of the waveform data according to the dimension reduction waveform data and the low-frequency waveform data.
Further, when the processor determines the second characterization vector of the waveform data according to the dimension-reduced waveform data and the low-frequency waveform data, the processor is specifically configured to:
performing feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data;
extracting the characteristics of the low-frequency waveform data to obtain a vector of the low-frequency waveform data;
and determining a second characterization vector of the waveform data according to the vector of the dimensionality reduction waveform data and the vector of the low-frequency waveform data.
Further, when the processor determines, according to the intervention data, a second target characterization vector corresponding to the intervention data, the processor is specifically configured to:
determining an intervention time corresponding to the intervention data from the intervention data, wherein the intervention data comprises one or more of medication data, examination data, procedure data for each patient in the sample set prior to the first target time;
acquiring text data and waveform data corresponding to the cardiac ultrasonic report at the intervention time;
and determining a second target characterization vector corresponding to the intervention data according to the text data and the waveform data corresponding to the cardiac ultrasonic report corresponding to the intervention time.
Further, when the processor determines the intervention time corresponding to the intervention data according to the intervention data, the processor is specifically configured to:
determining an intervention indication vector corresponding to the intervention data, the intervention indication vector indicating whether each patient in the sample set has intervened at a time prior to the first target time;
and determining the intervention time of each patient in the sample set according to the intervention indication vector.
In a second aspect, an embodiment of the present invention provides a coronary heart disease risk prediction method based on multi-modal data, including:
acquiring a cardiac ultrasound report of each patient in a sample set before a first target time, extracting text data from the cardiac ultrasound report, and determining a first characterization vector of the text data;
obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data;
performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data;
inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient;
and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
In a third aspect, an embodiment of the present invention provides a coronary heart disease risk prediction apparatus based on multi-modal data, including:
the acquisition unit is used for acquiring a cardiac ultrasonic report of each patient in a sample set before a first target moment, extracting text data from the cardiac ultrasonic report and determining a first characterization vector of the text data;
a first determining unit, configured to obtain waveform data of each patient in the sample set before the first target time, and determine a second characterization vector of the waveform data;
the fusion unit is used for carrying out fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
a second determining unit, configured to obtain intervention data of each patient in the sample set before the first target time, and determine a second target characterization vector corresponding to the intervention data according to the intervention data;
the processing unit is used for inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient;
and the prediction unit is used for acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease, and obtaining a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
In a fourth aspect, the present invention provides a computer-readable storage medium, which stores a computer program, wherein the computer program is executed by a processor to implement the method of the second aspect.
According to the embodiment of the invention, a cardiac ultrasonic report of each patient in a sample set before a first target moment can be obtained, text data is extracted from the cardiac ultrasonic report, and a first characterization vector of the text data is determined; obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data; performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing; acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data; inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient; and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested. By analyzing the multi-mode data of three dimensions of the cardiac ultrasonic report, the waveform data and the intervention data, the time and the risk probability of the patient suffering from the coronary heart disease and the time and the risk probability of the patient dying can be predicted more efficiently and quickly, and the prediction effect of the coronary heart disease is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic structural diagram of a coronary heart disease prediction system according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a coronary heart disease risk prediction method based on multi-modal data according to an embodiment of the present invention;
fig. 3 is a schematic block diagram of a coronary heart disease risk prediction apparatus based on multi-modal data according to an embodiment of the present invention;
fig. 4 is a schematic block diagram of a coronary heart disease risk prediction apparatus based on multi-modal data according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The coronary heart disease risk prediction method based on multi-modal data provided by the embodiment of the invention can be applied to a coronary heart disease prediction system, and in some embodiments, the coronary heart disease prediction system comprises a medical server and a coronary heart disease risk prediction device based on multi-modal data. In some embodiments, the medical server may establish a communication connection with a coronary heart disease risk prediction device based on multimodal data. In some embodiments, the manner of the Communication connection may include, but is not limited to, Wi-Fi, Bluetooth, Near Field Communication (NFC), and the like. In certain embodiments, the medical server is configured to store monitoring data for a patient, wherein the monitoring data includes cardiac ultrasound reports, waveform data, and intervention data. In some embodiments, the cardiac ultrasound report is data collected by a cardiac ultrasound test device, which sends the collected data to a medical server for storage. In some embodiments, the waveform data is acquired by a coronary heart disease monitoring device, and the coronary heart disease monitoring device sends the acquired waveform data to the medical server for storage. In some embodiments, the intervention data includes, but is not limited to, medication data, examination data, surgery, etc. other than cardiac ultrasound reports and waveform data, which each medical device sends to the medical server.
The coronary heart disease prediction system provided by the embodiment of the invention is schematically illustrated with reference to fig. 1.
Referring to fig. 1, fig. 1 is a schematic structural diagram of a coronary heart disease prediction system according to an embodiment of the present invention. The coronary heart disease prediction system comprises: a coronary heart disease risk prediction device 11 based on multi-modal data and a medical server 12. In some embodiments, the coronary heart disease risk prediction device 11 and the medical server 12 based on the multi-modal data may establish a communication connection through a wireless communication connection; in some scenarios, the coronary heart disease risk prediction device 11 based on the multi-modal data and the medical server 12 may also establish a communication connection through a wired communication connection. In some embodiments, the coronary heart disease risk prediction device 11 based on multi-modal data may include, but is not limited to, a smart terminal device such as a smart phone, a tablet computer, a notebook computer, a desktop computer, and the like.
In an embodiment of the present invention, the coronary heart disease risk prediction device 11 based on multi-modal data may obtain a sample set from the medical server 12, wherein the training set includes cardiac ultrasound reports, waveform data and intervention data of a plurality of patients before the first target time. The coronary heart disease risk prediction device 11 based on multi-modal data may extract text data from the cardiac ultrasound report and determine a first characterization vector of the text data and determine a second characterization vector of the waveform data. The coronary heart disease risk prediction device 11 based on the multi-modal data may perform fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing; and obtaining intervention data of each patient in the sample set before the first target time, and determining a second target characterization vector corresponding to the intervention data according to the intervention data. The coronary heart disease risk prediction device 11 based on multi-modal data may input the first target characterization vector and the second target characterization vector into a designated classification model to obtain a classification result, and train the designated classification model according to the classification result to obtain a coronary heart disease risk prediction model, where the classification result includes a first time and a first risk probability that a patient suffers from coronary heart disease and a second time and a second risk probability that the patient dies. The coronary heart disease risk prediction device 11 based on the multi-modal data may obtain text data, waveform data, and intervention data of the cardiac ultrasound report of the patient to be tested at the second target time, and input the text data, the waveform data, and the intervention data into a trained coronary heart disease risk prediction model to obtain a third time and a third risk probability of coronary heart disease attack and a fourth time and a fourth risk probability of death of the patient to be tested.
By means of the method for analyzing the multi-mode data of three dimensions of the cardiac ultrasonic report, the waveform data and the intervention data, the time and the risk probability of coronary heart disease of a patient and the time and the risk probability of death of the patient can be predicted more efficiently and rapidly, and the prediction effect of the coronary heart disease is improved.
The coronary heart disease risk prediction method based on multi-modal data provided by the embodiment of the invention is schematically illustrated with reference to fig. 2.
Referring to fig. 2, fig. 2 is a schematic flow chart of a coronary heart disease risk prediction method based on multi-modal data according to an embodiment of the present invention, and as shown in fig. 2, the method may be executed by a coronary heart disease risk prediction device based on multi-modal data, and the specific explanation of the coronary heart disease risk prediction device based on multi-modal data is as described above and is not repeated here. Specifically, the method of the embodiment of the present invention includes the following steps.
S201: a cardiac ultrasound report is acquired for each patient in a sample set prior to a first target time instant, and textual data is extracted from the cardiac ultrasound report, and a first characterization vector for the textual data is determined.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can acquire a cardiac ultrasonic report of each patient in a sample set before a first target moment, extract text data from the cardiac ultrasonic report, and determine a first characterization vector of the text data. In some embodiments, the first target time is any time before the current time.
In one embodiment, a coronary heart disease risk prediction device based on multimodal data may obtain cardiac ultrasound reports for each patient in a sample set within a specified time frame prior to a first target time. For example, assuming that the first target time is time t and the specified time range is 24 hours, the coronary heart disease risk prediction device based on multi-modal data may obtain a cardiac ultrasound report for each patient in the sample set 24 hours before time t.
In one embodiment, the text data carries a first tag and a second tag; when determining a first characterization vector of the text data, the coronary heart disease risk prediction device based on multi-modal data can input the text data into a first convolution neural network model to obtain a word vector corresponding to the text data; and determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, wherein the first label is used for indicating whether the patient suffers from coronary heart disease, and the second label is used for indicating whether the patient dies. In certain embodiments, the first label and the second label are not the same, the first label including but not limited to numbers, letters, etc., and the second label including but not limited to numbers, letters, etc. For example, a first label of 1 is used to indicate coronary heart disease, a first label of 0 is used to indicate no coronary heart disease, a second label of 11 is used to indicate patient death, and a second label of 00 is used to indicate no death.
In one embodiment, when determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, the coronary heart disease risk prediction device based on multi-modal data may input the word vector and the first label into a first classification model to obtain a first vector corresponding to the first label in the text data; inputting the word vector and the second label into a second classification model to obtain a second vector corresponding to the second label in the text data; determining a first characterization vector corresponding to the first label and the second label of the text data according to the first vector and the second vector.
In this way, a first characterization vector corresponding to the first tag and the second tag can be determined from the text data corresponding to the cardiac ultrasound report.
S202: waveform data for each patient in the sample set prior to the first target time is obtained and a second characterization vector for the waveform data is determined.
In the embodiment of the present invention, the coronary heart disease risk prediction device based on multi-modal data may obtain waveform data of each patient in the sample set before the first target time, and determine a second characterization vector of the waveform data.
In one embodiment, the waveform data includes high frequency waveform data and low frequency waveform data; when determining a second characterization vector of the waveform data, the coronary heart disease risk prediction device based on multi-modal data can acquire high-frequency waveform data and low-frequency waveform data from the waveform data according to the sampling frequency; carrying out dimensionality reduction processing on the high-frequency waveform data to obtain dimensionality reduction waveform data taking hours as a unit; and determining a second characterization vector of the waveform data according to the dimension reduction waveform data and the low-frequency waveform data. In some embodiments, the high-frequency waveform data is waveform data sampled at a high frequency, and the low-frequency waveform data is waveform data sampled at a low frequency and fixed time, wherein the low-frequency waveform data is data collected in hours.
In one embodiment, when determining the second characterization vector of the waveform data according to the dimensionality reduction waveform data and the low-frequency waveform data, the coronary heart disease risk prediction device based on multi-modal data may perform feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data; extracting the characteristics of the low-frequency waveform data to obtain a vector of the low-frequency waveform data; and determining a second characterization vector of the waveform data according to the vector of the dimensionality reduction waveform data and the vector of the low-frequency waveform data.
In an embodiment, when the coronary heart disease risk prediction device based on multi-modal data performs feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data, the dimensionality reduction waveform data can be input into a second convolution neural network model to obtain the vector of the dimensionality reduction waveform data.
S203: and performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can perform fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing.
In an embodiment, when the first token vector and the second token vector are fused to obtain a fused first target token vector, the coronary heart disease risk prediction device based on multi-modal data may input the first token vector and the second token vector into a recurrent neural network model to fuse the first token vector and the second token vector to obtain the first target token vector.
In one embodiment, when the first token vector and the second token vector are fused, the coronary heart disease risk prediction device based on multi-modal data may use a dual attention mechanism to perform weighted summation on the first token vector and the second token vector to obtain the first target token vector.
In one embodiment, the coronary heart disease risk prediction device based on multi-modal data may generate two attention weights for each time instant, which are α and β respectively, according to the first token vector and the second token vector of each time instant within a specified range before the first target time instant, and then perform weighted summation on the two attention weights α and β and the first token vector and the second token vector of each time instant to obtain a first target token vector. For example, assuming that the first token vector at a certain time is a and the second token vector is B, the first target token vector at the time is α a + β B.
S204: intervention data of each patient in the sample set before the first target moment is obtained, and a second target characterization vector corresponding to the intervention data is determined according to the intervention data.
In the embodiment of the present invention, the coronary heart disease risk prediction device based on multi-modal data may obtain intervention data of each patient in the sample set before the first target time, and determine a second target characterization vector corresponding to the intervention data according to the intervention data.
In one embodiment, the coronary heart disease risk prediction device based on multi-modal data may determine an intervention time corresponding to the intervention data from the intervention data when determining a second target characterization vector corresponding to the intervention data from the intervention data, wherein the intervention data comprises one or more of medication data, examination data, surgical data for each patient in the sample set prior to the first target time; acquiring text data and waveform data corresponding to the cardiac ultrasonic report at the intervention time; and determining a second target characterization vector corresponding to the intervention data according to the text data and the waveform data corresponding to the cardiac ultrasonic report corresponding to the intervention time.
In one embodiment, the coronary heart disease risk prediction device based on multimodal data, when determining an intervention instant corresponding to the intervention data from the intervention data, may determine an intervention indication vector corresponding to the intervention data, the intervention indication vector indicating whether each patient in the sample set intervenes at a time before the first target time; and determining the intervention time of each patient in the sample set according to the intervention indication vector. For example, assuming that the intervention indication vector is an indication vector of an intravenous injection, it is marked as a if an intravenous injection is performed at a certain time, otherwise it is marked as b.
S205: inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a coronary heart disease risk prediction model, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can input the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and train the specified classification model according to the classification result to obtain the coronary heart disease risk prediction model, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient. In some embodiments, the first time and the second time are greater than the first target time, for example, assuming that the first target time is time t, the first time may be time t +1, and the second time may be time t + 2. In some embodiments, the first time and the second time may be the same or different.
In one embodiment, when the coronary heart disease risk prediction device based on multi-modal data trains the specified classification model according to the classification result to obtain the coronary heart disease risk prediction model, the first time and the first risk probability of coronary heart disease suffered by each patient and the second time and the second risk probability of death of each patient obtained in the classification result may be compared with the first label and the second label of each patient in the sample set, and the specified classification model is trained according to the comparison result to obtain the coronary heart disease risk prediction model.
S206: and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained coronary heart disease risk prediction model to obtain a third moment and a third risk probability of coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can acquire text data, waveform data and intervention data of a cardiac ultrasonic report of a patient to be tested at a second target moment, and input the text data, the waveform data and the intervention data into a trained coronary heart disease risk prediction model to obtain a third moment and a third risk probability of the patient to be tested suffering from the coronary heart disease and a fourth moment and a fourth risk probability of death. In some embodiments, the third time and the fourth time are greater than the second target time, for example, assuming that the second target time is the current time n, the first time may be time n +1, and the second time may be time n + 2; in some embodiments, the third time and the fourth time may be the same or different.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can acquire a cardiac ultrasonic report of each patient in a sample set before a first target moment, extract text data from the cardiac ultrasonic report, and determine a first characterization vector of the text data; obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data; performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing; acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data; inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient; and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested. By analyzing the multi-mode data of three dimensions of the cardiac ultrasonic report, the waveform data and the intervention data, the time and the risk probability of the patient suffering from the coronary heart disease and the time and the risk probability of the patient dying can be predicted more efficiently and quickly, and the prediction effect of the coronary heart disease is improved.
The embodiment of the invention also provides a coronary heart disease risk prediction device based on multi-modal data, which is used for executing the unit of the method. Specifically, referring to fig. 3, fig. 3 is a schematic block diagram of a coronary heart disease risk prediction apparatus based on multi-modal data according to an embodiment of the present invention. The coronary heart disease risk prediction device based on multi-modal data of the embodiment comprises: an acquisition unit 301, a first determination unit 302, a fusion unit 303, a second determination unit 304, a processing unit 305, and a prediction unit 306.
An obtaining unit 301, configured to obtain a cardiac ultrasound report of each patient in a sample set before a first target time, extract text data from the cardiac ultrasound report, and determine a first characterization vector of the text data;
a first determining unit 302, configured to obtain waveform data of each patient in the sample set before the first target time and determine a second characterization vector of the waveform data;
a fusion unit 303, configured to perform fusion processing on the first token vector and the second token vector to obtain a first target token vector after the fusion processing;
a second determining unit 304, configured to obtain intervention data of each patient in the sample set before the first target time, and determine a second target characterization vector corresponding to the intervention data according to the intervention data;
the processing unit 305 is configured to input the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and train the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, where the classification result includes a first time and a first risk probability that the patient suffers from the coronary heart disease and a second time and a second risk probability that the patient dies;
the prediction unit 306 is configured to obtain text data, waveform data, and intervention data of the cardiac ultrasound report of the patient to be tested at the second target time, and input the text data, the waveform data, and the intervention data into a trained risk prediction model of the coronary heart disease, so as to obtain a third time and a third risk probability of coronary heart disease suffered by the patient to be tested, and a fourth time and a fourth risk probability of death by the patient to be tested.
Further, the text data carries a first tag and a second tag; when determining the first characterization vector of the text data, the obtaining unit 301 is specifically configured to:
inputting the text data into a first convolution neural network model to obtain a word vector corresponding to the text data;
and determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, wherein the first label is used for indicating whether the patient suffers from coronary heart disease, and the second label is used for indicating whether the patient dies.
Further, when the obtaining unit 301 determines, according to the word vector, the first tag, and the second tag, a first characterization vector corresponding to the first tag and the second tag of the text data, specifically configured to:
inputting the word vector and the first label into a first classification model to obtain a first vector corresponding to the first label in the text data;
inputting the word vector and the second label into a second classification model to obtain a second vector corresponding to the second label in the text data;
determining a first characterization vector corresponding to the first label and the second label of the text data according to the first vector and the second vector.
Further, the waveform data includes high frequency waveform data and low frequency waveform data; when the first determining unit 302 determines the second characterization vector of the waveform data, it is specifically configured to:
acquiring high-frequency waveform data and low-frequency waveform data from the waveform data according to the sampling frequency;
carrying out dimensionality reduction processing on the high-frequency waveform data to obtain dimensionality reduction waveform data taking hours as a unit;
and determining a second characterization vector of the waveform data according to the dimension reduction waveform data and the low-frequency waveform data.
Further, when the first determining unit 302 determines the second characterization vector of the waveform data according to the dimension-reduced waveform data and the low-frequency waveform data, it is specifically configured to:
performing feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data;
extracting the characteristics of the low-frequency waveform data to obtain a vector of the low-frequency waveform data;
and determining a second characterization vector of the waveform data according to the vector of the dimensionality reduction waveform data and the vector of the low-frequency waveform data.
Further, when the second determining unit 304 determines, according to the intervention data, a second target characterization vector corresponding to the intervention data, specifically configured to:
determining an intervention time corresponding to the intervention data from the intervention data, wherein the intervention data comprises one or more of medication data, examination data, procedure data for each patient in the sample set prior to the first target time;
acquiring text data and waveform data corresponding to the cardiac ultrasonic report at the intervention time;
and determining a second target characterization vector corresponding to the intervention data according to the text data and the waveform data corresponding to the cardiac ultrasonic report corresponding to the intervention time.
Further, when the second determining unit 304 determines the intervention time corresponding to the intervention data according to the intervention data, specifically, it is configured to:
determining an intervention indication vector corresponding to the intervention data, the intervention indication vector indicating whether each patient in the sample set has intervened at a time prior to the first target time;
and determining the intervention time of each patient in the sample set according to the intervention indication vector.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can acquire a cardiac ultrasonic report of each patient in a sample set before a first target moment, extract text data from the cardiac ultrasonic report, and determine a first characterization vector of the text data; obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data; performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing; acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data; inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient; and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested. By analyzing the multi-mode data of three dimensions of the cardiac ultrasonic report, the waveform data and the intervention data, the time and the risk probability of the patient suffering from the coronary heart disease and the time and the risk probability of the patient dying can be predicted more efficiently and quickly, and the prediction effect of the coronary heart disease is improved.
Referring to fig. 4, fig. 4 is a schematic block diagram of a coronary heart disease risk prediction apparatus based on multi-modal data according to an embodiment of the present invention. The apparatus in this embodiment as shown in the figure may comprise: one or more processors 401 and memory 402. The memory 402 is used for storing computer programs, including programs, and the processor 401 is used for executing the programs stored in the memory 402. Wherein the processor 401 is configured to invoke the program to perform:
acquiring a cardiac ultrasound report of each patient in a sample set before a first target time, extracting text data from the cardiac ultrasound report, and determining a first characterization vector of the text data;
obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data;
performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data;
inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a coronary heart disease risk prediction model, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient;
and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained coronary heart disease risk prediction model to obtain a third moment and a third risk probability of coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
Further, the text data carries a first tag and a second tag; when the processor 401 determines the first characterization vector of the text data, it is specifically configured to:
inputting the text data into a first convolution neural network model to obtain a word vector corresponding to the text data;
and determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, wherein the first label is used for indicating whether the patient suffers from coronary heart disease, and the second label is used for indicating whether the patient dies.
Further, when the processor 401 determines, according to the word vector, the first tag, and the second tag, a first characterization vector corresponding to the first tag and the second tag of the text data, specifically configured to:
inputting the word vector and the first label into a first classification model to obtain a first vector corresponding to the first label in the text data;
inputting the word vector and the second label into a second classification model to obtain a second vector corresponding to the second label in the text data;
determining a first characterization vector corresponding to the first label and the second label of the text data according to the first vector and the second vector.
Further, the waveform data includes high frequency waveform data and low frequency waveform data; when the processor 401 determines the second characterization vector of the waveform data, it is specifically configured to:
acquiring high-frequency waveform data and low-frequency waveform data from the waveform data according to the sampling frequency;
carrying out dimensionality reduction processing on the high-frequency waveform data to obtain dimensionality reduction waveform data taking hours as a unit;
and determining a second characterization vector of the waveform data according to the dimension reduction waveform data and the low-frequency waveform data.
Further, when the processor 401 determines the second characterization vector of the waveform data according to the dimension-reduced waveform data and the low-frequency waveform data, it is specifically configured to:
performing feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data;
extracting the characteristics of the low-frequency waveform data to obtain a vector of the low-frequency waveform data;
and determining a second characterization vector of the waveform data according to the vector of the dimensionality reduction waveform data and the vector of the low-frequency waveform data.
Further, when the processor 401 determines, according to the intervention data, a second target characterization vector corresponding to the intervention data, specifically configured to:
determining an intervention time corresponding to the intervention data from the intervention data, wherein the intervention data comprises one or more of medication data, examination data, procedure data for each patient in the sample set prior to the first target time;
acquiring text data and waveform data corresponding to the cardiac ultrasonic report at the intervention time;
and determining a second target characterization vector corresponding to the intervention data according to the text data and the waveform data corresponding to the cardiac ultrasonic report corresponding to the intervention time.
Further, when the processor 401 determines, according to the intervention data, an intervention time corresponding to the intervention data, specifically configured to:
determining an intervention indication vector corresponding to the intervention data, the intervention indication vector indicating whether each patient in the sample set has intervened at a time prior to the first target time;
and determining the intervention time of each patient in the sample set according to the intervention indication vector.
In the embodiment of the invention, the coronary heart disease risk prediction device based on multi-modal data can acquire a cardiac ultrasonic report of each patient in a sample set before a first target moment, extract text data from the cardiac ultrasonic report, and determine a first characterization vector of the text data; obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data; performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing; acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data; inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient; and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested. By analyzing the multi-mode data of three dimensions of the cardiac ultrasonic report, the waveform data and the intervention data, the time and the risk probability of the patient suffering from the coronary heart disease and the time and the risk probability of the patient dying can be predicted more efficiently and quickly, and the prediction effect of the coronary heart disease is improved.
It should be understood that, in the embodiment of the present invention, the Processor 401 may be a Central Processing Unit (CPU), and the Processor may also be other general processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable gate arrays (FPGAs) or other Programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, and the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 402 may include both read-only memory and random access memory, and provides instructions and data to the processor 401. A portion of the memory 402 may also include non-volatile random access memory. For example, the memory 402 may also store device type information.
The embodiment of the present invention further provides a computer-readable storage medium, where a computer program is stored, and when the computer program is executed by a processor, the method for predicting coronary heart disease risk based on multi-modal data described in the embodiment corresponding to fig. 2 may be implemented, or the device for predicting coronary heart disease risk based on multi-modal data described in the embodiment corresponding to fig. 3 may also be implemented, which is not described herein again.
The computer readable storage medium may be an internal storage unit of the coronary heart disease risk prediction device based on multi-modal data according to any of the foregoing embodiments, for example, a hard disk or an internal memory of the coronary heart disease risk prediction device based on multi-modal data. The computer readable storage medium may also be an external storage device of the coronary heart disease risk prediction device based on multi-modality data, such as a plug-in hard disk, a Smart Memory Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, which are equipped on the coronary heart disease risk prediction device based on multi-modality data. Further, the computer readable storage medium may also comprise both an internal storage unit and an external storage device of the coronary heart disease risk prediction device based on multimodal data. The computer readable storage medium is for storing the computer program and other programs and data required by the apparatus for coronary heart disease risk prediction based on multimodal data. The computer readable storage medium may also be used to temporarily store data that has been output or is to be output.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention essentially or partially contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product stored in a computer-readable storage medium, which includes several instructions for causing a computer device (which may be a personal computer, a terminal, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned computer-readable storage media comprise: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes. The computer-readable storage medium may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function, and the like; the storage data area may store data created according to the use of the blockchain node, and the like.
The block chain is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism and an encryption algorithm. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
The above description is only a part of the embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present invention, and these modifications or substitutions should be covered within the scope of the present invention.
Claims (10)
1. A coronary heart disease risk prediction device based on multi-modal data, the device comprising: a memory and a processor;
the memory to store program instructions;
the processor, configured to invoke the program instructions, and when the program instructions are executed, configured to:
acquiring a cardiac ultrasound report of each patient in a sample set before a first target time, extracting text data from the cardiac ultrasound report, and determining a first characterization vector of the text data;
obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data;
performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data;
inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient;
and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
2. The apparatus according to claim 1, wherein the text data carries a first tag and a second tag; when the processor determines the first characterization vector of the text data, the processor is specifically configured to:
inputting the text data into a first convolution neural network model to obtain a word vector corresponding to the text data;
and determining a first characterization vector corresponding to the first label and the second label of the text data according to the word vector, the first label and the second label, wherein the first label is used for indicating whether the patient suffers from coronary heart disease, and the second label is used for indicating whether the patient dies.
3. The device according to claim 2, wherein the processor, when determining, from the word vector, the first label and the second label, a first characterization vector corresponding to the first label and the second label of the text data, is specifically configured to:
inputting the word vector and the first label into a first classification model to obtain a first vector corresponding to the first label in the text data;
inputting the word vector and the second label into a second classification model to obtain a second vector corresponding to the second label in the text data;
determining a first characterization vector corresponding to the first label and the second label of the text data according to the first vector and the second vector.
4. The apparatus of claim 1, wherein the waveform data comprises high frequency waveform data and low frequency waveform data; when the processor determines the second characterization vector of the waveform data, the processor is specifically configured to:
acquiring high-frequency waveform data and low-frequency waveform data from the waveform data according to the sampling frequency;
carrying out dimensionality reduction processing on the high-frequency waveform data to obtain dimensionality reduction waveform data taking hours as a unit;
and determining a second characterization vector of the waveform data according to the dimension reduction waveform data and the low-frequency waveform data.
5. The apparatus of claim 4, wherein the processor, when determining the second characterization vector of the waveform data from the reduced-dimension waveform data and the low-frequency waveform data, is specifically configured to:
performing feature extraction on the dimensionality reduction waveform data to obtain a vector of the dimensionality reduction waveform data;
extracting the characteristics of the low-frequency waveform data to obtain a vector of the low-frequency waveform data;
and determining a second characterization vector of the waveform data according to the vector of the dimensionality reduction waveform data and the vector of the low-frequency waveform data.
6. The device according to claim 1, wherein the processor, when determining, from the intervention data, a second target characterization vector corresponding to the intervention data, is configured to:
determining an intervention time corresponding to the intervention data from the intervention data, wherein the intervention data comprises one or more of medication data, examination data, procedure data for each patient in the sample set prior to the first target time;
acquiring text data and waveform data corresponding to the cardiac ultrasonic report at the intervention time;
and determining a second target characterization vector corresponding to the intervention data according to the text data and the waveform data corresponding to the cardiac ultrasonic report corresponding to the intervention time.
7. The device according to claim 6, wherein the processor, when determining, from the intervention data, an intervention time corresponding to the intervention data, is configured to:
determining an intervention indication vector corresponding to the intervention data, the intervention indication vector indicating whether each patient in the sample set has intervened at a time prior to the first target time;
and determining the intervention time of each patient in the sample set according to the intervention indication vector.
8. A coronary heart disease risk prediction method based on multi-modal data is characterized by comprising the following steps:
acquiring a cardiac ultrasound report of each patient in a sample set before a first target time, extracting text data from the cardiac ultrasound report, and determining a first characterization vector of the text data;
obtaining waveform data of each patient in the sample set before the first target time and determining a second characterization vector of the waveform data;
performing fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
acquiring intervention data of each patient in the sample set before the first target moment, and determining a second target characterization vector corresponding to the intervention data according to the intervention data;
inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease of a patient and a second moment and a second risk probability of death of the patient;
and acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, and inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease to obtain a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
9. A coronary heart disease risk prediction device based on multi-modal data, comprising:
the acquisition unit is used for acquiring a cardiac ultrasonic report of each patient in a sample set before a first target moment, extracting text data from the cardiac ultrasonic report and determining a first characterization vector of the text data;
a first determining unit, configured to obtain waveform data of each patient in the sample set before the first target time, and determine a second characterization vector of the waveform data;
the fusion unit is used for carrying out fusion processing on the first characterization vector and the second characterization vector to obtain a first target characterization vector after the fusion processing;
a second determining unit, configured to obtain intervention data of each patient in the sample set before the first target time, and determine a second target characterization vector corresponding to the intervention data according to the intervention data;
the processing unit is used for inputting the first target characterization vector and the second target characterization vector into a specified classification model to obtain a classification result, and training the specified classification model according to the classification result to obtain a risk prediction model of the coronary heart disease, wherein the classification result comprises a first moment and a first risk probability of coronary heart disease suffered by a patient and a second moment and a second risk probability of death of the patient;
and the prediction unit is used for acquiring text data, waveform data and intervention data of the cardiac ultrasonic report of the patient to be tested at a second target moment, inputting the text data, the waveform data and the intervention data into a trained risk prediction model of the coronary heart disease, and obtaining a third moment and a third risk probability of the coronary heart disease of the patient to be tested and a fourth moment and a fourth risk probability of death of the patient to be tested.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which is executed by a processor to implement the method of claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011142479.4A CN112017784B (en) | 2020-10-22 | 2020-10-22 | Coronary heart disease risk prediction method based on multi-modal data and related equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011142479.4A CN112017784B (en) | 2020-10-22 | 2020-10-22 | Coronary heart disease risk prediction method based on multi-modal data and related equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112017784A true CN112017784A (en) | 2020-12-01 |
| CN112017784B CN112017784B (en) | 2021-02-09 |
Family
ID=73527844
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011142479.4A Active CN112017784B (en) | 2020-10-22 | 2020-10-22 | Coronary heart disease risk prediction method based on multi-modal data and related equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112017784B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112842297A (en) * | 2020-12-16 | 2021-05-28 | 大连医科大学附属第一医院 | Signal processing method, signal processing device, computer equipment and storage medium |
| CN113707309A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Disease prediction method and device based on machine learning |
| CN112842297B (en) * | 2020-12-16 | 2024-05-14 | 大连医科大学附属第一医院 | Signal processing method, apparatus, computer device and storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108962387A (en) * | 2018-06-14 | 2018-12-07 | 暨南大学附属第医院（广州华侨医院） | A kind of thyroid nodule Risk Forecast Method and system based on big data |
| EP3477657A1 (en) * | 2017-10-26 | 2019-05-01 | Advanced MR Analytics AB | Evaluating an individual's characteristics of at least one phenotype variable |
| CN109935326A (en) * | 2019-02-28 | 2019-06-25 | 生活空间(沈阳)数据技术服务有限公司 | A kind of probability of illness prediction meanss and storage medium |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| WO2020056372A1 (en) * | 2018-09-14 | 2020-03-19 | Krishnan Ramanathan | Multimodal learning framework for analysis of clinical trials |
| CN111292853A (en) * | 2020-01-15 | 2020-06-16 | 长春理工大学 | Cardiovascular disease risk prediction network model based on multiple parameters and construction method thereof |
| CN111317464A (en) * | 2020-02-13 | 2020-06-23 | 中国科学院自动化研究所 | Electrocardiogram analysis method and device |
| CN111354464A (en) * | 2018-12-24 | 2020-06-30 | 深圳先进技术研究院 | CAD prediction model establishing method and device and electronic equipment |
| CN111400560A (en) * | 2020-03-10 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Method and system for predicting based on heterogeneous graph neural network model |
-
2020
- 2020-10-22 CN CN202011142479.4A patent/CN112017784B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3477657A1 (en) * | 2017-10-26 | 2019-05-01 | Advanced MR Analytics AB | Evaluating an individual's characteristics of at least one phenotype variable |
| CN108962387A (en) * | 2018-06-14 | 2018-12-07 | 暨南大学附属第医院（广州华侨医院） | A kind of thyroid nodule Risk Forecast Method and system based on big data |
| WO2020056372A1 (en) * | 2018-09-14 | 2020-03-19 | Krishnan Ramanathan | Multimodal learning framework for analysis of clinical trials |
| CN111354464A (en) * | 2018-12-24 | 2020-06-30 | 深圳先进技术研究院 | CAD prediction model establishing method and device and electronic equipment |
| CN109935326A (en) * | 2019-02-28 | 2019-06-25 | 生活空间(沈阳)数据技术服务有限公司 | A kind of probability of illness prediction meanss and storage medium |
| CN110347837A (en) * | 2019-07-17 | 2019-10-18 | 电子科技大学 | A kind of unplanned Risk Forecast Method of being hospitalized again of cardiovascular disease |
| CN111292853A (en) * | 2020-01-15 | 2020-06-16 | 长春理工大学 | Cardiovascular disease risk prediction network model based on multiple parameters and construction method thereof |
| CN111317464A (en) * | 2020-02-13 | 2020-06-23 | 中国科学院自动化研究所 | Electrocardiogram analysis method and device |
| CN111400560A (en) * | 2020-03-10 | 2020-07-10 | 支付宝(杭州)信息技术有限公司 | Method and system for predicting based on heterogeneous graph neural network model |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112842297A (en) * | 2020-12-16 | 2021-05-28 | 大连医科大学附属第一医院 | Signal processing method, signal processing device, computer equipment and storage medium |
| CN112842297B (en) * | 2020-12-16 | 2024-05-14 | 大连医科大学附属第一医院 | Signal processing method, apparatus, computer device and storage medium |
| CN113707309A (en) * | 2021-08-31 | 2021-11-26 | 平安科技（深圳）有限公司 | Disease prediction method and device based on machine learning |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112017784B (en) | 2021-02-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111985584B (en) | Disease auxiliary detection equipment, method, device and medium based on multi-mode data | |
| CN112133441B (en) | Method and terminal for establishing MH postoperative crack state prediction model | |
| CN112259246B (en) | Disease prediction method integrating medical concept hierarchy structure and related equipment | |
| CN111933284B (en) | Complication risk prediction system, method, device, equipment and medium | |
| CN112259180B (en) | Disease prediction method based on heterogeneous medical knowledge graph and related equipment | |
| CN109545387B (en) | Abnormal case recognition method and computing equipment based on neural network | |
| CN111192660A (en) | Image report analysis method, equipment and computer storage medium | |
| CN112634231A (en) | Image classification method and device, terminal equipment and storage medium | |
| CN112017784B (en) | Coronary heart disease risk prediction method based on multi-modal data and related equipment | |
| CN116797554A (en) | Image processing method and device | |
| Abou Ghayda et al. | Artificial intelligence in andrology: from semen analysis to image diagnostics | |
| CN116612879B (en) | Diagnostic result prediction method, diagnostic result prediction device, electronic equipment and storage medium | |
| CN112802598A (en) | Real-time auxiliary diagnosis and treatment method and system based on voice diagnosis and treatment data | |
| CN108831560B (en) | Method and device for determining medical data attribute data | |
| CN111145852A (en) | Medical information processing method and device and computer-readable storage medium | |
| CN112037918B (en) | Chronic disease medical insurance cost prediction method fusing complication risks and related equipment | |
| CN113678147A (en) | Search method and information processing system | |
| CN115985506A (en) | Information extraction method and device, storage medium and computer equipment | |
| CN115775635A (en) | Medicine risk identification method and device based on deep learning model and terminal equipment | |
| WO2022036328A1 (en) | Method and apparatus for biopsy | |
| CN113553840A (en) | Text information processing method, device, equipment and storage medium | |
| CN109558398B (en) | Data cleaning method based on big data and related device | |
| CN113707344A (en) | AI-based patient rehabilitation tracking method, device, server and medium | |
| CN112435745A (en) | Consultation strategy recommendation method and device, electronic equipment and storage medium | |
| CN112117015B (en) | Sepsis early warning equipment, sepsis early warning method, sepsis early warning device and sepsis early warning storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |