[go: up one dir, main page]

WO2021056342A1 - Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation - Google Patents

Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation Download PDF

Info

Publication number
WO2021056342A1
WO2021056342A1 PCT/CN2019/108210 CN2019108210W WO2021056342A1 WO 2021056342 A1 WO2021056342 A1 WO 2021056342A1 CN 2019108210 W CN2019108210 W CN 2019108210W WO 2021056342 A1 WO2021056342 A1 WO 2021056342A1
Authority
WO
WIPO (PCT)
Prior art keywords
parameter
ultrasound
neuromodulation
result
prediction model
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.)
Ceased
Application number
PCT/CN2019/108210
Other languages
English (en)
Chinese (zh)
Inventor
郑海荣
牛丽丽
王壹博
孟龙
肖杨
边天元
林争荣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Institute of Advanced Technology of CAS
Original Assignee
Shenzhen Institute of Advanced Technology of CAS
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Institute of Advanced Technology of CAS filed Critical Shenzhen Institute of Advanced Technology of CAS
Priority to PCT/CN2019/108210 priority Critical patent/WO2021056342A1/fr
Publication of WO2021056342A1 publication Critical patent/WO2021056342A1/fr
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
    • A61N1/00—Electrotherapy; Circuits therefor
    • A61N1/18—Applying electric currents by contact electrodes
    • A61N1/32—Applying electric currents by contact electrodes alternating or intermittent currents
    • A61N1/36—Applying electric currents by contact electrodes alternating or intermittent currents for stimulation

Definitions

  • This application belongs to the field of computer technology, and in particular relates to a method, device, computer-readable storage medium, and terminal equipment for predicting neuromodulation results.
  • Transcranial ultrasound stimulation as a new type of non-invasive neuromodulation technology, has great potential to be applied to the regulation of the nervous system.
  • transcranial ultrasound stimulation mainly uses ultrasound to change the activity of neurons to obtain nerve regulation results.
  • the neuromodulation results obtained will also be different.
  • experimental tests are often performed on preset animals to determine the neuromodulation results corresponding to each ultrasound parameter, and the efficiency of determining the neuromodulation results is low. Therefore, how to accurately and efficiently determine the neuromodulation results corresponding to different ultrasound parameters has become an urgent technical problem to be solved by those skilled in the art.
  • the embodiments of the present application provide a method, a device, a computer-readable storage medium, and a terminal device for predicting a neuromodulation result, which can solve the problem that the neuromodulation result corresponding to different ultrasound parameters cannot be accurately and efficiently determined in the prior art.
  • an embodiment of the present application provides a method for predicting neuromodulation results, including:
  • the ultrasound information includes ultrasound parameters, and information about stimulation types and stimulation targets corresponding to the ultrasound parameters;
  • the parameter feature information is input to the neuromodulation result prediction model for processing, and the neuromodulation result output by the neuromodulation result prediction model is obtained.
  • the neuromodulation result prediction model is obtained through training in the following steps:
  • the training nerve regulation result and the first target regulation result adjust the model parameters of the neuroregulation result prediction model, and continue to execute the input of the sample parameter feature information into the neuroregulation result prediction model.
  • the processing steps and subsequent steps until the neuroregulation result prediction model meets the preset training condition, and the neuroregulation result prediction model that satisfies the preset training condition is determined as the trained neuroregulation result prediction model.
  • the obtaining the first training data according to the information of the stimulation type and the stimulation target corresponding to the neuroregulation result prediction model includes:
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample includes:
  • the performing feature extraction on the ultrasound parameters to obtain parameter feature information corresponding to the ultrasound parameters includes:
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample includes:
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample includes:
  • Analyze the ultrasound parameter sample determine the divergence corresponding to each preset parameter feature in the ultrasound parameter sample, and/or determine the difference between each preset parameter feature in the ultrasound parameter sample and the first target control result Relevance
  • the determining the divergence corresponding to each preset parameter feature in the ultrasound parameter sample includes:
  • the variance corresponding to each preset parameter feature in the ultrasound parameter sample is calculated, and the divergence corresponding to each preset parameter feature is determined according to the variance corresponding to each preset parameter feature.
  • the determining the correlation between each preset parameter feature in the ultrasound parameter sample and the first target regulation result includes:
  • the method includes:
  • the prediction accuracy is less than the preset accuracy threshold, adjusting the model parameters of the neuromodulation result prediction model, and continue to use the first training data to train the neuromodulation result prediction model;
  • the training of the neural regulation result prediction model is ended.
  • an embodiment of the present application provides a neuromodulation result prediction device, including:
  • An ultrasound information acquisition module configured to acquire ultrasound information to be predicted, where the ultrasound information includes ultrasound parameters and information about the stimulation types and stimulation targets corresponding to the ultrasound parameters;
  • a predictive model determining module configured to determine a neuromodulation result predictive model corresponding to the ultrasound parameter according to the stimulation type and the information of the stimulation target;
  • the feature extraction module is configured to perform feature extraction on the ultrasound parameters to obtain parameter feature information corresponding to the ultrasound parameters;
  • the regulation result prediction module is used to input the parameter feature information into the neuroregulation result prediction model for processing to obtain the neuroregulation result output by the neuroregulation result prediction model.
  • an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and running on the processor.
  • the processor executes the computer program, The method for predicting neuromodulation results according to any one of the above-mentioned first aspects is realized.
  • an embodiment of the present application provides a computer-readable storage medium that stores a computer program that, when executed by a processor, implements any one of the above-mentioned aspects of the first aspect Methods of predicting the outcome of neuromodulation.
  • the embodiments of the present application provide a computer program product, which when the computer program product runs on a terminal device, causes the terminal device to execute the neuromodulation result prediction method described in any one of the above-mentioned first aspects.
  • multiple neuromodulation result prediction models can be constructed in advance according to the stimulation type and the stimulation target.
  • the parameters of the ultrasound parameters corresponding to the ultrasound information can be extracted first Feature information, and the neuromodulation result prediction model corresponding to the ultrasound parameter can be determined by the stimulation type corresponding to the ultrasound information and the stimulation target information, so that the parameter feature information of the ultrasound parameter can be input to the determined neuromodulation
  • the result prediction model predicts the neuromodulation result corresponding to the ultrasound information, so as to reduce the time and labor cost of neuromodulation result prediction, and improve the accuracy and efficiency of neuromodulation result prediction.
  • FIG. 1 is a schematic flowchart of a method for predicting neuromodulation results according to an embodiment of the present application
  • FIG. 2 is a schematic flowchart of training a neural control result prediction model in an application scenario of the neural control result prediction method provided by an embodiment of the present application;
  • FIG. 3 is a schematic diagram of the process of obtaining first training data in an application scenario by the method for predicting neuromodulation results according to an embodiment of the present application;
  • FIG. 4 is a schematic flow chart of testing a neuromodulation result prediction model in an application scenario according to the method for predicting neuromodulation results according to an embodiment of the present application;
  • Figure 5 is a schematic structural diagram of a neuromodulation result prediction device provided by an embodiment of the present application.
  • Fig. 6 is a schematic structural diagram of a terminal device provided by an embodiment of the present application.
  • the term “if” can be construed as “when” or “once” or “in response to determination” or “in response to detecting “.
  • the phrase “if determined” or “if detected [described condition or event]” can be interpreted as meaning “once determined” or “in response to determination” or “once detected [described condition or event]” depending on the context ]” or “in response to detection of [condition or event described]”.
  • ultrasound parameters are often selected based on existing literature or experimental experience.
  • the ultrasound parameters may include acoustic intensity (Acoustic Intensity). intensity, AI), fundamental frequency (Fundamental frequency, FF), duty cycle (duty cycle, DC) and pulse repetition frequency (pulse Repetition frequency, PRF), etc., and determine the neuromodulation results corresponding to the selected ultrasound parameters through experimental tests on preset animals, which takes a lot of time and labor costs, and it is easy to reduce the efficiency and efficiency of neuromodulation results determination. accuracy.
  • the embodiments of the present application provide a method, a device, a computer-readable storage medium, and a terminal device for predicting neuromodulation results.
  • the corresponding training data can be obtained in advance according to the information of the stimulus type and the stimulus target, and through The corresponding training data is used to train each neuromodulation result prediction model, and multiple trained neuromodulation result prediction models are obtained.
  • the ultrasound information corresponding to the The information of the stimulus type and the stimulus target determines the corresponding neuromodulation result prediction model, so that the neuromodulation result corresponding to the ultrasound information can be predicted through the determined neuromodulation result prediction model, so as to reduce the time and time for predicting the neuromodulation result.
  • Manpower cost improves the accuracy and efficiency of the prediction of neuroregulation results.
  • an embodiment of the present application provides a method for predicting neuromodulation results.
  • the execution subject of the embodiments of the present application is a terminal device, and the terminal device includes, but is not limited to, computing devices such as desktop computers, notebooks, palmtop computers, and cloud servers.
  • the method for predicting the neuroregulation result may include:
  • the user when a user wants to predict the nerve regulation result corresponding to a certain ultrasound stimulation, the user can send the ultrasound information to be predicted corresponding to the ultrasound stimulation to the terminal device, where the sent
  • the ultrasound information may include the specific ultrasound parameters corresponding to the ultrasound stimulation and the stimulation type and stimulation target information that the ultrasound stimulation implements.
  • the stimulation type refers to the object to which the ultrasound stimulation is implemented
  • the stimulation target refers to For the specific location where ultrasound stimulation is performed in each object, the information of the stimulation target may be an identification or type corresponding to the specific location.
  • the type of stimulation may be rodents, primates, humans, etc.
  • the stimulation target may be the central nervous system and peripheral nerves, etc.
  • the information of the stimulation target may be corresponding to the central nervous system or peripheral nerves.
  • S102 Determine a neural regulation result prediction model corresponding to the ultrasound parameter according to the stimulation type and the information of the stimulation target;
  • the neuromodulation results obtained may be different.
  • ultrasounds with the same ultrasound parameters are used to simultaneously treat the central nervous system and the central nervous system of rodents.
  • peripheral nerves are stimulated by ultrasound, it is possible that the nerve regulation results obtained by the central nerve are effective and inhibited, while the nerve regulation results obtained by the peripheral nerves are effective and exciting; another example is the use of ultrasound with the same ultrasonic parameters to simultaneously affect the central nerve and spirit of rodents.
  • the terminal device is provided with a plurality of neural control result prediction models for different stimulation types and stimulation targets, and each of the neural control result prediction models can predict the result according to the parameter characteristic information of the input ultrasonic parameters.
  • the neuromodulation results corresponding to the ultrasound parameters can be obtained through corresponding training data training, wherein the training process of the neural control result prediction model will be described in detail in the following content.
  • the terminal device can determine the neuroregulation result prediction model corresponding to the ultrasound stimulation according to the information of the stimulation type and the stimulation target.
  • the ultrasound stimulation can be determined by matching the information of the stimulation type and the stimulation target corresponding to the ultrasound stimulation with the information of the stimulation type and the stimulation target corresponding to each neuroregulation result prediction model set in the terminal device. Corresponding prediction model of neuroregulation results.
  • the neuromodulation result prediction model corresponds to the feature extraction method one-to-one, after the terminal device determines the neuromodulation result prediction model corresponding to the ultrasound stimulation, it can extract the features corresponding to the neuromodulation result prediction model The method performs feature extraction on the ultrasound parameters corresponding to the ultrasound stimulation.
  • the terminal device can obtain the feature vector determined by the principal component analysis method during the training process, and can use the feature vector to Perform feature extraction on the ultrasound parameters to obtain parameter feature information corresponding to the ultrasound parameters.
  • the terminal device can obtain the mixed matrix determined by the independent component analysis method during the training process, and use the mixed matrix to Perform feature extraction on the ultrasound parameters to obtain parameter feature information corresponding to the ultrasound parameters.
  • the terminal device may be based on the divergence and/or correlation determined during the training process.
  • the ultrasound parameters may be standardized first, that is, the ultrasound parameters may be processed first. After the z-score standardization process, feature extraction is performed on the ultrasonic parameters after the standardization process.
  • Step S104 Input the parameter characteristic information into the neuromodulation result prediction model for processing, and obtain the neuromodulation result output by the neuromodulation result prediction model.
  • the terminal device After extracting the parameter feature information corresponding to the ultrasound parameter, the terminal device can input the parameter feature information into the aforementioned determined neuromodulation result prediction model for processing, and obtain all the output of the neuromodulation result prediction model.
  • the nerve control results corresponding to the ultrasound parameters are described so that the user can optimize and adjust the ultrasound stimulation parameters according to the predicted nerve control results, and the user can accurately predict the therapeutic effect of the ultrasound stimulation.
  • FIG. 2 shows a schematic flow chart of a neuromodulation result prediction method provided by an embodiment of the present application for training a neuromodulation result prediction model in an application scenario.
  • the execution body of the neuroregulation result prediction model training can be the same as the execution body in the above S101 to S104, that is, the terminal device that executes S101 to S104.
  • the execution body of the neuroregulation result prediction model training can be the same as the above.
  • the execution subject in S101 to S104 is different.
  • the training process of the neural regulation result prediction model may specifically include:
  • multiple neural control result prediction models can be constructed in advance based on the information of the stimulation type and the stimulation target, where the stimulation type refers to the object to which the ultrasound stimulation is performed, and the stimulation target refers to the ultrasound stimulation
  • the information of the stimulation target may be an identification or type corresponding to the specific location.
  • the type of stimulation may be rodents, primates, humans, etc.
  • the stimulation targets may be central nerves and peripheral nerves, etc.
  • the information of the stimulation targets may be identifiers corresponding to the central nerves or peripheral nerves.
  • neuroregulation result prediction model A corresponding to rodents and central nervous system
  • neuroregulation result prediction model B corresponding to rodents and peripheral nerves
  • neuroregulation result prediction model B corresponding to primates and central nervous system
  • Neuroregulation outcome prediction model C construction of neuroregulation outcome prediction model D corresponding to primates and peripheral nerves
  • construction of neuroregulation outcome prediction model E corresponding to humans and central nerves
  • prediction of neuroregulation results corresponding to humans and peripheral nerves Model F etc.
  • the first training data corresponding to each neuromodulation result prediction model can be obtained according to the information of the stimulus type and the stimulation target. For example, the first training data corresponding to each neuromodulation result prediction model can be obtained first. The historical experimental data related to ultrasound parameters can then be classified according to the stimulus type and the information of the stimulus target to obtain the first training data corresponding to each neuromodulation result prediction model.
  • the first training data may include an ultrasound parameter sample and a first target adjustment result corresponding to the ultrasound parameter sample.
  • the first target adjustment result may be an experiment result in each historical experiment data, and the experiment
  • the results can include effective and inhibited, effective and excited and ineffective, wherein the effective and inhibited refers to the effective inhibition of the activity of neurons in the stimulation target, and the effective and exciting refers to the effective activation of the nerves in the stimulated target. Yuan's activities.
  • each neuromodulation result prediction model is the same, and the basic principles are similar.
  • the training methods of each neuromodulation result prediction model are the same, and the basic principles are similar.
  • the following will take the neuromodulation result prediction model A as an example for illustration, and the training of other neuromodulation result prediction models can refer to neuromodulation
  • the result predicts the training of model A.
  • the stimulus type and the stimulation target corresponding to the neuromodulation result prediction model may include:
  • the first training data can be pre-processed and standardized data, that is, the neuroregulation result prediction model can be obtained based on the neuroregulation result prediction model A corresponding to the stimulation type and stimulation target information.
  • the initial training data corresponding to A can then be preprocessed respectively on each of the initial training data to obtain the preprocessed second training data.
  • the consistency check and the check of invalid and missing values can be used to check each of the initial training data.
  • Data cleaning is performed on the initial training data to remove duplicate data and invalid data in the initial training data, so as to improve the effectiveness of the training data and improve the training efficiency and prediction accuracy of the neural regulation result prediction model.
  • each of the cleaned second training data may be standardized, and the standardized second training data may be determined as the first training data.
  • each of the cleaned second training data may be determined as the first training data.
  • S202 Perform feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample;
  • the data in the first training data A can be compared Perform feature extraction for each ultrasound parameter sample A to obtain sample parameter feature information A corresponding to each ultrasound parameter sample A, so as to reduce the computational complexity of the neuromodulation result prediction model through parameter feature extraction, thereby improving the neuromodulation result prediction The predictive efficiency of the model.
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample may include:
  • Step a Analyze the ultrasonic parameter sample by using the principal component analysis method to obtain the feature vector corresponding to the ultrasonic parameter sample;
  • Step b Perform feature extraction on the ultrasound parameter sample according to the feature vector to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the principal component analysis method can be used to extract features of the ultrasound parameter sample A of the neuromodulation result prediction model A.
  • the principal component analysis method can be used to perform principal component analysis on all the ultrasound parameter samples A first.
  • the feature vector corresponding to the neuromodulation result prediction model A is obtained, and then the feature vector can be used to perform feature extraction on each ultrasound parameter sample A in the neuromodulation result prediction model A.
  • the process of using the principal component analysis method to perform sample parameter feature extraction on the ultrasonic parameter sample A can refer to the existing process of using the principal component analysis method to perform data dimensionality reduction, which is not limited in the embodiment of the present application.
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample may include:
  • Step c Analyze the ultrasound parameter samples using an independent component analysis method to obtain a mixing matrix corresponding to the ultrasound parameter samples;
  • Step d Perform feature extraction on the ultrasound parameter sample according to the hybrid matrix to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the independent component analysis method can also be used to extract features of the ultrasound parameter sample A of the neuromodulation result prediction model A.
  • the independent component analysis method can be used to perform independent component analysis on all the ultrasound parameter samples A.
  • the hybrid matrix corresponding to the neuromodulation result prediction model A is obtained, and then the feature extraction of each ultrasound parameter sample A in the neuromodulation result prediction model A can be performed through the hybrid matrix.
  • the process of using the independent component analysis method to perform sample parameter feature extraction on the ultrasonic parameter sample A can also refer to the existing process of using the independent component analysis method to perform data dimensionality reduction, which is not limited in the embodiment of the present application.
  • the performing feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample may include:
  • Step e Analyze the ultrasound parameter sample, determine the divergence corresponding to each preset parameter feature in the ultrasound parameter sample, and/or determine whether each preset parameter feature in the ultrasound parameter sample and the first target control Correlation between results;
  • each ultrasound parameter sample A in the neuromodulation result prediction model A it is also possible to analyze the preset parameter features of each ultrasound parameter sample A in the neuromodulation result prediction model A to determine the corresponding preset parameter feature in the neuromodulation result prediction model A Divergence and determining the correlation between each preset parameter feature in the neural control result prediction model A and the first target control result, for example, determine the divergence and fundamental frequency corresponding to the sound intensity in the neural control result prediction model A The corresponding divergence, etc., as well as the correlation between the duty cycle and the first target control result, the correlation between the pulse repetition frequency and the first target control result in the neuroregulation result prediction model A, etc. are determined.
  • the determining the divergence corresponding to each preset parameter feature in the ultrasound parameter sample may include: calculating the variance corresponding to each preset parameter feature in the ultrasound parameter sample, and according to each preset parameter feature The corresponding variance determines the divergence corresponding to each of the preset parameter features.
  • the feature value corresponding to each preset parameter feature in each ultrasound parameter sample can be obtained first, and then the variance corresponding to each preset parameter feature can be calculated according to each feature value corresponding to each preset parameter feature.
  • the variance a corresponding to the preset parameter feature a can be calculated by the feature values a1, a2, a3,..., an corresponding to the preset parameter feature a (such as sound intensity), and the variance a corresponding to the preset parameter feature a can be calculated through the preset parameter feature b (such as Fundamental frequency) corresponding to the feature values b1, b2, b3,..., bn to calculate the variance b corresponding to the preset parameter feature b, and the feature values c1, c2, corresponding to the preset parameter feature c (such as duty cycle) can be used c3,...,cn to calculate the variance c corresponding to the preset parameter feature c, and so on.
  • the divergence corresponding to each preset parameter feature can be determined according to the calculated variances.
  • the variance corresponding to each preset parameter feature can be used to directly determine the divergence corresponding to the preset parameter feature, or it can be based on the variance and divergence.
  • the smaller the variance corresponding to the preset parameter feature the smaller the divergence corresponding to the preset parameter feature, which means that each ultrasound There is basically no difference between the parameter samples in the preset parameter characteristics.
  • the determining the correlation between each preset parameter feature in the ultrasound parameter sample and the first target regulation result may include:
  • the mutual information between each preset parameter feature and the first target regulation result can be calculated first, and the mutual information between each preset parameter feature and the first target regulation result can be determined based on the mutual information.
  • the mutual information corresponding to each preset parameter feature can be directly determined as the correlation between the preset parameter feature and the first target regulation result, or it can be based on the preset correspondence between the mutual information and the correlation Relationship to determine the correlation between each preset parameter feature and the first target regulation result.
  • Step f Perform feature extraction on the ultrasound parameter sample according to the divergence and/or the correlation to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the ultrasonic parameter sample can be feature extracted according to the divergence and/or correlation to obtain the ultrasonic parameter sample corresponding Among them, the predicted parameter features with greater divergence or greater correlation are easier to be extracted, and the extracted preset parameter features can constitute the sample parameter feature information corresponding to the ultrasound parameter sample.
  • the feature extraction method described in step e and step f can be used simultaneously with the aforementioned principal component analysis method or independent component analysis method.
  • the parameter sample is extracted for the first time, and then the feature extraction method described in step e and step f can be used to perform further feature extraction on the parameter features extracted by the principal component analysis method or the independent component analysis method to improve the parameter characteristics The effectiveness of the extraction, thereby improving the prediction accuracy of the neural regulation result prediction model.
  • the feature extraction method described in step e and step f can also be used to extract the features of the ultrasonic parameter sample for the first time, and then the principal component analysis method or the independent component analysis method can be used to perform the first feature extraction.
  • the parameter features extracted at one time are subjected to further feature extraction.
  • the parameter feature information of each sample can be input into the neuroregulation result prediction model for training, and Obtain the training neural control result corresponding to each sample parameter feature output by the neural control result prediction model.
  • the model parameters of the neuromodulation result prediction model can be adjusted according to the training neuromodulation results and the first target regulation results corresponding to the characteristic information of each sample parameter.
  • the training loss value corresponding to the neural control result prediction model can be calculated according to each training neural control result and each first target control result, and the model parameters of the neural control result prediction model can be modified by back propagation of the training loss value ,
  • use the neuromodulation result prediction model after the model parameter modification to continue to execute the step of inputting the sample parameter feature information into the neuromodulation result prediction model for processing and subsequent steps until the neuromodulation result prediction model satisfies the preset training Condition, wherein the preset training condition may be that the number of training times reaches a preset threshold; the preset training condition may also be that the training error reaches an optimal value; the preset training condition may also be that the number of training times reaches a preset value The threshold of times or the training error reaches the optimal value.
  • the method may include:
  • S401 Use preset test data to test the neuromodulation result prediction model, and obtain a test neuromodulation result corresponding to each of the test data output by the neuromodulation result prediction model;
  • the test data is similar to the first training data, and may also include ultrasound parameter samples and second target adjustment results corresponding to the ultrasound parameter samples.
  • the obtained training data can be divided into two parts, one part is the first training Data, and the other part is the test data.
  • 7500 pieces of training data can be randomly selected from the 10,000 pieces of training data as the neuroregulation result prediction model A
  • the remaining 2500 training data can be used as the test data corresponding to the neural regulation result prediction model A.
  • the training data corresponding to each neuroregulation result prediction model after obtaining the training data corresponding to each neuroregulation result prediction model according to the information of the stimulus type and the stimulus target, there may be a replacement to randomly extract the first number of training from the obtained training data Data is used as the first training data, and a second amount of training data can be randomly selected from the acquired training data as the test data.
  • a second amount of training data can be randomly selected from the acquired training data as the test data.
  • 7500 pieces of training data can be randomly selected from the 10,000 pieces of training data as the first training data corresponding to the neuroregulation result prediction model A
  • 2500 pieces of training data can be randomly selected from the 10,000 pieces of training data.
  • the test data corresponding to model A is predicted as the result of neuroregulation.
  • the test data can be used to test the neuromodulation result prediction model, that is, the ultrasonic sample parameters corresponding to each test data can be feature extracted, and the feature information of the extracted sample parameters can be input to
  • the trained neural control result prediction model is processed to obtain test neural control results corresponding to each of the test data output by the neural control result prediction model.
  • the test data can be considered to be accurate in prediction. On the contrary, the test data is considered to be inaccurate.
  • S402 Determine the prediction accuracy of the neuroregulation result prediction model according to the second target regulation result corresponding to each test data and the test neuroregulation result;
  • the second target control results and the test nerve control results corresponding to each test data can be counted to obtain the number of accurate predictions and the number of inaccurate predictions, respectively.
  • the prediction accuracy of the neural regulation result prediction model can be calculated according to the number of accurate predictions and the number of inaccurate predictions.
  • the prediction accuracy is less than the preset accuracy threshold, adjust the model parameters of the neuromodulation result prediction model, and continue to use the first training data to predict the neuromodulation result model Training; if the prediction accuracy is greater than or equal to the accuracy threshold, the training of the neuromodulation result prediction model is ended, and the neuromodulation result prediction model after the training is determined to be the final prediction of the neuromodulation result
  • the model can subsequently be used to predict the neuromodulation results corresponding to a certain ultrasound parameter.
  • the accuracy threshold may be specifically set according to actual conditions. For example, the accuracy threshold may be set to 90%, 98%, or other values.
  • multiple neuromodulation result prediction models can be constructed in advance according to the stimulation type and the stimulation target.
  • the parameters of the ultrasound parameters corresponding to the ultrasound information can be extracted first Feature information, and the neuromodulation result prediction model corresponding to the ultrasound parameter can be determined by the stimulation type corresponding to the ultrasound information and the stimulation target information, so that the parameter feature information of the ultrasound parameter can be input to the determined neuromodulation
  • the result prediction model predicts the neuromodulation result corresponding to the ultrasound information, so as to reduce the time and labor cost of neuromodulation result prediction, and improve the accuracy and efficiency of neuromodulation result prediction.
  • FIG. 5 shows a structural block diagram of the neuromodulation result prediction device provided in an embodiment of the present application. For ease of description, only the information related to the embodiment of the present application is shown. section.
  • the neuroregulation result prediction device includes:
  • the ultrasound information acquisition module 501 is configured to acquire ultrasound information to be predicted, where the ultrasound information includes ultrasound parameters and information about the stimulation types and stimulation targets corresponding to the ultrasound parameters;
  • a predictive model determining module 502 configured to determine a neuromodulation result predictive model corresponding to the ultrasound parameter according to the stimulus type and the information of the stimulus target;
  • the feature extraction module 503 is configured to perform feature extraction on the ultrasound parameters to obtain parameter feature information corresponding to the ultrasound parameters;
  • the regulation result prediction module 504 is configured to input the parameter feature information into the neuroregulation result prediction model for processing, and obtain the neuroregulation result output by the neuroregulation result prediction model.
  • the neuromodulation result prediction device may further include:
  • the training data acquisition module is configured to acquire first training data according to the information of the stimulation type and the stimulation target corresponding to the neuromodulation result prediction model, where the first training data includes ultrasound parameter samples and the ultrasound parameter samples The corresponding first target regulation result;
  • a sample feature extraction module configured to perform feature extraction on the ultrasound parameter sample to obtain sample parameter feature information corresponding to the ultrasound parameter sample
  • a training module configured to input the sample parameter feature information into the neuroregulation result prediction model for processing, and obtain the training neuroregulation result output by the neuroregulation result prediction model;
  • the first parameter adjustment module is configured to adjust the model parameters of the neural adjustment result prediction model according to the training nerve adjustment result and the first target adjustment result, and continue to execute the input of the sample parameter feature information into The processing steps and subsequent steps of the neuroregulation result prediction model until the neuroregulation result prediction model satisfies a preset training condition, and the neuroregulation result prediction model that satisfies the preset training condition is determined as the trained nerve Control result prediction model.
  • the training data acquisition module may include:
  • An initial training data acquisition unit configured to acquire initial training data according to the information of the stimulation type and the stimulation target corresponding to the neural regulation result prediction model
  • a preprocessing unit configured to preprocess the initial training data to obtain second training data
  • the standardization unit is configured to perform standardization processing on the second training data, and determine the second training data after the standardization processing as the first training data.
  • the sample feature extraction module may include:
  • the feature vector acquiring unit is configured to analyze the ultrasound parameter sample by using the principal component analysis method to obtain the feature vector corresponding to the ultrasound parameter sample;
  • the first sample feature extraction unit is configured to perform feature extraction on the ultrasound parameter sample according to the feature vector to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the feature extraction module 503 is specifically configured to perform feature extraction on the ultrasound parameter according to the feature vector to obtain parameter feature information corresponding to the ultrasound parameter.
  • the sample feature extraction module may further include:
  • a hybrid matrix acquiring unit configured to analyze the ultrasonic parameter sample by using an independent component analysis method to obtain a hybrid matrix corresponding to the ultrasonic parameter sample;
  • the second sample feature extraction unit is configured to perform feature extraction on the ultrasound parameter sample according to the mixing matrix to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the sample feature extraction module may further include:
  • the divergence determination unit is used to analyze the ultrasonic parameter sample, determine the divergence corresponding to each preset parameter feature in the ultrasonic parameter sample, and/or determine the relationship between each preset parameter feature in the ultrasonic parameter sample and the The correlation between the regulation results of the first target;
  • the third sample feature extraction unit is configured to perform feature extraction on the ultrasound parameter sample according to the divergence and/or the correlation to obtain sample parameter feature information corresponding to the ultrasound parameter sample.
  • the divergence determination unit may include:
  • the variance calculation subunit is used to calculate the variance corresponding to each preset parameter feature in the ultrasound parameter sample, and determine the divergence corresponding to each preset parameter feature according to the variance corresponding to each preset parameter feature.
  • the divergence determination unit may further include:
  • the mutual information calculation subunit is used to calculate the mutual information between each preset parameter feature in the ultrasound parameter sample and the first target control result, and determine the relationship between each preset parameter feature and the first target adjustment result according to the mutual information. The correlation between the regulation results of the first target is described.
  • the neuromodulation result prediction device may further include:
  • the test module is configured to use preset test data to test the neuromodulation result prediction model, and obtain the test neuromodulation result corresponding to each of the test data output by the neuromodulation result prediction model;
  • An accuracy determination module configured to determine the prediction accuracy of the neuroregulation result prediction model according to the second target regulation result corresponding to each test data and the test neuroregulation result;
  • the second parameter adjustment module is configured to adjust the model parameters of the neuromodulation result prediction model if the prediction accuracy is less than a preset accuracy threshold, and continue to use the first training data to adjust the neuromodulation result Predictive model for training;
  • the training end module is configured to end the training of the neural control result prediction model if the prediction accuracy is greater than or equal to the accuracy threshold.
  • FIG. 6 is a schematic structural diagram of a terminal device provided by an embodiment of this application.
  • the terminal device 6 of this embodiment includes: at least one processor 60 (only one is shown in FIG. 6), a processor, a memory 61, and a processor that is stored in the memory 61 and can be processed in the at least one processor.
  • a computer program 62 running on the processor 60 when the processor 60 executes the computer program 62, the steps in any of the foregoing neuromodulation result prediction method embodiments are implemented.
  • the terminal device 6 may be a computing device such as a desktop computer, a notebook, a palmtop computer, and a cloud server.
  • the terminal device may include, but is not limited to, a processor 60 and a memory 61.
  • FIG. 6 is only an example of the terminal device 6 and does not constitute a limitation on the terminal device 6. It may include more or less components than shown in the figure, or a combination of certain components, or different components. , For example, can also include input and output devices, network access devices, and so on.
  • the processor 60 may be a central processing unit (Central Processing Unit, CPU), the processor 60 may also be other general-purpose processors or digital signal processors (Digital Signal Processor, DSP), Application Specific Integrated Circuit (ASIC), ready-made programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc.
  • the general-purpose processor may be a microprocessor or the processor may also be any conventional processor or the like.
  • the memory 61 may be an internal storage unit of the terminal device 6 in some embodiments, such as a hard disk or a memory of the terminal device 6. In other embodiments, the memory 61 may also be an external storage device of the terminal device 6, such as a plug-in hard disk equipped on the terminal device 6, a smart memory card (Smart Media Card, SMC), and a secure digital (Secure Digital, SD) card, flash memory card (Flash Card) and so on. Further, the memory 61 may also include both an internal storage unit of the terminal device 6 and an external storage device.
  • the memory 61 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program. The memory 61 can also be used to temporarily store data that has been output or will be output.
  • the embodiments of the present application also provide a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the foregoing method embodiments can be implemented.
  • the embodiments of the present application provide a computer program product.
  • the terminal device can implement the steps in the foregoing method embodiments when executed.
  • the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer readable storage medium.
  • the computer program can be stored in a computer-readable storage medium. When executed by the processor, the steps of the foregoing method embodiments can be implemented.
  • the computer program includes computer program code, and the computer program code may be in the form of source code, object code, executable file, or some intermediate forms.
  • the computer-readable medium may include at least: any entity or device capable of carrying computer program code to the photographing device/terminal device, recording medium, computer memory, read-only memory (Read-Only Memory, ROM), and random access memory (Random Access Memory, RAM), electrical carrier signals, telecommunications signals, and software distribution media.
  • any entity or device capable of carrying computer program code to the photographing device/terminal device recording medium, computer memory, read-only memory (Read-Only Memory, ROM), and random access memory (Random Access Memory, RAM), electrical carrier signals, telecommunications signals, and software distribution media.
  • ROM read-only memory
  • RAM random access memory
  • electrical carrier signals telecommunications signals
  • software distribution media for example, U disk, mobile hard disk, floppy disk or CD-ROM, etc.
  • computer-readable media cannot be electrical carrier signals and telecommunication signals.
  • the disclosed device/terminal device and method may be implemented in other ways.
  • the device/terminal device embodiments described above are merely illustrative.
  • the division of the modules or units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units.
  • components can be combined or integrated into another system, or some features can be omitted or not implemented.
  • the displayed or discussed mutual coupling or direct coupling or communication connection may be indirect coupling or communication connection through some interfaces, devices or units, and may be in electrical, mechanical or other forms.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the objectives of the solutions of the embodiments.

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Radiology & Medical Imaging (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Veterinary Medicine (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

La présente demande a trait au domaine de la technologie informatique et concerne plus particulièrement un procédé, un dispositif et un appareil terminal de prédiction d'un résultat de neuromodulation. Le procédé de prédiction d'un résultat de neuromodulation comprend : l'acquisition d'informations ultrasonores pour les soumettre à une prédiction, les informations ultrasonores comprenant un paramètre ultrasonore et des informations concernant un type de stimulation et un site cible de stimulation correspondant au paramètre ultrasonore ; la détermination, sur la base des informations concernant le type de stimulation et le site cible de stimulation, d'un modèle de prédiction de résultat de neuromodulation correspondant au paramètre ultrasonore ; la réalisation d'une extraction de caractéristiques sur le paramètre ultrasonore pour obtenir des informations de caractéristiques de paramètre correspondant au paramètre ultrasonore ; et l'entrée des informations de caractéristiques de paramètre dans le modèle de prédiction de résultat de neuromodulation et la réalisation d'un traitement pour obtenir un résultat de neuromodulation émis par le modèle de prédiction de résultat de neuromodulation. Selon la présente invention, un modèle de prédiction de résultat de neuromodulation pré-entraîné est utilisé pour effectuer une prédiction de résultat de neuromodulation, réduire le temps et les coûts de main-d'œuvre associés à la prédiction de résultat de neuromodulation, et augmenter la précision et l'efficacité de celle-ci.
PCT/CN2019/108210 2019-09-26 2019-09-26 Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation Ceased WO2021056342A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/108210 WO2021056342A1 (fr) 2019-09-26 2019-09-26 Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2019/108210 WO2021056342A1 (fr) 2019-09-26 2019-09-26 Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation

Publications (1)

Publication Number Publication Date
WO2021056342A1 true WO2021056342A1 (fr) 2021-04-01

Family

ID=75165500

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/108210 Ceased WO2021056342A1 (fr) 2019-09-26 2019-09-26 Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation

Country Status (1)

Country Link
WO (1) WO2021056342A1 (fr)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102056547A (zh) * 2008-06-03 2011-05-11 株式会社日立医疗器械 医用图像处理装置及医用图像处理方法
CN104548390A (zh) * 2014-12-26 2015-04-29 中国科学院深圳先进技术研究院 一种超声深部脑刺激方法及系统
WO2016033543A1 (fr) * 2014-08-28 2016-03-03 Medtronic Ardian Luxembourg S.A.R.L. Procédés d'évaluation de l'efficacité d'une neuromodulation rénale et systèmes et dispositifs associés
CN106823137A (zh) * 2016-12-30 2017-06-13 王征 一种优化神经调控的方法和装置
WO2019173525A1 (fr) * 2018-03-09 2019-09-12 General Electric Company Techniques de neuromodulation ultrasonore
WO2019173518A1 (fr) * 2018-03-09 2019-09-12 General Electric Company Techniques de neuromodulation

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102056547A (zh) * 2008-06-03 2011-05-11 株式会社日立医疗器械 医用图像处理装置及医用图像处理方法
WO2016033543A1 (fr) * 2014-08-28 2016-03-03 Medtronic Ardian Luxembourg S.A.R.L. Procédés d'évaluation de l'efficacité d'une neuromodulation rénale et systèmes et dispositifs associés
CN104548390A (zh) * 2014-12-26 2015-04-29 中国科学院深圳先进技术研究院 一种超声深部脑刺激方法及系统
CN106823137A (zh) * 2016-12-30 2017-06-13 王征 一种优化神经调控的方法和装置
WO2019173525A1 (fr) * 2018-03-09 2019-09-12 General Electric Company Techniques de neuromodulation ultrasonore
WO2019173518A1 (fr) * 2018-03-09 2019-09-12 General Electric Company Techniques de neuromodulation

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
BLACKMORE JOSEPH; SHRIVASTAVA SHAMIT; SALLET JEROME; BUTLER CHRIS R.; CLEVELAND ROBIN O.: "Ultrasound Neuromodulation: A Review of Results, Mechanisms and Safety", ULTRASOUND IN MEDICINE AND BIOLOGY., NEW YORK, NY, US, vol. 45, no. 7, 1 January 1900 (1900-01-01), US, pages 1509 - 1536, XP085704226, ISSN: 0301-5629, DOI: 10.1016/j.ultrasmedbio.2018.12.015 *
SHEN XUELIAN, YAN FEI, ZHAO YUN, ZHOU JUN: "Research progress of neuromodulation with ultrasound", J CLIN ULTRASOUND IN MED, vol. 18, no. 11, 1 November 2016 (2016-11-01), pages 764 - 766, XP055794855, DOI: 10.16245/j.cnki.issn1008-6978.2016.11.016 *
WANG, JUN, LI SUI, AINAN CAI, YONGLIANG WU: "Effects of ultrasound stimulation parameters on neuromodulation", CHINESE JOURNAL OF MEDICAL PHYSICS, vol. 35, no. 2, 1 February 2018 (2018-02-01), pages 236 - 242, XP055794850, DOI: 10.3969/j.issn.1005-202X.2018.02.023 *

Similar Documents

Publication Publication Date Title
CN109636061B (zh) 医保欺诈预测网络的训练方法、装置、设备及存储介质
US11983620B2 (en) Simplification of spiking neural network models
CN110222992A (zh) 一种基于受骗群体画像的网络诈骗预警方法及装置
CN110647849B (zh) 一种神经调控结果预测方法、装置及终端设备
GB2524639A (en) Methods and systems for analyzing healthcare data
WO2019242155A1 (fr) Procédé et appareil de gestion de la santé basés sur la reconnaissance vocale et dispositif informatique
CN114403899B (zh) 一种大脑神经元锋电位与局部场电位结合的抑郁检测装置
WO2019080502A1 (fr) Procédé de prédiction de maladie basée sur la voix, serveur d'application et support d'informations lisible par ordinateur
CN112635053A (zh) 基于大数据的居民健康预警方法、装置、设备和系统
CN116936117A (zh) 基于ai分析模型的慢性病大数据识别和分析处理方法
CN116311539A (zh) 基于毫米波的睡眠动作捕捉方法、装置、设备及存储介质
CN109934723B (zh) 一种医保欺诈行为识别方法、装置及设备
KR20250043829A (ko) 단백질-단백질 상호작용 네트워크를 이용한 약물 효과 예측 방법 및 시스템
CN109559821B (zh) 基于数据处理的风湿性心脏瓣膜病认证方法及相关设备
Berger et al. Efficient identification of assembly neurons within massively parallel spike trains
CN107944754A (zh) 康复治疗质量评定的方法、装置、存储介质及电子设备
CN114462925B (zh) 库存异常资产识别方法、装置及终端设备
WO2021056342A1 (fr) Procédé, dispositif et appareil terminal de prédiction de résultat de neuromodulation
CN112035361B (zh) 医疗诊断模型的测试方法、装置、计算机设备和存储介质
CN114743619A (zh) 一种用于疾病风险预测的调查问卷质量评价方法及系统
CN118213088B (zh) 罕见病临床试验样本含量估算方法及系统
CN110706803B (zh) 一种确定心肌纤维化的方法、装置、可读介质及电子设备
CN112819430A (zh) 一种资源筹集项目申请验证方法、系统和电子设备
CN113779635B (zh) 一种医疗数据的校验方法、装置、设备及存储介质
CN119479956B (zh) 一种临床试验信息录入方法、装置、存储介质及电子设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19947385

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19947385

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 19947385

Country of ref document: EP

Kind code of ref document: A1