Adjusting method of millimeter wave therapeutic instrument

Abstract

The invention relates to the field of millimeter wave therapeutic apparatuses, in particular to a millimeter wave therapeutic apparatus and an adjusting method thereof. The heart rate, the blood pressure and the skin temperature of a patient in the treatment process are subjected to real-time monitoring and data time sequence analysis by adopting an artificial intelligence technology based on deep learning, time sequence change rules of various physiological parameters are mined respectively, and multi-modal data time sequence characteristic principal component fusion analysis is carried out on the multiple physiological parameters. The comprehensive description of the physiological state of the patient is obtained, and then self-adaptive adjustment of the running power of the millimeter wave therapeutic apparatus main body is carried out, so that real-time monitoring and personalized treatment of the physiological state of the patient can be realized, the performance of the millimeter wave therapeutic apparatus is effectively improved, and the treatment effect and the intelligent level of the equipment are improved.

Description

Technical Field

[0001] The present application relates to the field of millimeter wave therapeutic apparatus, and in particular to a method for adjusting a millimeter wave therapeutic apparatus. Background Art

[0002] With the development of science and technology, medical equipment is moving towards intelligence and personalization. As a new type of physical therapy, millimeter wave therapy has been widely used in clinical medicine because of its good penetration and non-thermal effect.

[0003] Traditional millimeter wave therapy devices mainly rely on the doctor's experience to set treatment parameters. Although this method is effective, it has certain limitations in terms of accuracy and personalized treatment, and the adjustment efficiency is low. In order to improve the intelligence level of millimeter wave therapy devices and achieve personalized treatment, there are currently some products or research projects on the market that attempt to apply sensor technology and data analysis methods to millimeter wave therapy.

[0004] However, most of the existing related technologies are still at a relatively early stage. For example, the preset operating mode of the therapeutic device is switched according to the patient's physiological parameter threshold, but there is a lack of in-depth analysis and dynamic response to the patient's real-time physiological state, and true real-time dynamic adjustment cannot be achieved, resulting in insufficient flexibility and adaptability of the millimeter wave therapeutic device.

[0005] Therefore, an optimized millimeter wave therapy device and its adjustment method are expected. Summary of the invention

[0006] This application is made in view of the above problems. One object of this application is to provide a method for adjusting a millimeter wave therapeutic apparatus.

[0007] The embodiment of the present application provides a millimeter wave therapeutic device, which includes: a millimeter wave therapeutic device body, a data acquisition module, a data transmission module and a parameter adjustment controller;

[0008] The data acquisition module is used to receive a time queue of the patient's physiological parameters, wherein the patient's physiological parameters include heart rate value, blood pressure value and skin temperature value;

[0009] The data transmission module is used to transmit the time queue of the patient's physiological parameters to the parameter adjustment controller;

[0010] The parameter adjustment controller is used to process the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body;

[0011] Wherein, the parameter adjustment controller comprises:

[0012] A patient physiological parameter time series encoding module, used for extracting time series features of the time series of the patient physiological parameters based on parameter samples to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector and a skin temperature time series correlation feature vector;

[0013] A multimodal feature matching and fusion module, used for performing principal component feature matching and fusion processing on the heart rate value time series associated feature vector, the blood pressure value time series associated feature vector and the skin temperature time series associated feature vector to obtain a multimodal time series significant fusion representation vector of a physiological state;

[0014] The adjustment instruction generation module is used to generate the parameter adaptive adjustment instruction based on the physiological state multimodal time series significant fusion representation vector, and the parameter adaptive adjustment instruction includes a decoded value of the recommended power value at the next time point.

[0015] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the patient physiological parameter timing encoding module includes:

[0016] A data regularization unit, used for performing data regularization on the time queue of the patient's physiological parameters according to the parameter sample dimension to obtain a time queue of heart rate values, a time queue of blood pressure values, and a time queue of skin temperature values;

[0017] A parameter timing encoding unit is used to perform sequence encoding on the time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector.

[0018] For example, according to the millimeter wave therapy device of an embodiment of the present application, the parameter timing encoding unit is used to: input the time queue of the heart rate value, the time queue of the blood pressure value, and the time queue of the skin temperature value into a sequence encoder based on a Bi-LSTM model to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector, and the skin temperature timing association feature vector.

[0019] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the multimodal feature matching and fusion module includes:

[0020] A heart rate and blood pressure time series association fusion unit, used for inputting the heart rate value time series association feature vector and the blood pressure value time series association feature vector into a matching fusion network based on the feature principal component as the granularity to obtain a heart rate-blood pressure time series association significant fusion representation vector;

[0021] The full-modal fusion representation unit is used to input the heart rate-blood pressure time series association significant fusion representation vector and the skin temperature time series association feature vector into the matching fusion network based on the feature principal component as the granularity to obtain the physiological state multi-modal time series significant fusion representation vector.

[0022] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the heart rate and blood pressure timing correlation fusion unit includes:

[0023] A standardization subunit, used for performing standardization processing on the heart rate value time series association feature vector and the blood pressure value time series association feature vector to obtain a standardized heart rate value time series association feature vector and a standardized blood pressure value time series association feature vector;

[0024] A sample covariance calculation subunit, used to calculate the sample covariance matrices of the standardized heart rate value time series associated feature vector and the standardized blood pressure value time series associated feature vector respectively to obtain a heart rate sample covariance matrix and a blood pressure sample covariance matrix;

[0025] A principal component feature extraction subunit, used for performing matrix decomposition-based feature vector extraction on the heart rate sample covariance matrix and the blood pressure sample covariance matrix to obtain a set of heart rate principal component feature vectors and a set of blood pressure principal component feature vectors;

[0026] A query matching subunit, used for inputting the set of the heart rate principal component feature vectors and the set of the blood pressure principal component feature vectors into a maximum approximate query matching network to obtain a set of best matching pairs of the heart rate principal component feature vectors and the blood pressure principal component feature vectors;

[0027] A fine-grained semantic joint subunit, used for inputting each best matching pair in the set of best matching pairs of the heart rate principal component feature vector and the blood pressure principal component feature vector into a semantic fine-grained joint analysis module to obtain a set of heart rate-blood pressure time series principal component joint feature vectors;

[0028] The principal component matching feature global aggregation subunit is used to cascade the set of the heart rate-blood pressure time series principal component joint feature vectors to obtain the heart rate-blood pressure time series association significant fusion representation vector.

[0029] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the query matching subunit is used to:

[0030] Calculating the cosine similarity between the i-th heart rate principal component eigenvector in the set of heart rate principal component eigenvectors and each blood pressure principal component eigenvector in the set of blood pressure principal component eigenvectors to obtain a set of cosine similarities;

[0031] The blood pressure principal component eigenvector corresponding to the maximum value in the set of cosine similarities and the i-th heart rate principal component eigenvector form the best matching pair.

[0032] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the fine-grained semantic association subunit is used to:

[0033] Calculate the position point addition, position point subtraction and position point multiplication between the heart rate principal component eigenvector and the blood pressure principal component eigenvector in the best matching pair to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result;

[0034] After the first matching fusion result, the second matching fusion result and the third matching fusion result are cascaded into a heart rate-blood pressure principal component multi-scale fusion feature vector, one-dimensional convolution processing and maximum pooling processing are performed on them to obtain the heart rate-blood pressure time series principal component joint feature vector.

[0035] For example, according to the millimeter wave therapeutic apparatus of an embodiment of the present application, the adjustment instruction generation module is used to: input the multimodal temporal significant fusion representation vector of the physiological state into the decoder-based adjustment instruction generator to obtain the parameter adaptive adjustment instruction.

[0036] The embodiment of the present application also provides a method for adjusting a millimeter wave therapeutic apparatus, which comprises:

[0037] receiving a time queue of physiological parameters of a patient, wherein the physiological parameters of the patient include a heart rate value, a blood pressure value, and a skin temperature value;

[0038] transmitting the time queue of the patient's physiological parameters to a parameter adjustment controller;

[0039] Processing the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, wherein the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body;

[0040] The time queue of the patient's physiological parameters is processed to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body, including:

[0041] Extracting time series features based on parameter samples from the time queue of the patient's physiological parameters to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector, and a skin temperature time series correlation feature vector;

[0042] Performing principal component feature matching and fusion processing on the heart rate value time series associated feature vector, the blood pressure value time series associated feature vector, and the skin temperature time series associated feature vector to obtain a multi-modal time series significant fusion representation vector of a physiological state;

[0043] Based on the physiological state multimodal temporal significant fusion representation vector, the parameter adaptive adjustment instruction is generated, and the parameter adaptive adjustment instruction includes a decoded value of a recommended power value at a next time point.

[0044] For example, according to the adjustment method of the millimeter wave therapeutic apparatus of the embodiment of the present application, the time queue of the patient's physiological parameters is subjected to time series feature extraction based on parameter samples to obtain a heart rate value time series associated feature vector, a blood pressure value time series associated feature vector, and a skin temperature time series associated feature vector, including:

[0045] Regularizing the time queue of the patient's physiological parameters according to the parameter sample dimension to obtain a time queue of heart rate values, a time queue of blood pressure values, and a time queue of skin temperature values;

[0046] The time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value are respectively sequence-encoded to obtain the heart rate value time series association feature vector, the blood pressure value time series association feature vector and the skin temperature time series association feature vector.

[0047] The present application has at least the following technical effects: Compared with the prior art, the millimeter wave therapy device and its adjustment method of the embodiment of the present application use artificial intelligence technology based on deep learning to perform real-time monitoring and data timing analysis on the patient's heart rate, blood pressure and skin temperature during the treatment process, and respectively mine the timing change rules of various physiological parameters, and perform multimodal data timing feature principal component fusion analysis on multiple physiological parameters to obtain a comprehensive description of the patient's physiological state, and then use this to adaptively adjust the main operating power of the millimeter wave therapy device, which can realize real-time monitoring and personalized treatment of the patient's physiological state, effectively improve the performance of the millimeter wave therapy device, and improve the treatment effect and the intelligence level of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings of the embodiments of the present application are briefly introduced below. Obviously, the drawings described below only relate to some embodiments of the present application, and are not intended to limit the present application.

[0049] Figure 1 A schematic diagram of the structure of a millimeter wave therapeutic apparatus in an embodiment of the present application is shown;

[0050] Figure 2A schematic diagram of the structure of a parameter adjustment controller of a millimeter wave therapeutic apparatus in an embodiment of the present application is shown;

[0051] Figure 3 A flow chart showing a method for adjusting a millimeter wave therapeutic apparatus in an embodiment of the present application is shown;

[0052] Figure 4 A diagram showing an application scenario of the millimeter wave therapy device in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0053] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work also fall within the scope of protection of the present application.

[0054] The terms used in this specification are those common terms currently widely used in the art in consideration of the functions of the present application, but these terms may vary according to the intentions of those skilled in the art, precedents, or new technologies in the art. In addition, specific terms may be selected, and in this case, their detailed meanings will be described in the detailed description of the present application. Therefore, the terms used in the specification should not be understood as simple names, but rather as a general description based on the meaning of the terms and the present application.

[0055] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules can be used and run on the user terminal and / or server. The modules are only illustrative, and different aspects of the system and method can use different modules.

[0056] Flowcharts are used in the present application to illustrate the operations performed by the system according to the embodiments of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps may be processed in reverse order or simultaneously as required. At the same time, other operations may also be added to these processes, or a certain step or several steps of operations may be removed from these processes.

[0057] In addition, the application architecture diagram in the embodiment of the present application is intended to more clearly illustrate the technical solution in the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. Of course, for other application architectures and business applications, the technical solution provided in the embodiment of the present application is also applicable to similar problems.

[0058] The following is a non-restrictive description of the millimeter wave therapy device provided according to at least one embodiment of the present application through several examples or embodiments. As described below, different features in these specific examples or embodiments can be combined with each other without conflicting with each other to obtain new examples or embodiments, and these new examples or embodiments also fall within the scope of protection of the present application.

[0059] In response to the above-mentioned technical problems, the technical concept of the present application is to use artificial intelligence technology based on deep learning to conduct real-time monitoring and data time series analysis of the patient's heart rate, blood pressure and skin temperature during the treatment process, and to respectively explore the time series change rules of various physiological parameters, and to perform principal component fusion analysis of multimodal data time series features of multiple physiological parameters to obtain a comprehensive description of the patient's physiological state, and then use this to adaptively adjust the main operating power of the millimeter wave therapy device, which can realize real-time monitoring and personalized treatment of the patient's physiological state, effectively improve the performance of the millimeter wave therapy device, and improve the treatment effect and the intelligence level of the equipment.

[0060] Based on this, the present application proposes a millimeter wave therapeutic device, such as Figure 1 As shown, it includes: a millimeter wave therapeutic device body 810, a data acquisition module 820, a data transmission module 830 and a parameter adjustment controller 840; the data acquisition module 820 is used to receive a time queue of a patient's physiological parameters, wherein the patient's physiological parameters include a heart rate value, a blood pressure value and a skin temperature value; the data transmission module 830 is used to transmit the time queue of the patient's physiological parameters to the parameter adjustment controller 840; the parameter adjustment controller 840 is used to process the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body 810.

[0061] Among them, heart rate, blood pressure and skin temperature are important physiological indicators that change in real time during the treatment process, and can reflect the patient's response and adaptability to millimeter wave treatment. Therefore, in the technical solution of this application, the patient's treatment effect and physical condition are evaluated by collecting the patient's physiological parameters in real time, and the working parameters of the millimeter wave treatment device are adjusted in real time accordingly.

[0062] Accordingly, if Figure 2As shown, the parameter adjustment controller 840 includes: a patient physiological parameter timing encoding module 841, which is used to perform timing feature extraction based on parameter samples on the time queue of the patient's physiological parameters to obtain a heart rate value timing association feature vector, a blood pressure value timing association feature vector and a skin temperature timing association feature vector; a multimodal feature matching and fusion module 842, which is used to perform principal component feature matching and fusion processing on the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector to obtain a physiological state multimodal timing significant fusion representation vector; an adjustment instruction generation module 843, which is used to generate the parameter adaptive adjustment instruction based on the physiological state multimodal timing significant fusion representation vector, and the parameter adaptive adjustment instruction includes a decoded value of a recommended power value at the next time point.

[0063] Specifically, in order to ensure the pertinence and accuracy of subsequent data processing, the present application first regularizes the time queue of the patient's physiological parameters according to the parameter sample dimension, splits the original parameter set into three independent time queues of heart rate value, blood pressure value and skin temperature value, and maintains the time synchronization of the three to facilitate subsequent multimodal data analysis.

[0064] Since the patient's heart rate, blood pressure and skin temperature are dynamic data that change continuously as the treatment progresses. Therefore, in order to fully capture the dynamic change law of the patient's physiological state, this application uses a sequence encoder based on the Bi-LSTM model to perform time series encoding on the time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value, respectively, and uses the bidirectional long-term memory network structure of the Bi-LSTM model to model the bidirectional long-term dependency in the time series data, thereby mining the dynamic change pattern of the patient's heart rate, blood pressure and skin temperature values ​​in the time dimension, so as to obtain the heart rate value time series association feature vector, blood pressure value time series association feature vector and skin temperature time series association feature vector.

[0065] Correspondingly, the patient physiological parameter timing encoding module 841 includes: a data regularization unit, which is used to regularize the time queue of the patient physiological parameters according to the parameter sample dimension to obtain the time queue of heart rate values, the time queue of blood pressure values ​​and the time queue of skin temperature values; a parameter timing encoding unit, which is used to sequence encode the time queue of heart rate values, the time queue of blood pressure values ​​and the time queue of skin temperature values ​​respectively to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector.

[0066] Among them, the parameter timing encoding unit is used to: input the time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value into a sequence encoder based on the Bi-LSTM model to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector.

[0067] Next, considering that during the treatment process, the patient's heart rate and blood pressure changes usually have a strong correlation, and the acceleration or slowing of the heart rate is often accompanied by an increase or decrease in blood pressure. Therefore, when performing multi-source data fusion analysis, the present application first performs a joint analysis process on the heart rate value time series correlation feature vector and the blood pressure value time series correlation feature vector to emphasize the correlation between heart rate and blood pressure. Specifically, the present application introduces a matching fusion network based on the granularity of the characteristic principal component to realize this joint analysis process. The network first standardizes the heart rate value time series correlation feature vector and the blood pressure value time series correlation feature vector to eliminate the influence of different dimensions and magnitudes. Then, the principal component analysis algorithm is used to calculate the sample covariance and extract the principal component features of the standardized heart rate value time series correlation feature vector and the blood pressure value time series correlation feature vector, and the main components are mined and redundant information is removed to generate a set of heart rate principal component feature vectors and a set of blood pressure principal component feature vectors. Then, based on the maximum approximate query matching between feature vectors, the best matching pair is found by calculating and comparing the similarity between the two principal component feature vector sets, and the significant correlation features between the two are extracted by performing fine-grained semantic joint analysis on each best matching pair. Finally, the feature joint results of each best matching pair are cascaded to aggregate multiple core component joint features of the patient's heart rate features and blood pressure features to generate a global heart rate-blood pressure time series association significant fusion representation vector.

[0068] Then, on this basis, the present application further uses the matching fusion network based on the feature principal component as the granularity to perform principal component feature extraction and feature query matching fusion processing on the heart rate-blood pressure time series association significant fusion representation vector and the skin temperature time series association feature vector. While retaining the key information of heart rate and blood pressure, the additional dimension of skin temperature is introduced to more comprehensively reflect the patient's physiological state, so as to obtain a multimodal time series significant fusion representation vector of the physiological state.

[0069] Correspondingly, the multimodal feature matching fusion module 842 includes: a heart rate and blood pressure time series association fusion unit, which is used to input the heart rate value time series association feature vector and the blood pressure value time series association feature vector into a matching fusion network based on feature principal components as the granularity to obtain a heart rate-blood pressure time series association significant fusion representation vector; and a full modal fusion representation unit, which is used to input the heart rate-blood pressure time series association significant fusion representation vector and the skin temperature time series association feature vector into the matching fusion network based on feature principal components as the granularity to obtain the physiological state multimodal time series significant fusion representation vector.

[0070] Among them, the heart rate and blood pressure time series association fusion unit includes: a standardization subunit, which is used to standardize the heart rate value time series association feature vector and the blood pressure value time series association feature vector to obtain a standardized heart rate value time series association feature vector and a standardized blood pressure value time series association feature vector; a sample covariance calculation subunit, which is used to respectively calculate the sample covariance matrices of the standardized heart rate value time series association feature vector and the standardized blood pressure value time series association feature vector to obtain a heart rate sample covariance matrix and a blood pressure sample covariance matrix; a principal component feature extraction subunit, which is used to perform matrix decomposition-based feature vector extraction on the heart rate sample covariance matrix and the blood pressure sample covariance matrix to obtain a set of heart rate principal component feature vectors and a blood pressure sample covariance matrix. A set of principal component feature vectors; a query matching subunit, used to input the set of heart rate principal component feature vectors and the set of blood pressure principal component feature vectors into a maximum approximate query matching network to obtain a set of best matching pairs of heart rate principal component feature vectors and blood pressure principal component feature vectors; a fine-grained semantic joint subunit, used to input each best matching pair in the set of best matching pairs of heart rate principal component feature vectors and blood pressure principal component feature vectors into a semantic fine-grained joint analysis module to obtain a set of heart rate-blood pressure time series principal component joint feature vectors; a principal component matching feature global aggregation subunit, used to cascade the set of heart rate-blood pressure time series principal component joint feature vectors to obtain the heart rate-blood pressure time series association significant fusion representation vector.

[0071] Specifically, the query matching subunit is used to: calculate the cosine similarity between the i-th heart rate principal component eigenvector in the set of heart rate principal component eigenvectors and each blood pressure principal component eigenvector in the set of blood pressure principal component eigenvectors to obtain a set of cosine similarities; and form the best matching pair with the blood pressure principal component eigenvector corresponding to the maximum value in the set of cosine similarities.

[0072] Specifically, the fine-grained semantic union subunit is used to: calculate the point-by-point addition, point-by-point subtraction and point-by-point multiplication between the heart rate principal component feature vector and the blood pressure principal component feature vector in the best matching pair to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; after cascading the first matching fusion result, the second matching fusion result and the third matching fusion result into a heart rate-blood pressure principal component multi-scale fusion feature vector, perform one-dimensional convolution processing and maximum pooling processing on it to obtain the heart rate-blood pressure time series principal component joint feature vector.

[0073] In a specific example, the heart rate and blood pressure time series association fusion unit is used to: fuse the heart rate value time series association feature vector and the blood pressure value time series association feature vector using the following principal component query matching formula to obtain the heart rate-blood pressure time series association significant fusion representation vector, wherein the principal component query matching formula is:

[0074]

[0075]

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087] in, represents the temporal correlation feature vector of the heart rate value, and represent the mean and standard deviation of the heart rate value time series associated feature vector respectively, represents the time series correlation feature vector of the normalized heart rate value, represents the blood pressure value time series correlation feature vector, and represent the mean and standard deviation of the blood pressure value time series correlation feature vector respectively, represents the time series correlation feature vector of the standardized blood pressure value, The value of is the length of the heart rate value time series associated feature vector, The value of is the length of the blood pressure value time series associated feature vector, represents the heart rate sample covariance matrix, represents the blood pressure sample covariance matrix, represents a matrix consisting of a set of principal component eigenvectors of the heart rate obtained by performing eigenvalue decomposition on the covariance matrix of the heart rate samples, represents a diagonal matrix consisting of a set of eigenvalues ​​of the heart rate principal components obtained by performing eigenvalue decomposition on the heart rate sample covariance matrix, , , and Respectively represent the first, second, and third and heart rate principal component eigenvectors, , and respectively represent the first, second and third eigenvalues ​​of the heart rate principal component The principal component eigenvalues ​​of heart rate, The value of is the number of eigenvalues ​​of the heart rate principal component, represents a matrix consisting of a set of blood pressure principal component eigenvectors obtained by performing eigenvalue decomposition on the blood pressure sample covariance matrix, represents a diagonal matrix consisting of a set of blood pressure principal component eigenvalues ​​obtained by performing eigenvalue decomposition on the blood pressure sample covariance matrix, , , , and Respectively represent the first, second, and third , and The principal component eigenvectors of blood pressure, , and respectively represent the first, second and third eigenvalues ​​of the blood pressure principal component The principal component eigenvalues ​​of blood pressure, is the diag function, which is used to construct a diagonal matrix. represents the inner product, represents the magnitude of the eigenvector, Indicates the index corresponding to the maximum value. Indicates that The index of the blood pressure principal component eigenvector that forms the best matching pair of heart rate principal component eigenvectors, Indicates cascade operation, represents the maximum pooling operation, represents a one-dimensional convolution operation, represents dot product, Indicates point addition, Indicates point reduction, Indicates heart rate-blood pressure time series principal component joint eigenvectors, The representation vector represents the significant fusion of the heart rate-blood pressure temporal association.

[0088] Finally, the multimodal temporal significant fusion representation vector of the physiological state is input into the decoder-based adjustment instruction generator to obtain the parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction includes the decoded value of the recommended power value at the next time point. In the technical solution of the present application, the decoder-based adjustment instruction generator learns the mapping relationship between the different physiological states of the patient and the operating parameters of the therapeutic device through training, so that it can dynamically generate appropriate therapeutic device parameter adjustment instructions based on the patient's real-time physiological state information contained in the multimodal temporal significant fusion representation vector of the physiological state, ensuring that the operating power of the therapeutic device matches the patient's physiological state to achieve personalized treatment.

[0089] Correspondingly, the adjustment instruction generation module 843 is used to: input the physiological state multimodal temporal significant fusion representation vector into the decoder-based adjustment instruction generator to obtain the parameter adaptive adjustment instruction.

[0090] Preferably, considering that the heart rate-blood pressure temporal association significant fusion representation vector and the skin temperature temporal association feature vector respectively represent the significant fusion features of the heart rate-blood pressure temporal association based on feature principal component query and the skin temperature local temporal context association features, when performing significant fusion based on feature principal component query matching, there will be insufficient feature representation of long-distance temporal feature fusion caused by the imbalance of feature query matching significance under the difference in local temporal feature context expression, affecting the accuracy of the parameter adaptive adjustment instructions obtained by the adjustment instruction generator based on the decoder input.

[0091] Therefore, when the multimodal temporal significant fusion representation vector of the physiological state is input into the adjustment instruction generator based on the decoder to obtain the parameter adaptive adjustment instruction, the multimodal temporal significant fusion representation vector of the physiological state is optimized, including the steps of:

[0092] Arranging the eigenvalues ​​of the physiological state multimodal temporal significant fusion representation vector in ascending order to form a physiological state multimodal temporal significant fusion sequential encoding vector;

[0093] In response to the physiological state multimodal temporal significant fusion sequence encoding vector The eigenvalue and The absolute value of the difference between the eigenvalues ​​is less than or equal to the distance difference hyperparameter , calculate the The eigenvalues ​​are similar to the The weighted sum between the eigenvalues ​​is the optimized Eigenvalue;

[0094] Calculate the square root of the sum of squares of all eigenvalues ​​of the significant fusion representation vector of the multimodal time series of the physiological state:

[0095]

[0096] in, The first part of the multimodal temporal significant fusion representation vector of the physiological state is represented by Eigenvalues, The length of the multimodal temporal significant fusion representation vector representing the physiological state, The square root of the sum of squares of all eigenvalues ​​of the multimodal time series significant fusion representation vector representing the physiological state;

[0097] The square root of the sum of the squares of all eigenvalues ​​of the physiological state multimodal time series significant fusion representation vector is multiplied by 2 and then divided by the square of the length of the physiological state multimodal time series significant fusion representation vector to obtain the physiological state multimodal time series significant fusion space primitive value ;

[0098] In response to the physiological state multimodal temporal significant fusion sequence encoding vector The eigenvalue and The absolute value of the difference between the eigenvalues ​​is greater than the distance difference hyperparameter , multiply the physiological state multimodal time series significant fusion space primitive value by the first After the eigenvalue, calculate the product with the first The weighted reduction between the eigenvalues ​​is the optimized Eigenvalue;

[0099] On the basis of keeping the first eigenvalue of the multimodal temporal significant fusion sequence encoding vector of the physiological state unchanged, the first eigenvalue of the combined optimization is The eigenvalues ​​are used to obtain the optimized multi-modal temporal significant fusion representation vector of the physiological state.

[0100] Therefore, in order to address the problem that the feature set of the physiological state multimodal time series significant fusion representation vector has insufficient global interactive response representation capability due to the long distance exceeding the predetermined local distribution interval threshold under the predetermined eigenvalue sequential distribution, the high-dimensional feature space primitive representation based on self-inner product fusion of the physiological state multimodal time series significant fusion representation vector is used to capture the complex structure of the global network interaction of its eigenvalues, thereby reconstructing the interactive response relationship between the eigenvalues ​​of the physiological state multimodal time series significant fusion representation vector by simulating the scale-based high-dimensional feature space potential primitives, so as to achieve the encoding reconstruction of the real sequence distribution behavior of the physiological state multimodal time series significant fusion representation vector under long distance, improve the interactive response expression effect of the physiological state multimodal time series significant fusion representation vector, and improve the accuracy of the parameter adaptive adjustment instructions obtained by the adjustment instruction generator based on the decoder input of the physiological state multimodal time series significant fusion representation vector.

[0101] Based on the above embodiments, see Figure 3 As shown in FIG. 1 , it is a flow chart of a method for adjusting a millimeter wave therapeutic apparatus in an embodiment of the present application. Figure 3 As shown, the adjustment method of the millimeter wave therapeutic device according to the embodiment of the present application includes the steps of: S510, receiving a time queue of the patient's physiological parameters, wherein the patient's physiological parameters include a heart rate value, a blood pressure value and a skin temperature value; S520, transmitting the time queue of the patient's physiological parameters to a parameter adjustment controller; S530, processing the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body.

[0102] Among them, in step S530, the time queue of the patient's physiological parameters is processed to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body, including: performing time series feature extraction based on parameter samples on the time queue of the patient's physiological parameters to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector and a skin temperature time series correlation feature vector; performing principal component feature matching and fusion processing on the heart rate value time series correlation feature vector, the blood pressure value time series correlation feature vector and the skin temperature time series correlation feature vector to obtain a physiological state multimodal time series significant fusion representation vector; based on the physiological state multimodal time series significant fusion representation vector, generating the parameter adaptive adjustment instruction, the parameter adaptive adjustment instruction includes a decoded value of the recommended power value at the next time point.

[0103] Specifically, a time series feature extraction based on parameter samples is performed on the time queue of the patient's physiological parameters to obtain a time series associated feature vector of heart rate value, a time series associated feature vector of blood pressure value and a time series associated feature vector of skin temperature, including: performing data regularization on the time queue of the patient's physiological parameters according to the parameter sample dimension to obtain a time queue of heart rate value, a time queue of blood pressure value and a time queue of skin temperature value; and performing sequence encoding on the time queue of heart rate value, the time queue of blood pressure value and the time queue of skin temperature value, respectively, to obtain a time series associated feature vector of heart rate value, a time series associated feature vector of blood pressure value and a time series associated feature vector of skin temperature.

[0104] The specific operations of each step in the adjustment method of the above-mentioned millimeter wave therapeutic device have been referred to above. Figure 1 to Figure 2 The millimeter wave therapy device 800 is described in detail, and therefore, its repeated description will be omitted.

[0105] Figure 4 FIG. 1 is an application scenario diagram of a millimeter wave therapeutic apparatus according to an embodiment of the present application. Figure 4 As shown, in this application scenario, first, a time queue of the patient's physiological parameters is received (for example, Figure 4 D), wherein the patient's physiological parameters include heart rate value, blood pressure value and skin temperature value, and then the time queue of the patient's physiological parameters is input to a server (for example, Figure 4 In S) as shown in , the server can use the adjustment algorithm of the millimeter wave therapeutic device to process the time queue of the patient's physiological parameters to obtain parameter adaptive adjustment instructions.

[0106] Based on the above embodiment, another exemplary embodiment of an electronic device is also provided in the embodiment of the present application. In some possible implementations, the electronic device in the embodiment of the present application may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor can implement the steps of the adjustment method of the millimeter wave therapeutic device in the above embodiment when executing the program.

[0107] The embodiment of the present application also provides a computer-readable storage medium, on which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, the adjustment method of the millimeter wave therapy device according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include a random access memory (RAM) and / or a cache memory (cache), etc. The non-volatile memory may, for example, include a read-only memory (ROM), a hard disk, a flash memory, etc.

[0108] The embodiment of the present application also provides a computer program product or a computer program, which includes computer executable instructions, and the computer executable instructions are stored in a computer readable storage medium. The processor of the computer device reads the computer executable instructions from the computer readable storage medium, and the processor executes the computer executable instructions, so that the computer device executes the adjustment method of the millimeter wave therapy device according to the embodiment of the present application.

[0109] Those skilled in the art will appreciate that the contents disclosed in this application may be subject to various variations and improvements. For example, the various devices or components described above may be implemented by hardware, or by software, firmware, or a combination of some or all of the three.

[0110] In addition, although the present application makes various references to certain units in the system according to embodiments of the present application, any number of different units can be used and run on the client and / or server. The units are only illustrative, and different aspects of the system and method can use different units.

[0111] Those skilled in the art will appreciate that all or part of the steps in the above method can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk. Optionally, all or part of the steps in the above embodiment can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or in the form of software function modules. The present application is not limited to any particular form of combination of hardware and software.

[0112] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It should also be understood that terms such as those defined in common dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined as such herein.

[0113] The above is an explanation of the present application and should not be considered as limiting thereof. Although several exemplary embodiments of the present application have been described, those skilled in the art will readily appreciate that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present application.

Claims

1. A millimeter wave therapeutic apparatus, characterized in that: include: Millimeter wave therapeutic device main body, data acquisition module, data transmission module and parameter adjustment controller; The data acquisition module is used to receive a time queue of the patient's physiological parameters, wherein the patient's physiological parameters include heart rate value, blood pressure value and skin temperature value; The data transmission module is used to transmit the time queue of the patient's physiological parameters to the parameter adjustment controller; The parameter adjustment controller is used to process the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body; Wherein, the parameter adjustment controller comprises: A patient physiological parameter time series encoding module, used for extracting time series features of the time series of the patient physiological parameters based on parameter samples to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector and a skin temperature time series correlation feature vector; A multimodal feature matching and fusion module, used for performing principal component feature matching and fusion processing on the heart rate value time series associated feature vector, the blood pressure value time series associated feature vector and the skin temperature time series associated feature vector to obtain a multimodal time series significant fusion representation vector of a physiological state; The adjustment instruction generation module is used to generate the parameter adaptive adjustment instruction based on the physiological state multimodal time series significant fusion representation vector, and the parameter adaptive adjustment instruction includes a decoded value of the recommended power value at the next time point.

2. The millimeter wave therapeutic apparatus according to claim 1, characterized in that: The patient physiological parameter time series encoding module comprises: A data regularization unit, used for performing data regularization on the time queue of the patient's physiological parameters according to the parameter sample dimension to obtain a time queue of heart rate values, a time queue of blood pressure values, and a time queue of skin temperature values; A parameter timing encoding unit is used to perform sequence encoding on the time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector.

3. The millimeter wave therapeutic apparatus according to claim 2, characterized in that: The parameter timing encoding unit is used to: input the time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value into a sequence encoder based on the Bi-LSTM model to obtain the heart rate value timing association feature vector, the blood pressure value timing association feature vector and the skin temperature timing association feature vector.

4. The millimeter wave therapeutic apparatus according to claim 3, characterized in that: The multimodal feature matching and fusion module includes: A heart rate and blood pressure time series association fusion unit, used for inputting the heart rate value time series association feature vector and the blood pressure value time series association feature vector into a matching fusion network based on the feature principal component as the granularity to obtain a heart rate-blood pressure time series association significant fusion representation vector; The full-modal fusion representation unit is used to input the heart rate-blood pressure time series association significant fusion representation vector and the skin temperature time series association feature vector into the matching fusion network based on the feature principal component as the granularity to obtain the physiological state multi-modal time series significant fusion representation vector.

5. The millimeter wave therapeutic apparatus according to claim 4, characterized in that: The heart rate and blood pressure time series association fusion unit comprises: A standardization subunit, used for performing standardization processing on the heart rate value time series association feature vector and the blood pressure value time series association feature vector to obtain a standardized heart rate value time series association feature vector and a standardized blood pressure value time series association feature vector; A sample covariance calculation subunit, used to calculate the sample covariance matrices of the standardized heart rate value time series associated feature vector and the standardized blood pressure value time series associated feature vector respectively to obtain a heart rate sample covariance matrix and a blood pressure sample covariance matrix; A principal component feature extraction subunit, used for performing matrix decomposition-based feature vector extraction on the heart rate sample covariance matrix and the blood pressure sample covariance matrix to obtain a set of heart rate principal component feature vectors and a set of blood pressure principal component feature vectors; A query matching subunit, used for inputting the set of the heart rate principal component feature vectors and the set of the blood pressure principal component feature vectors into a maximum approximate query matching network to obtain a set of best matching pairs of the heart rate principal component feature vectors and the blood pressure principal component feature vectors; A fine-grained semantic joint subunit, used for inputting each best matching pair in the set of best matching pairs of the heart rate principal component feature vector and the blood pressure principal component feature vector into a semantic fine-grained joint analysis module to obtain a set of heart rate-blood pressure time series principal component joint feature vectors; The principal component matching feature global aggregation subunit is used to cascade the set of the heart rate-blood pressure time series principal component joint feature vectors to obtain the heart rate-blood pressure time series association significant fusion representation vector.

6. The millimeter wave therapeutic apparatus according to claim 5, characterized in that: The query matching subunit is used to: Calculating the cosine similarity between the i-th heart rate principal component eigenvector in the set of heart rate principal component eigenvectors and each blood pressure principal component eigenvector in the set of blood pressure principal component eigenvectors to obtain a set of cosine similarities; The blood pressure principal component eigenvector corresponding to the maximum value in the set of cosine similarities and the i-th heart rate principal component eigenvector form the best matching pair.

7. The millimeter wave therapeutic apparatus according to claim 6, characterized in that: The fine-grained semantic association subunit is used to: Calculate the position point addition, position point subtraction and position point multiplication between the heart rate principal component eigenvector and the blood pressure principal component eigenvector in the best matching pair to obtain a first matching fusion result, a second matching fusion result and a third matching fusion result; After the first matching fusion result, the second matching fusion result and the third matching fusion result are cascaded into a heart rate-blood pressure principal component multi-scale fusion feature vector, one-dimensional convolution processing and maximum pooling processing are performed on them to obtain the heart rate-blood pressure time series principal component joint feature vector.

8. The millimeter wave therapeutic apparatus according to claim 7, characterized in that: The adjustment instruction generation module is used to: input the physiological state multi-modal temporal significant fusion representation vector into the decoder-based adjustment instruction generator to obtain the parameter adaptive adjustment instruction.

9. A method for adjusting a millimeter wave therapeutic apparatus, characterized in that: include: receiving a time queue of physiological parameters of a patient, wherein the physiological parameters of the patient include a heart rate value, a blood pressure value, and a skin temperature value; transmitting the time queue of the patient's physiological parameters to a parameter adjustment controller; Processing the time queue of the patient's physiological parameters to obtain a parameter adaptive adjustment instruction, wherein the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body; The time queue of the patient's physiological parameters is processed to obtain a parameter adaptive adjustment instruction, and the parameter adaptive adjustment instruction is used to control the power value of the millimeter wave therapeutic device body, including: Extracting time series features based on parameter samples from the time queue of the patient's physiological parameters to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector, and a skin temperature time series correlation feature vector; Performing principal component feature matching and fusion processing on the heart rate value time series associated feature vector, the blood pressure value time series associated feature vector, and the skin temperature time series associated feature vector to obtain a multi-modal time series significant fusion representation vector of a physiological state; Based on the physiological state multimodal temporal significant fusion representation vector, the parameter adaptive adjustment instruction is generated, and the parameter adaptive adjustment instruction includes a decoded value of a recommended power value at a next time point.

10. The adjustment method of the millimeter wave therapeutic apparatus according to claim 9, characterized in that: Extracting the time series features of the time series of the patient's physiological parameters based on parameter samples to obtain a heart rate value time series correlation feature vector, a blood pressure value time series correlation feature vector, and a skin temperature time series correlation feature vector, including: Regularizing the time queue of the patient's physiological parameters according to the parameter sample dimension to obtain a time queue of heart rate values, a time queue of blood pressure values, and a time queue of skin temperature values; The time queue of the heart rate value, the time queue of the blood pressure value and the time queue of the skin temperature value are respectively sequence-encoded to obtain the heart rate value time series association feature vector, the blood pressure value time series association feature vector and the skin temperature time series association feature vector.

Images