A biological information capture and storage device

Abstract

The present application relates to the field of biological signal processing, and specifically discloses a biological information capture and storage device, which collects biological signals through a biological information receiving and processing device, and uses a signal processing and analysis technology based on deep learning to decompose and encode the biological signals, thereby automatically judging whether the biological signal time-frequency component is a noise component based on the matching characteristics between multiple decomposed signal time-frequency components, and inversely transforming the remaining biological signal time-frequency components after elimination to obtain a denoised biological signal. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, thereby performing noise reduction processing on the original collected biological signal, providing a higher quality data signal for subsequent signal processing and analysis.

Description

[0001] The present application relates to the field of biological signal processing, and more specifically, to a biological information capture and storage device. Background Art

[0002] Bioinformation reflects the movement state and pattern of organisms and is the product of rhythmic activities at all levels of biological organisms. In organisms, countless cells move in a self-organized manner, gradually evolving from the initial disordered state to an ordered state. The normal operation of life depends on constant metabolism and balance. Only when cells maintain a normal state of movement can bioinformation remain normal, thereby ensuring the normal operation of life through a benign interaction with cells.

[0003] With the development of biotechnology, especially in the fields of genomics, transcriptomics, proteomics, etc., the amount of data generated has increased dramatically. These data usually contain noise and incomplete information and need to be processed and analyzed through filtering technology to extract useful biological information.

[0004] However, traditional biosignal filtering techniques, such as analog filters, cannot effectively handle high-frequency noise in the signal, which may cause signal distortion. In addition, traditional filtering techniques may not be flexible enough when dealing with nonlinear or non-stationary signals because they are usually designed and optimized for specific signal characteristics.

[0005] Therefore, an optimized biological information capture and storage device is desired. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. An embodiment of the present application provides a biological information capture and storage device, which collects biological signals through a biological information receiving and processing device, and uses a signal processing and analysis technology based on deep learning to decompose and encode the biological signals, so as to automatically determine whether the biological signal time-frequency component is a noise component based on the matching characteristics between multiple decomposed signal time-frequency components, and inversely transform the remaining biological signal time-frequency components after elimination to obtain a denoised biological signal. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, so as to perform noise reduction processing on the original collected biological signal, providing a higher quality data signal for subsequent signal processing and analysis.

[0007] According to one aspect of the present application, a biological information capture and storage device is provided, comprising:

[0008] Wherein, the biological information receiving and processing device is electrically connected to the biological information digitizing and processing device, and the biological information digitizing and processing device is electrically connected to the biological information storage device;

[0009] Wherein, the biological information digital processing device comprises:

[0010] A biological signal acquisition module, used to acquire the biological signal collected by the biological information receiving and processing device;

[0011] A signal decomposition module, used for performing multi-level decomposition on the biological signal to obtain multiple biological signal time-frequency components;

[0012] A biological signal encoding module, used for respectively encoding each biological signal time-frequency component of the plurality of biological signal time-frequency components to obtain a plurality of biological signal component time-frequency associated feature vectors;

[0013] A one-way interactive matching module, used to use each biological signal component time-frequency association feature vector of the multiple biological signal component time-frequency association feature vectors as a query feature vector, and calculate its one-way interactive matching result relative to the feature library composed of the multiple biological signal component time-frequency association feature vectors to obtain multiple one-way matching feature vectors as one-way matching features;

[0014] A recognition result generation module, used to obtain multiple recognition results based on the one-way matching feature, wherein the recognition results are used to indicate whether the time-frequency component of the biological signal is a noise component;

[0015] A noise component removal module, used to remove the noise components in the multiple biological signal time-frequency components to obtain the remaining biological signal time-frequency components;

[0016] The noise reduction signal generation module is used to restore the remaining time-frequency components of the biological signal to obtain a noise-reduced biological signal.

[0017] In the above-mentioned biological information capturing and storing device, the signal decomposition module is used to: use wavelet decomposition technology to perform multi-level decomposition on the biological signal to obtain the multiple biological signal time-frequency components.

[0018] In the above-mentioned biological information capture and storage device, the biological signal encoding module is used to: input each biological signal time-frequency component of the multiple biological signal time-frequency components into a signal encoder based on the 1D-CNN model to obtain the time-frequency correlation feature vectors of the multiple biological signal components.

[0019] In the above-mentioned biological information capture and storage device, the one-way interactive matching module includes: a predetermined feature extraction unit, used to extract a predetermined biological signal component time-frequency correlation feature vector from the multiple biological signal component time-frequency correlation feature vectors as a query feature vector; a correlation calculation unit, used to calculate the correlation between the query feature vector and other biological signal component time-frequency correlation feature vectors except the query feature vector to obtain the one-way matching feature vector corresponding to the query feature vector.

[0020] In the above-mentioned biological information capture and storage device, the correlation calculation unit includes: a logarithmic value calculation subunit, which is used to calculate the position-based logarithmic function value with base 2 of each biological signal component time-frequency correlation feature vector in the other biological signal component time-frequency correlation feature vectors to obtain a sequence of biological signal component time-frequency correlation logarithmic function value vectors; a correlation interaction subunit, which is used to calculate the position point multiplication between each biological signal component time-frequency correlation logarithmic function value vector in the sequence of biological signal component time-frequency correlation logarithmic function value vectors and the query feature vector to obtain a sequence of biological signal component time-frequency correlation interaction feature vectors; a global mean calculation subunit, which is used to calculate the global mean between the positional feature values ​​of each biological signal component time-frequency correlation interaction feature vector in the sequence of biological signal component time-frequency correlation interaction feature vectors to obtain the one-way matching feature vector.

[0021] In the above-mentioned biological information capture and storage device, the recognition result generation module is used to: input the multiple one-way matching feature vectors into a component identifier based on a classifier to obtain the multiple recognition results.

[0022] In the above-mentioned biological information capture and storage device, the recognition result generation module is used to: use the fully connected layer of the classifier-based component identifier to fully connect the one-way matching feature vector to obtain a one-way matching fully connected encoded feature vector; input the one-way matching fully connected encoded feature vector into the Softmax classification function of the classifier-based component identifier to obtain the probability value of the one-way matching feature vector belonging to each classification label, and the classification label includes a value for indicating that the biological signal time-frequency component is a noise component and a value for indicating that the biological signal time-frequency component is not a noise component; and determine the classification label corresponding to the largest of the probability values ​​as the recognition result.

[0023] In the above-mentioned biological information capturing and storing device, the noise reduction signal generating module is used to: input the remaining biological signal time-frequency components into a signal restorer based on inverse wavelet transform to obtain the noise-reduced biological signal.

[0024] Compared with the prior art, the bio-information capture and storage device provided by the present application collects bio-signals through a bio-information receiving and processing device, and uses a signal processing and analysis technology based on deep learning to decompose and encode the bio-signals, thereby automatically determining whether the bio-signal time-frequency components are noise components based on the matching characteristics between multiple decomposed signal time-frequency components, and inversely transforming the remaining bio-signal time-frequency components after elimination to obtain a denoised bio-signal. In this way, the automated bio-signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, thereby performing noise reduction processing on the original collected bio-signals, providing a higher quality data signal for subsequent signal processing and analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0026] Figure 1 4 is a block diagram of a biological information capture and storage device according to an embodiment of the present application.

[0027] Figure 2 4 is a block diagram of a biological information digitization processing device in a biological information capture and storage device according to an embodiment of the present application.

[0028] Figure 3 Schematic diagram of the structure of a biological information digitization processing device in a biological information capture and storage device according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0030] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0031] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0032] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0033] Bioinformation reflects the movement state and pattern of organisms and is the product of rhythmic activities at all levels of biological organisms. In organisms, countless cells move in a self-organized manner, gradually evolving from the initial disordered state to an ordered state. The normal operation of life depends on constant metabolism and balance. Only when cells maintain a normal state of movement can bioinformation remain normal, thereby ensuring the normal operation of life through a benign interaction with cells.

[0034] With the development of biotechnology, especially in the fields of genomics, transcriptomics, proteomics, etc., the amount of data generated has increased dramatically. These data usually contain noise and incomplete information and need to be processed and analyzed through filtering technology to extract useful biological information.

[0035] Based on this, the present application proposes a biological information capture and storage device, which includes: a biological information receiving and processing device, a biological information digitization processing device and a biological information storage device, wherein the biological information receiving and processing device is electrically connected to the biological information digitization processing device, and the biological information digitization processing device is electrically connected to the biological information storage device.

[0036] However, traditional biosignal filtering techniques, such as analog filters, cannot effectively handle high-frequency noise in the signal, which may cause signal distortion. In addition, traditional filtering techniques may not be flexible enough when dealing with nonlinear or non-stationary signals because they are usually designed and optimized for specific signal characteristics.

[0037] Therefore, in response to the above technical problems, the technical concept of the present application is to collect biological signals through a biological information receiving and processing device, and use signal processing and analysis technology based on deep learning to decompose and encode the biological signals, so as to automatically determine whether the biological signal time-frequency component is a noise component based on the matching characteristics between multiple decomposed signal time-frequency components, and inversely transform the remaining biological signal time-frequency components after elimination to obtain the denoised biological signal. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, so as to perform noise reduction processing on the original collected biological signal, providing a higher quality data signal for subsequent signal processing and analysis.

[0038] Figure 1 FIG. 1 is a block diagram of a biological information capture and storage device according to an embodiment of the present application. Figure 1 As shown, in the biological information capturing and storing device 100, it includes: a biological information receiving and processing device 110, a biological information digitizing and processing device 120, and a biological information storing device 130. Specifically, the biological information receiving and processing device is electrically connected to the biological information digitizing and processing device, and the biological information digitizing and processing device is electrically connected to the biological information storing device.

[0039] Specifically, the biological information receiving and processing device 110 is a highly complex device that plays a vital role in the body. In detail, the biological information receiving and processing device can receive and process various biological signals, such as chemical signals, electrical signals and physical signals of the organism, and these signals are essential for maintaining the normal progress of life activities. Specifically, in an embodiment of the present application, the biological information receiving and processing device includes: an antenna, a biological information collection and processing integrated circuit board, a first insulating rubber seat, a first filling rubber layer and a plurality of first pin connectors, the antenna is arranged on the biological information collection and processing integrated circuit board, the biological information collection and processing integrated circuit board is fixed on the first insulating rubber seat, the first filling rubber layer is filled in the gap between the biological information collection and processing integrated circuit board and the first insulating rubber seat, and each of the first pin connectors vertically passes through the first insulating rubber seat and the top end is electrically connected to the biological information collection and processing integrated circuit board.

[0040] Figure 2 4 is a block diagram of a biological information digitization processing device in a biological information capture and storage device according to an embodiment of the present application. Figure 3 FIG. 1 is a schematic diagram of the structure of the biological information digital processing device in the biological information capture and storage device according to an embodiment of the present application. Figure 2 and Figure 3As shown, the biological information digital processing device 120 includes: a biological signal acquisition module 121, used to obtain the biological signal collected by the biological information receiving and processing device; a signal decomposition module 122, used to perform multi-level decomposition on the biological signal to obtain multiple biological signal time-frequency components; a biological signal encoding module 123, used to perform signal encoding on each of the multiple biological signal time-frequency components to obtain multiple biological signal component time-frequency correlation feature vectors; a one-way interactive matching module 124, used to use each biological signal component time-frequency correlation feature vector in the multiple biological signal component time-frequency correlation feature vectors as a query A query feature vector is used to calculate the one-way interactive matching result relative to the feature library composed of the time-frequency associated feature vectors of the multiple biological signal components to obtain multiple one-way matching feature vectors as one-way matching features; a recognition result generation module 125 is used to obtain multiple recognition results based on the one-way matching features, and the recognition results are used to indicate whether the biological signal time-frequency component is a noise component; a noise component elimination module 126 is used to eliminate the noise components in the multiple biological signal time-frequency components to obtain the remaining biological signal time-frequency components; a noise reduction signal generation module 127 is used to perform signal restoration on the remaining biological signal time-frequency components to obtain a denoised biological signal.

[0041] Specifically, in an embodiment of the present application, the biological information digitization processing device includes: a biological information digitization processing integrated circuit board, a second insulating rubber seat, a second filling rubber layer and a plurality of second pin connectors, the biological information digitization processing integrated circuit board is fixed on the second insulating rubber seat and electrically connected to the plurality of first pin connectors, the second filling rubber layer is filled in the gap between the biological information digitization processing integrated circuit board and the second insulating rubber seat, each of the second pin connectors vertically passes through the second insulating rubber seat and the top end is electrically connected to the biological information digitization processing integrated circuit board.

[0042] In the embodiment of the present application, the biological signal acquisition module 121 is used to obtain the biological signal collected by the biological information receiving and processing device. It should be understood that the biological signal refers to various signals emitted by an organism that can reflect its physiological, biochemical state and function, and participate in regulating various physiological activities of the organism. Based on this, in order to filter, reduce noise, extract features and restore signals more accurately in the future. Based on this, in the technical solution of the present application, it is first necessary to collect the biological signals collected by the biological information receiving and processing device.

[0043] In an embodiment of the present application, the signal decomposition module 122 is used to perform multi-level decomposition on the biological signal to obtain multiple biological signal time-frequency components. Specifically, in an embodiment of the present application, the signal decomposition module is used to: use wavelet decomposition technology to perform multi-level decomposition on the biological signal to obtain the multiple biological signal time-frequency components. Accordingly, it is considered that different local signal fragments in the biological signal have different time-frequency feature information. Wavelet decomposition is a time-frequency analysis method that can provide better time domain and frequency domain resolution, and can decompose the signal into different frequency components of signal and noise. Based on this, in the technical solution of the present application, the biological signal is decomposed into multiple levels using wavelet decomposition technology to obtain multiple biological signal time-frequency components. In other words, wavelet decomposition of the biological signal can decompose the biological signal more finely in the time domain and frequency domain into multiple sub-signals of different frequency ranges, so as to better capture the time-frequency characteristics of the local signal to distinguish the noise component and improve the detailed time-frequency information and characteristics of the biological signal.

[0044] In an embodiment of the present application, the biosignal encoding module 123 is used to perform signal encoding on each of the multiple biosignal time-frequency components to obtain multiple biosignal component time-frequency correlation feature vectors. Specifically, in an embodiment of the present application, the biosignal encoding module is used to: input each of the multiple biosignal time-frequency components into a signal encoder based on a 1D-CNN model to obtain the multiple biosignal component time-frequency correlation feature vectors. It should be understood that considering that each of the multiple biosignal time-frequency components expresses the local feature information of the time-frequency signal within different frequency ranges, and considering that the 1D-CNN model is suitable for processing time series data, it can effectively capture the local features in the signal. Therefore, in the technical solution of the present application, each of the multiple biosignal time-frequency components is input into a signal encoder based on a 1D-CNN model to capture and extract the correlation between the local time-frequency of each biosignal time-frequency feature in each biosignal time-frequency component, thereby obtaining multiple biosignal component time-frequency correlation feature vectors.

[0045] In the embodiment of the present application, the one-way interactive matching module 124 is used to use each biological signal component time-frequency association feature vector in the multiple biological signal component time-frequency association feature vectors as a query feature vector, and calculate its one-way interactive matching result relative to the feature library composed of the multiple biological signal component time-frequency association feature vectors to obtain multiple one-way matching feature vectors as one-way matching features. It should be understood that considering that the multiple biological signal component time-frequency association feature vectors contain biological signal component time-frequency association feature vectors that are similar to other biological signal component time-frequency association features, there are also noise components that differ greatly from the overall biological signal component time-frequency association features. Based on this, in order to accurately and quickly screen out biological signal noise components that differ greatly from the overall component features to improve the quality of biological signal filtering, in the technical solution of the present application, each biological signal component time-frequency association feature vector in the multiple biological signal component time-frequency association feature vectors is used as a query feature vector, and its one-way interactive matching result relative to the feature library composed of the multiple biological signal component time-frequency association feature vectors is calculated to obtain multiple one-way matching feature vectors. In detail, first, a predetermined biological signal component time-frequency correlation feature vector is extracted from the multiple biological signal component time-frequency correlation feature vectors as a query vector to specifically find the feature most relevant to the biological signal component time-frequency correlation feature vector, and then the correlation between the query feature vector and other biological signal component time-frequency correlation feature vectors is calculated to evaluate the time-frequency similarity of the biological signal components therebetween, which helps to determine the degree of matching between the query feature vector and the other feature vectors, improves the efficiency of time-frequency feature matching of biological signal components, and provides an important reference for subsequent analysis and identification of noise components.

[0046] Specifically, in an embodiment of the present application, the one-way interactive matching module includes: a predetermined feature extraction unit, used to extract a predetermined biological signal component time-frequency correlation feature vector from the multiple biological signal component time-frequency correlation feature vectors as a query feature vector; a correlation calculation unit, used to calculate the correlation between the query feature vector and other biological signal component time-frequency correlation feature vectors other than the query feature vector to obtain the one-way matching feature vector corresponding to the query feature vector.

[0047] More specifically, in an embodiment of the present application, the correlation calculation unit includes: a logarithmic value calculation subunit, which is used to calculate the position-based logarithmic function value with base 2 of each biological signal component time-frequency correlation feature vector in the other biological signal component time-frequency correlation feature vectors to obtain a sequence of biological signal component time-frequency correlation logarithmic function value vectors; a correlation interaction subunit, which is used to calculate the position-based multiplication between each biological signal component time-frequency correlation logarithmic function value vector in the sequence of biological signal component time-frequency correlation logarithmic function value vectors and the query feature vector to obtain a sequence of biological signal component time-frequency correlation interaction feature vectors; a global mean calculation subunit, which is used to calculate the global mean between the positional feature values ​​of each biological signal component time-frequency correlation interaction feature vector in the sequence of biological signal component time-frequency correlation interaction feature vectors to obtain the one-way matching feature vector.

[0048] In the embodiment of the present application, specifically, the correlation calculation unit is used to calculate the correlation between the query feature vector and the time-frequency correlation feature vectors of other biological signal components other than the query feature vector using the following interaction formula to obtain a one-way matching feature vector; wherein the interaction formula is:

[0049] in, and are the query feature vector and the value of each position in the time-frequency association feature vector of each biological signal component, respectively, is the length of the query feature vector and the time-frequency correlation feature vector of each of the biosignal components, represents the eigenvalues ​​of each position in the one-way matching feature vector, Represents the value of the logarithmic function with base 2.

[0050] In an embodiment of the present application, the recognition result generation module 125 is used to obtain multiple recognition results based on the one-way matching features, and the recognition results are used to indicate whether the time-frequency components of the biological signal are noise components. Specifically, in an embodiment of the present application, the recognition result generation module is used to: input the multiple one-way matching feature vectors into a component identifier based on a classifier to obtain the multiple recognition results. It should be understood that the classification processing is performed by using the one-way matching features obtained by one-way interactive matching of the multiple biological signal component time-frequency associated feature vectors, so as to automatically determine whether the biological signal time-frequency components are noise components based on the matching features between the multiple decomposed signal time-frequency components. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, thereby improving the accuracy and quality of signal processing.

[0051] More specifically, in an embodiment of the present application, the recognition result generation module is used to: use the fully connected layer of the classifier-based component identifier to fully connect the one-way matching feature vector to obtain a one-way matching fully connected encoded feature vector; input the one-way matching fully connected encoded feature vector into the Softmax classification function of the classifier-based component identifier to obtain the probability value of the one-way matching feature vector belonging to each classification label, and the classification label includes a value for indicating that the time-frequency component of the biological signal is a noise component and a value for indicating that the time-frequency component of the biological signal is not a noise component; and determine the classification label corresponding to the largest of the probability values ​​as the recognition result.

[0052] In particular, here, for the time series distribution differences of the time-frequency associated feature vectors of the multiple biological signal components after the one-dimensional convolution extraction of the source time-frequency distribution differences of the multiple biological signal time-frequency components, after the one-way interactive matching relative to the feature library composed of the time-frequency associated feature vectors of the multiple biological signal components, the multiple one-way matching feature vectors will also have the problem of insufficient distribution aggregation of the class one-way matching based on the classifier, and will make the one-way matching logic convergence of the class probability density distribution of the in-class features and the out-of-class features after the clustering operation inconsistent.

[0053] Preferably, inputting the multiple one-way matching feature vectors into a component identifier based on a classifier to obtain multiple recognition results includes: cascading the multiple one-way matching feature vectors into a one-way matching cascade feature vector; performing a clustering operation on the one-way matching cascade feature vector, for example, performing a clustering operation based on the distance between eigenvalues, and determining the number of in-class features and the number of out-of-class features; dividing the total number of eigenvalues ​​of the one-way matching cascade feature vector by the number of in-class features and the number of out-of-class features respectively to obtain a class importance value and an out-of-class proportion value, and calculating the inverse of the class importance value to obtain a class constraint value; calculating each eigenvalue of the one-way matching cascade feature vector with the class importance value as an index A power function of a number is generated, and after adding an exponential value with a natural function as the base and the class constraint value as the exponent, it is multiplied by the out-of-class proportion value to obtain a one-way matching cascade modulation vector; the one-way matching cascade feature vector is point multiplied with the class importance value to obtain a one-way matching cascade ontology vector; the weighted sum of the one-way matching cascade modulation vector and the one-way matching cascade ontology vector is calculated with a weight hyperparameter to obtain an optimized one-way matching cascade feature vector; the optimized one-way matching cascade feature vector is split into multiple optimized one-way matching feature vectors; the multiple optimized one-way matching feature vectors are input into the classifier-based component identifier to obtain the multiple recognition results.

[0054] Specifically, in this preferred embodiment, the multiple one-way matching feature vectors are input into a component identifier based on a classifier to obtain multiple recognition results, including:

[0055] The multiple one-way matching feature vectors are optimized to obtain multiple optimized one-way matching feature vectors. The process formula is expressed as follows:

[0056]

[0057] in, is the one-way matching cascade feature vector, is the total number of eigenvalues ​​of the one-way matching cascade eigenvector, is the number of intra-class features of the one-way matching cascade feature vector, It is the point multiplication by position. is the weight hyperparameter, is added by position, It means to perform exponential operation on the eigenvalues ​​at each position in the eigenvector. are the multiple optimized one-way matching feature vectors;

[0058] The multiple optimized one-way matching feature vectors are input into the classifier-based component identifier to obtain the multiple identification results.

[0059] Therefore, the applicant of the present application performs clustering operation on the one-way matching cascade feature vector obtained by aggregating the multiple one-way matching feature vectors, and further performs responsive modulation with the out-of-class scale factor by combining the clustering importance metric of the feature value of the one-way matching cascade feature vector with the class constraint metric of the clustering operation, and establishes the simplicity and effectiveness of the feature purification of the feature set of the one-way matching cascade feature vector based on the clustering scaling represented by the feature ontology of the one-way matching cascade feature vector, and suppresses the inconsistency of the one-way matching logic convergence of the class probability density distribution caused by the clustering operation, so as to improve the consistency of the multiple recognition results obtained by the component identifier based on the classifier when the multiple one-way matching feature vectors are input, thereby improving the recognition accuracy of the noise component and improving the signal quality of the finally obtained denoised biological signal. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, so as to perform denoising processing on the original collected biological signal, and provide a higher quality data signal for subsequent signal processing and analysis.

[0060] In an embodiment of the present application, the noise component removal module 126 is used to remove the noise components in the multiple biological signal time-frequency components to obtain the remaining biological signal time-frequency components. Accordingly, considering that the noise components will interfere with the accurate recognition and analysis of biological signals, in order to improve the recognition accuracy of biological signals, so as to more accurately analyze and understand the characteristics and properties of biological signals, in the technical solution of the present application, the noise components in the multiple biological signal time-frequency components are removed to obtain the remaining biological signal time-frequency components. That is, the noise component will introduce unnecessary interference and affect the correct understanding and interpretation of the biological signal, and by removing the noise component, such interference can be reduced, making the analysis results of the biological signal clearer and more reliable.

[0061] In an embodiment of the present application, the noise reduction signal generation module 127 is used to perform signal restoration on the remaining time-frequency components of the biological signal to obtain the biological signal after noise reduction. Specifically, in an embodiment of the present application, the noise reduction signal generation module inputs the remaining time-frequency components of the biological signal into a signal restorer based on an inverse wavelet transform to obtain the biological signal after noise reduction. It should be understood that in order to minimize the loss of biological information and retain important features and information in the biological signal, in the technical solution of the present application, the remaining time-frequency components of the biological signal are input into a signal restorer based on an inverse wavelet transform to obtain the biological signal after noise reduction. Specifically, the inverse wavelet transform can help convert the time-frequency components of the biological signal that has been subjected to noise reduction processing back to the original signal domain, thereby restoring the original quality and characteristics of the signal, restoring the true information of the biological signal, and reducing the distortion introduced by the noise reduction processing.

[0062] It is worth mentioning that the signal restorer based on inverse wavelet transform is a signal processing tool that uses the inverse process of wavelet transform to reconstruct or restore the signal. In detail, in biological signal processing, wavelet transform is able to decompose biological signals into sub-signals of different scales, which can reveal different characteristics of the signal. The inverse wavelet transform is the reverse operation of this process, which can recombine the time-frequency components of the biological signals after noise reduction to restore the biological signals after noise reduction. In particular, the signal restorer based on inverse wavelet transform has applications in many fields, such as medical imaging, audio processing, seismic data analysis, etc., which helps professionals analyze and interpret data more accurately.

[0063] Specifically, the main function of the bio-information storage device 130 is to preserve bio-information for a long time for future scientific research and medical applications. In particular, the bio-information storage device can protect biological samples from environmental factors such as oxygen, temperature and microorganisms. Specifically, in an embodiment of the present application, the bio-information storage device includes: a bio-information storage integrated circuit board, a third insulating rubber seat and a third filling rubber layer, the bio-information storage integrated circuit board is fixed on the third insulating rubber seat and is electrically connected to the plurality of second pin connectors, and the third filling rubber layer is filled in the gap between the bio-information storage integrated circuit board and the third insulating rubber seat.

[0064] In summary, the biological information capture and storage device 100 based on the embodiment of the present application is explained, which collects biological signals through a biological information receiving and processing device, and uses a signal processing and analysis technology based on deep learning to decompose and encode the biological signals, so as to automatically determine whether the biological signal time-frequency component is a noise component based on the matching characteristics between multiple decomposed signal time-frequency components, and inversely transform the remaining biological signal time-frequency components after elimination to obtain a denoised biological signal. In this way, the automated biological signal receiving and digital processing device can quickly collect and carefully analyze the time-frequency characteristics of the signal, so as to perform noise reduction processing on the original collected biological signal, providing a higher quality data signal for subsequent signal processing and analysis.

[0065] As described above, the biological information capture and storage device 100 according to the embodiment of the present application can be implemented in various terminal devices, such as a server for capturing and storing biological information. In one example, the biological information capture and storage device 100 according to the embodiment of the present application can be integrated into the terminal device as a software module and / or a hardware module. For example, the biological information capture and storage device 100 can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the biological information capture and storage device 100 can also be one of the many hardware modules of the terminal device.

[0066] Alternatively, in another example, the biometric information capturing and storing device 100 and the terminal device may be separate devices, and the biometric information capturing and storing device 100 may be connected to the terminal device via a wired and / or wireless network and transmit interactive information in accordance with an agreed data format.

[0067] The foregoing is merely an example of the principles of the present disclosure, and various modifications may be made by those skilled in the art without departing from the scope of the present disclosure. The above embodiments are presented for the purpose of illustration and not limitation. The present disclosure may also take many forms other than those explicitly described herein. Therefore, it is emphasized that the present disclosure is not limited to the methods, systems and devices explicitly disclosed, but is intended to include variations and modifications within the spirit and scope of the appended claims.

Claims

1. A biological information capture and storage device, comprising: Biological information receiving and processing device, biological information digitizing and processing device, and biological information storage device; Wherein, the biological information receiving and processing device is electrically connected to the biological information digitizing and processing device, and the biological information digitizing and processing device is electrically connected to the biological information storage device; The biological information digital processing device is characterized by comprising: A biological signal acquisition module, used to acquire the biological signal collected by the biological information receiving and processing device; A signal decomposition module, used for performing multi-level decomposition on the biological signal to obtain multiple biological signal time-frequency components; A biological signal encoding module, used for respectively encoding each biological signal time-frequency component of the plurality of biological signal time-frequency components to obtain a plurality of biological signal component time-frequency associated feature vectors; A one-way interactive matching module, used to use each biological signal component time-frequency association feature vector of the multiple biological signal component time-frequency association feature vectors as a query feature vector, and calculate its one-way interactive matching result relative to the feature library composed of the multiple biological signal component time-frequency association feature vectors to obtain multiple one-way matching feature vectors as one-way matching features; A recognition result generation module, used to obtain multiple recognition results based on the one-way matching feature, wherein the recognition results are used to indicate whether the time-frequency component of the biological signal is a noise component; A noise component removal module, used to remove the noise components in the multiple biological signal time-frequency components to obtain the remaining biological signal time-frequency components; A noise reduction signal generation module, used for performing signal restoration on the remaining time-frequency components of the biological signal to obtain a noise-reduced biological signal; Wherein, the one-way interactive matching module includes: A predetermined feature extraction unit, configured to extract a predetermined biological signal component time-frequency correlation feature vector from the plurality of biological signal component time-frequency correlation feature vectors as a query feature vector; A correlation calculation unit, used for calculating the correlation between the query feature vector and other biological signal component time-frequency correlation feature vectors other than the query feature vector to obtain the one-way matching feature vector corresponding to the query feature vector; Wherein, the correlation degree calculation unit includes: A logarithmic value calculation subunit, used to calculate the logarithmic function value with base 2 of each biological signal component time-frequency correlation feature vector in the other biological signal component time-frequency correlation feature vectors to obtain a sequence of biological signal component time-frequency correlation logarithmic function value vectors; An association interaction subunit, used for calculating the point-by-point multiplication between each biological signal component time-frequency association logarithmic function value vector in the sequence of biological signal component time-frequency association logarithmic function value vectors and the query feature vector to obtain a sequence of biological signal component time-frequency association interaction feature vectors; A global mean calculation subunit, used for calculating the global mean between the positional feature values ​​of each biological signal component time-frequency correlation interaction feature vector in the sequence of the biological signal component time-frequency correlation interaction feature vector to obtain the one-way matching feature vector; The correlation calculation unit is used to calculate the correlation between the query feature vector and other biological signal component time-frequency correlation feature vectors other than the query feature vector using the following interactive formula to obtain a one-way matching feature vector; wherein the interactive formula is: in, and are the query feature vector and the value of each position in the time-frequency association feature vector of each biological signal component, respectively, is the length of the query feature vector and the time-frequency correlation feature vector of each of the biosignal components, represents the eigenvalues ​​of each position in the one-way matching feature vector, Represents the value of the logarithmic function with base 2.

2. The biological information capture and storage device according to claim 1, characterized in that: The signal decomposition module is used to: use wavelet decomposition technology to perform multi-level decomposition on the biological signal to obtain the multiple biological signal time-frequency components.

3. The biological information capture and storage device according to claim 2, characterized in that: The biological signal encoding module is used to: input each biological signal time-frequency component of the multiple biological signal time-frequency components into a signal encoder based on the 1D-CNN model to obtain the time-frequency associated feature vectors of the multiple biological signal components.

4. The biological information capture and storage device according to claim 3, characterized in that: The recognition result generating module is used to: input the multiple one-way matching feature vectors into a classifier-based component identifier to obtain the multiple recognition results.

5. The biological information capture and storage device according to claim 4, characterized in that: The recognition result generating module is used for: Performing full connection encoding on the one-way matching feature vector using the fully connected layer of the classifier-based component identifier to obtain a one-way matching fully connected encoded feature vector; Inputting the one-way matching fully connected encoded feature vector into the Softmax classification function of the classifier-based component identifier to obtain the probability value of the one-way matching feature vector belonging to each classification label, wherein the classification label includes a label for indicating that the biological signal time-frequency component is a noise component and a label for indicating that the biological signal time-frequency component is not a noise component; The classification label corresponding to the largest probability value among the probability values ​​is determined as the recognition result.

6. The biological information capture and storage device according to claim 5, characterized in that: The noise reduction signal generating module is used to: input the remaining biological signal time-frequency components into a signal restorer based on inverse wavelet transform to obtain the noise-reduced biological signal.

Images