Method, system, device and storage medium for improving the accuracy of electrical impedance imaging

Abstract

The present invention discloses a method, system, device and storage medium for improving the accuracy of electrical impedance imaging, wherein the method comprises: inputting current information, collecting voltage information, obtaining conductivity information according to the current information and the voltage information; constructing and obtaining a conductivity image according to the conductivity information; inputting the conductivity image into a preset neural network for image reconstruction, and outputting a reconstruction area and a confidence level; after inputting the current information n times in sequence at different times, obtaining n reconstruction areas and confidence levels corresponding to the reconstruction areas; obtaining the optimal reconstruction area according to the confidence level as the final electrical impedance imaging result. The present invention uses a neural network to reconstruct the conductivity image, obtains the output reconstruction area and the confidence level, and after collecting multiple information, obtains the optimal reconstruction area according to the confidence level as the final electrical impedance imaging result, thereby improving the accuracy of electrical impedance imaging. The present invention can be widely used in the field of electrical impedance imaging.

Description

Technical Field

[0001] The present invention relates to the field of electrical impedance imaging, and in particular to a method, system, device and storage medium for improving the accuracy of electrical impedance imaging. Background Art

[0002] Electrical impedance tomography (EIT) has a basic principle of applying a small safe driving current / voltage to the human body by various methods according to the different resistance / conductivity of different tissues in the human body under different physiological and pathological conditions. The image of the resistivity distribution or its changes inside the human body is reconstructed through the measured response information of the driving current or voltage in the human body.

[0003] In hospitals or professional institutions, patients are tested with professional equipment. Since professional equipment is troublesome to wear, each measurement takes a certain amount of time, and the patient experience is poor. With the development of sensor technology, there are flexible fabric sensors that can be attached to everyday clothes, and after adding flexible sensors, there will be no difference in wearing comfort compared to ordinary clothes. For example, for conductivity imaging of female breasts, flexible fabric sensors can be built into women's bras. Through the built-in sensors, the corresponding voltage data can be obtained, and the conductivity of the internal tissue of the breast can be reconstructed with this data. By comparing the conductivity of normal breast tissue, it can be detected whether there is tumor tissue. Although this wearable device has a good experience, wearable devices are different from medical devices. The data collected by them will contain various noises, such as noise caused by wearing posture, size of wearable devices, measurement errors, etc., which will affect the accuracy of conductivity reconstruction. Summary of the invention

[0004] In order to solve at least one of the technical problems existing in the prior art to a certain extent, the present invention aims at providing a method, system, device and storage medium for improving the accuracy of electrical impedance imaging.

[0005] The technical solution adopted by the present invention is:

[0006] A method for improving the accuracy of electrical impedance imaging, comprising the following steps:

[0007] Inputting current information, collecting voltage information, and obtaining conductivity information according to the current information and the voltage information;

[0008] Constructing and obtaining a conductivity image according to the conductivity information;

[0009] Input the conductivity image into the preset neural network for image reconstruction, and output the reconstructed area and confidence level;

[0010] After inputting n current information in sequence at different times, n reconstruction regions and confidences corresponding to the reconstruction regions are obtained;

[0011] The optimal reconstruction area is obtained according to the confidence level as the final electrical impedance imaging result.

[0012] Furthermore, the confidence is obtained by:

[0013] The softmax function (layer) in the neural network is obtained, and the maximum value of the output is used as the confidence.

[0014] Furthermore, the neural network is a classification model, and the classification result of the classification model is expressed as:

[0015] P, C = F (I, θ), ∑ p i =1 C∈(0,1)

[0016] Among them, p i represents the probability of classification; C represents the confidence of the neural network output, which is used to characterize the degree of confidence in the classification result; p i It is obtained through the softmax function of the last layer of the neural network.

[0017] Furthermore, the confidence is obtained by:

[0018] Setting a confidence path in the neural network and setting a loss function Lc according to the confidence path;

[0019] Combine the loss function Lc with the original loss Ls of the neural network to obtain the total loss function;

[0020] The total loss function is used to train the neural network, and after training, the value of the loss function Lc is obtained as the confidence level;

[0021] Among them, when the reconstructed area output by the neural network is correct, the value of the loss function Lc is close to 0; when the reconstructed area output by the neural network is wrong, the value of the loss function Lc is close to 1.

[0022] Furthermore, after adding the confidence path to the neural network, the expression of the neural network output is as follows:

[0023] p′=C*p+(1-C)*Y

[0024] Among them, Y is the target classification result; C represents the confidence of the neural network output; the original loss Ls is calculated using the cross entropy loss function;

[0025] The calculation formula of the loss function Lc is as follows:

[0026] Lc=sigmoid(-log(C))

[0027] When the confidence is close to 1, the value of the loss function Lc is close to 0; when the confidence is close to 0, the value of the loss function Lc is close to 1.

[0028] Furthermore, the step of obtaining the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level includes:

[0029] The reconstructed area corresponding to the maximum confidence level is obtained as the final electrical impedance imaging result.

[0030] Furthermore, the step of obtaining the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level includes:

[0031] Arrange the n confidence levels in descending order, and obtain the first m confidence levels and the reconstruction areas corresponding to the m confidence levels;

[0032] Obtain the intersecting area between the m reconstruction areas as the overlapping area;

[0033] The reconstructed region corresponding to the area of ​​the largest overlapping region is obtained as the final electrical impedance imaging result.

[0034] Another technical solution adopted by the present invention is:

[0035] A system for improving the accuracy of electrical impedance imaging, comprising:

[0036] An information acquisition module, used to input current information, collect voltage information, and obtain conductivity information according to the current information and the voltage information;

[0037] A composition module, used for constructing and obtaining a conductivity image according to the conductivity information;

[0038] An image reconstruction module is used to input the conductivity image into a preset neural network for image reconstruction, and output the reconstructed area and confidence level;

[0039] A repeated mapping module is used to obtain n reconstruction areas and confidences corresponding to the reconstruction areas after inputting current information n times in sequence at different times;

[0040] The image selection module is used to obtain the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level.

[0041] Another technical solution adopted by the present invention is:

[0042] A device for improving the accuracy of electrical impedance imaging, comprising:

[0043] at least one processor;

[0044] at least one memory for storing at least one program;

[0045] When the at least one program is executed by the at least one processor, the at least one processor implements the above method.

[0046] Another technical solution adopted by the present invention is:

[0047] A storage medium stores a program executable by a processor, wherein the program executable by the processor is used to execute the method as described above when executed by the processor.

[0048] The beneficial effects of the present invention are as follows: the present invention uses a neural network to reconstruct the conductivity image, obtains the output reconstruction area and the confidence, and after collecting multiple information, obtains the optimal reconstruction area according to the confidence as the final impedance imaging result, thereby improving the accuracy of impedance imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments of the present invention or the drawings of related technical solutions in the prior art are introduced below. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0050] Figure 1 is a schematic structural diagram of a wearable device in an embodiment of the present invention;

[0051] Figure 2 is a circuit diagram of a sensor in a wearable device according to an embodiment of the present invention;

[0052] Figure 3 is a flow chart of a method for improving the accuracy of electrical impedance imaging in an embodiment of the present invention;

[0053] Figure 4 is a schematic diagram of a conductivity image in an embodiment of the present invention;

[0054] Figure 5 It is a schematic diagram of the intersection of three reconstruction areas E'1 in an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0056] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0057] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0058] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0059] like Figure 1 As shown, Figure 1 Schematic diagram of a wearable device. In this embodiment, the wearable device is a bra, but the wearable device is not limited to a bra and can be a tights, a hat, etc. Based on existing sensor technology, flexible fabric sensors can be attached to everyday clothes, and after adding flexible sensors, there will be no difference in wearing comfort compared to ordinary clothes. At the same time, after washing and friction, the performance of the sensor will not decrease. Based on this technology, clothing that collects signals from specific parts of the human body can be designed.

[0060] like Figure 2 As shown, Figure 2 is a circuit diagram of a sensor in a wearable device. In this embodiment, the sensor circuit includes 32 electrodes, which are used to output current information and collect voltage information. The 32 electrodes are scattered on the wearable device. Figure 1Specifically, a switching circuit is connected to the electrode to control the output of current and the acquisition of voltage. The current output branch includes a current source circuit and a D / A conversion circuit; the voltage acquisition branch includes a filter circuit, a programmable amplifier circuit and an A / D conversion circuit. The D / A conversion circuit in the current output branch and the A / D conversion circuit in the voltage acquisition branch are both connected to the FPGA.

[0061] Based on the above wearable devices, such as Figure 3 As shown, this embodiment provides a method for improving the accuracy of electrical impedance imaging, comprising the following steps:

[0062] S1. Input current information, collect voltage information, and obtain conductivity information based on the current information and voltage information;

[0063] S2. constructing and obtaining a conductivity image according to the conductivity information;

[0064] S3, inputting the conductivity image into a preset neural network for image reconstruction, and outputting the reconstructed area and confidence level;

[0065] S4, after inputting the current information n times in sequence at different times, obtaining n reconstruction regions and the confidences corresponding to the reconstruction regions;

[0066] S5. Obtain the optimal reconstruction area according to the confidence level as the final electrical impedance imaging result.

[0067] See also Figure 1 The entire wearable device contains 32 (M) electrodes, divided into two layers, each layer has 16 electrodes, and each layer selects 2 electrodes as excitation electrodes. The working steps based on 2 excitation electrodes are as follows:

[0068] Step 1: Based on the output of excitation current by two excitation electrodes, the remaining electrodes simultaneously collect voltage values; by replacing two different electrodes as excitation electrodes in turn and acquiring corresponding voltage data, a set of voltage values ​​can be obtained, with a total of (M-1)*(M-2) voltage data.

[0069] Step 2: Preprocess the collected voltage data to remove the noise in the data. Specifically, SVD decomposition or PCA decomposition can be used to remove some random noise.

[0070] Step 3: Obtain the conductivity based on the current and denoised voltage data, and construct a conductivity image based on the conductivity. Figure 4As shown, region E1, region E2 and region E3 represent different organ tissues in the female breast, respectively, wherein region E4 represents background tissue, that is, it is not a critical detection site. The goal of electrical impedance imaging is to reconstruct the conductivity of its internal tissues, that is, the conductivity of region E1, region E2, region E3 and region E4, through the measured boundary voltage data, and to give the accurate boundaries of region E1, region E2, region E3 and region E4. It should be noted that based on different parts of the human body, the number of regions that may be obtained may be different, and one region or more than one region may be obtained.

[0071] In order to improve the accuracy of reconstructed conductivity, this embodiment adopts a two-step reconstruction method, and the specific steps are as follows:

[0072] Step a: constructing a conductivity image I1 according to the conductivity. This step can be implemented by using an existing algorithm, such as a linear algorithm or a nonlinear algorithm.

[0073] Step b: The conductivity image I1 is used as the input of the neural network. The target image of the neural network is the pre-set ground truth conductivity image. The output of the neural network reconstructs the conductivity and the corresponding confidence. The neural network can be of any suitable structure. Based on the characteristics of the voltage data, a convolutional neural network is more suitable. In this embodiment, the output of the neural network includes the area E1 and the confidence Area E2 and Confidence Region E3 and Confidence wait.

[0074] The following takes region E1 as an example to illustrate how the neural network works.

[0075] After the conductivity image I1 is input into the neural network, the output area E'1 and the confidence level corresponding to the area E'1 are Where E'1 represents the reconstructed area of ​​E1. It can be obtained in many ways, such as directly taking the maximum value of the neural network after the softmax function as the confidence (that is, the maximum value output by the softmax function (layer) as the confidence), or setting a confidence path in the neural network, setting a loss function Lc for it, and combining it with the original loss Ls of the neural network, that is, the total loss function is Lt = Lc + a*Ls. The design principle of Lc ensures that when the neural network outputs the E1 area correctly, Lc is close to 0, and when the neural network outputs the E1 area incorrectly, Lc approaches 1.

[0076] The neural network is trained in the following manner: First, the neural network is trained, and the training set is the preliminary conductivity image, and the label of the training set is the real conductivity image. Through training, the neural network learns to establish a mapping relationship between input and output. After the neural network training is completed and the corresponding model is generated, the conductivity image calculated in the new step is input into the model to obtain the final conductivity reconstruction image. It should be noted that the neural network can be either a classification model or a regression model, and both can achieve the purpose of conductivity reconstruction. In this embodiment, the classification model is used as an example to illustrate how to construct the loss function.

[0077] After adding the confidence score, the input is the preliminary conductivity image, and the output is the classification result and the confidence score, which can be expressed by the following formula:

[0078] P, C = F (I, θ), ∑ p i =1 C∈(0,1)

[0079] p i represents the probability of classification, and C represents the confidence score output by the neural network, which represents the degree of confidence in the classification result.

[0080] Obviously, p i It is obtained through the softmax of the last layer of the neural network. C can be obtained through a monotonic linear or nonlinear function with a value range of (0, 1). The commonly used one is the sigmoid function.

[0081] After adding the confidence branch, the new output p′ is obtained as follows:

[0082] p′=C*p+(1-C)*Y

[0083] Y is the target classification result.

[0084] Based on p′, the classification loss Ls is calculated using the cross entropy loss function. Lc is the confidence loss function, and its logarithmic form is directly used, that is, Lc = sigmoid (-log (C)). When the confidence is close to 1, the confidence loss is 0, and when the confidence is close to 0, the confidence loss is close to 1.

[0085] Therefore, the total loss function is: Lt = Lc + a*Ls. a is a hyperparameter used to adjust the proportional relationship between the two losses.

[0086] Through the above steps, a complete conductivity reconstruction can be completed, and the reconstructed area and the confidence corresponding to the reconstructed area can be obtained. Since the wearable device has a good experience and patients can wear it for a long time, it has an advantage in detection time. The wearable device can perform multiple measurements, and the number of times is represented by N. According to the above steps, the results of N reconstructed areas E'1 and the corresponding confidence can be obtained. The confidence is represented by C = {c1, c2, …, cn}. According to the confidence C, the best reconstruction area is selected from the N reconstructed areas E'1 as area E1.

[0087] Specifically, the optimal reconstruction area E'1 can be obtained by:

[0088] In some optional embodiments, the reconstructed region corresponding to the maximum confidence C is selected as the optimal reconstructed region E'1.

[0089] In some optional embodiments, see Figure 5 , sort the confidences in descending order, obtain the reconstruction areas corresponding to the first three highest confidences, detect the intersection areas (i.e., overlapping areas) between the three reconstruction areas, and if any reconstruction area contains the most overlapping areas, then this reconstruction area is taken as the optimal reconstruction area E'1. Figure 5 In , the confidences of the three reconstructed regions are 0.9, 0.92, and 0.95, respectively. The overlap area of ​​the reconstructed region with a confidence of 0.92 is the largest, and this reconstructed region is selected as the optimal reconstructed region E'1.

[0090] By the above method, the optimal reconstruction area E'2 corresponding to the area E2 and the optimal reconstruction area E'3 corresponding to the area E3 can be obtained in the same way. The final electrical impedance imaging is constructed by the obtained reconstruction areas.

[0091] In summary, compared with the prior art, this embodiment has the following beneficial effects:

[0092] (1) A confidence score is introduced into the neural network for reconstructing conductivity to identify the degree of certainty (i.e., confidence) of the neural network calculation results and improve the accuracy of conductivity reconstruction.

[0093] (2) With the convenience of wearable devices, multiple measurements can be performed and screened using confidence indicators to further improve the accuracy of the calculation results.

[0094] (3) Ultimately, the same effects as those achieved on complex professional equipment in hospitals can be achieved on wearable clothing, which is convenient for users and allows doctors to understand the patient's condition in a timely manner.

[0095] This embodiment also provides a system for improving the accuracy of electrical impedance imaging, including:

[0096] An information acquisition module, used to input current information, collect voltage information, and obtain conductivity information according to the current information and the voltage information;

[0097] A composition module, used for constructing and obtaining a conductivity image according to the conductivity information;

[0098] An image reconstruction module is used to input the conductivity image into a preset neural network for image reconstruction, and output the reconstructed area and confidence level;

[0099] A repeated mapping module is used to obtain n reconstruction areas and confidences corresponding to the reconstruction areas after inputting current information n times in sequence at different times;

[0100] The image selection module is used to obtain the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level.

[0101] A system for improving the accuracy of electrical impedance imaging in this embodiment can execute a method for improving the accuracy of electrical impedance imaging provided by the method embodiment of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0102] This embodiment also provides a device for improving the accuracy of electrical impedance imaging, including:

[0103] at least one processor;

[0104] at least one memory for storing at least one program;

[0105] When the at least one program is executed by the at least one processor, the at least one processor implements Figure 3 The method shown.

[0106] A device for improving the accuracy of electrical impedance imaging in this embodiment can execute a method for improving the accuracy of electrical impedance imaging provided by the method embodiment of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0107] The present application also discloses a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. A processor of a computer device can read the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes Figure 3 The method shown.

[0108] This embodiment also provides a storage medium storing instructions or programs that can execute a method for improving the accuracy of electrical impedance imaging provided by an embodiment of the method of the present invention. When the instructions or programs are run, any combination of implementation steps of the method embodiment can be executed, and the corresponding functions and beneficial effects of the method can be obtained.

[0109] In some selectable embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided by way of example, for the purpose of providing a more comprehensive understanding of technology. The disclosed method is not limited to the operation and logic flow presented herein. Selectable embodiments are expected, wherein the order of various operations is changed and the sub-operation of a part for which is described as a larger operation is performed independently.

[0110] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise specified, one or more of the functions and / or features described may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the present invention. More specifically, in view of the properties, functions, and internal relationships of the various functional modules in the device disclosed herein, the actual implementation of the module will be understood within the conventional skills of the engineer. Therefore, those skilled in the art can implement the present invention set forth in the claims without excessive experimentation using ordinary techniques. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0111] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0112] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0113] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0114] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0115] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0116] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0117] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for improving the accuracy of electrical impedance imaging, characterized in that: The following steps are involved: Inputting current information, collecting voltage information, and obtaining conductivity information according to the current information and the voltage information; Constructing and obtaining a conductivity image according to the conductivity information; Input the conductivity image into the preset neural network for image reconstruction, and output the reconstructed area and confidence level; After inputting n current information sequentially at different times, n reconstruction regions and confidences corresponding to the reconstruction regions are obtained; Obtaining the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level; The neural network is a classification model, and the classification result of the classification model is expressed as: P,C=F(I,θ),∑p i =1 C∈(0,1) Among them, p i Represents the probability of classification; C represents the confidence of the neural network output, which is used to characterize the degree of confidence in the classification result; p i It is obtained through the softmax function of the last layer of the neural network; The confidence level is obtained by: Setting a confidence path in the neural network and setting a loss function Lc according to the confidence path; Combine the loss function Lc with the original loss Ls of the neural network to obtain the total loss function; The total loss function is used to train the neural network, and after training, the value of the loss function Lc is obtained as the confidence level; Among them, when the reconstruction area output by the neural network is correct, the value of the loss function Lc is close to 0; when the reconstruction area output by the neural network is wrong, the value of the loss function Lc is close to 1; After adding the confidence path to the neural network, the expression of the neural network output is as follows: p′=C*p+(1-C)*Y Among them, Y is the target classification result; C represents the confidence of the neural network output; the original loss Ls is calculated using the cross entropy loss function; The calculation formula of the loss function Lc is as follows: Lc=sigmoid(-log(C)) When the confidence is close to 1, the value of the loss function Lc is close to 0; when the confidence is close to 0, the value of the loss function Lc is close to 1; The step of obtaining the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level includes: Arrange the n confidence levels in descending order, and obtain the first m confidence levels and the reconstruction areas corresponding to the m confidence levels; Obtain the intersecting area between the m reconstruction areas as the overlapping area; The reconstructed region corresponding to the area of ​​the largest overlapping region is obtained as the final electrical impedance imaging result.

2. A method for improving the accuracy of electrical impedance imaging according to claim 1, characterized in that: The confidence level is obtained by: Obtain the softmax function that passes through the neural network, and take the maximum value in the output as the confidence.

3. A system for improving the accuracy of electrical impedance imaging, characterized in that: include: An information acquisition module, used to input current information, collect voltage information, and obtain conductivity information according to the current information and the voltage information; A composition module, used for constructing and obtaining a conductivity image according to the conductivity information; An image reconstruction module is used to input the conductivity image into a preset neural network for image reconstruction, and output the reconstructed area and confidence level; A repeated mapping module is used to obtain n reconstruction areas and confidences corresponding to the reconstruction areas after inputting current information n times in sequence at different times; An image selection module, used for obtaining an optimal reconstruction area as a final electrical impedance imaging result according to the confidence level; The neural network is a classification model, and the classification result of the classification model is expressed as: P,C=F(I,θ),∑p i =1C∈(0,1) Among them, p i Represents the probability of classification; C represents the confidence of the neural network output, which is used to characterize the degree of confidence in the classification result; p i It is obtained through the softmax function of the last layer of the neural network; The confidence level is obtained by: Setting a confidence path in the neural network and setting a loss function Lc according to the confidence path; Combine the loss function Lc with the original loss Ls of the neural network to obtain the total loss function; The total loss function is used to train the neural network, and after training, the value of the loss function Lc is obtained as the confidence; when the reconstruction area output by the neural network is correct, the value of the loss function Lc is close to 0; when the reconstruction area output by the neural network is wrong, the value of the loss function Lc is close to 1; After adding the confidence path to the neural network, the expression of the neural network output is as follows: p′=C*p+(1-C)*Y Among them, Y is the target classification result; C represents the confidence of the neural network output; the original loss Ls is calculated using the cross entropy loss function; The calculation formula of the loss function Lc is as follows: Lc=sigmoid(-log(C)) When the confidence is close to 1, the value of the loss function Lc is close to 0; when the confidence is close to 0, the value of the loss function Lc is close to 1; The step of obtaining the optimal reconstruction area as the final electrical impedance imaging result according to the confidence level includes: Arrange the n confidence levels in descending order, and obtain the first m confidence levels and the reconstruction areas corresponding to the m confidence levels; Obtain the intersecting area between the m reconstruction areas as the overlapping area; The reconstructed region corresponding to the area of ​​the largest overlapping region is obtained as the final electrical impedance imaging result.

4. A device for improving the accuracy of electrical impedance imaging, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the method according to any one of claims 1 to 2.

5. A storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to perform the method according to any one of claims 1 to 2 when executed by the processor.

Images