Methods and devices for assessing the corrective status of ankle and foot orthoses

Abstract

This invention proposes a method and device for judging the corrective state of an ankle-foot orthosis, relating to the field of orthotic technology. The method includes: training and validating a deep neural network using plantar pressure data and three-point force data in the sagittal, coronal, and horizontal planes corresponding to the ankle and foot, based on historical data from users who have worn ankle-foot orthoses, to obtain a three-point force prediction model for the ankle-foot orthosis; predicting target plantar pressure data in the anatomical region of the user's foot to obtain predicted three-point force data in the sagittal, coronal, and horizontal planes; collecting measured three-point force data of the user's bare foot in the sagittal, coronal, and horizontal planes; comparing the predicted three-point force data with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected corrective state. Thus, by establishing a mapping relationship between plantar pressure and the corrective three-point force, the ankle-foot orthosis fitting process is quantified, improving the accuracy and scientific rigor of orthotics.

Description

Technical Field

[0001] This invention relates to the field of orthotics, and more particularly to a method and apparatus for determining the corrective state of an ankle-foot orthosis. Background Technology

[0002] Ankle-foot orthoses are designed based on the three-point force correction principle of human biomechanics. They apply a three-point force system in different plane combinations to stabilize the ankle-foot joint, correct ankle-foot deformities, and restore and maintain normal biomechanical alignment. The clinical fit of ankle-foot orthoses is highly dependent on whether the applied three-point force achieves optimal correction. Quantitative assessment of the optimal three-point force remains a clinical challenge. It is usually determined by human experience, which is inefficient and inaccurate. Excessive three-point force may lead to excessive pressure on the soft tissues of the foot, causing local skin redness, pressure sores, or even nerve damage; insufficient three-point force cannot effectively counteract the deformity torque, resulting in insufficient correction or even dislodgement of the ankle-foot orthose. Summary of the Invention

[0003] The present invention aims to at least partially solve one of the technical problems in the related art.

[0004] Therefore, the first objective of this invention is to propose a method for judging the corrective state of an ankle-foot orthosis. By establishing a mapping relationship between plantar pressure and the three-point force of correction, the method quantifies the fit process of the ankle-foot orthosis, thereby improving the accuracy and scientific nature of the correction.

[0005] The second objective of this invention is to provide a device for determining the correction status of an ankle-foot orthosis.

[0006] The third objective of this invention is to provide an electronic device.

[0007] The fourth objective of this invention is to provide a non-transitory computer-readable storage medium storing computer instructions.

[0008] The fifth objective of this invention is to provide a computer program product.

[0009] To achieve the above objectives, a first aspect of the present invention provides a method for determining the corrective state of an ankle-foot orthosis, the method comprising:

[0010] Plantar pressure data is obtained by placing a first pressure sensor on the anatomical area of ​​the foot of a user who has historically worn ankle-foot orthoses and by measuring the user's weight. Three-point force data in the sagittal, coronal, and horizontal planes are collected by placing a second pressure sensor at three correction force points corresponding to the ankle and foot in the sagittal, coronal, and horizontal planes.

[0011] Using the plantar pressure data as input and the three-point force data in the sagittal, coronal, and horizontal planes as output, a deep neural network is constructed from the long short-term memory network for training and validation to obtain a three-point force prediction model for ankle-foot orthotics based on plantar pressure.

[0012] Collect target plantar pressure data of the anatomical region of the user's foot to be predicted, and input the target plantar pressure data into the three-point force prediction model of the ankle-foot orthosis to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes;

[0013] Collect force data at three points on the bare feet of the user to be predicted, corresponding to the sagittal, coronal, and horizontal planes.

[0014] By comparing the predicted three-point force data with the measured three-point force data, it can be determined whether the ankle-foot orthosis has achieved the expected corrective state.

[0015] To achieve the above objectives, a second aspect of the present invention provides a device for determining the correction status of an ankle-foot orthosis, the device comprising:

[0016] The first acquisition module is used to acquire plantar pressure data by using a first pressure sensor placed on the anatomical area of ​​the foot of a user who has historically worn an ankle-foot orthosis and the user's weight, and to acquire three-point force data in the sagittal, coronal and horizontal planes by using a second pressure sensor placed at three correction force points corresponding to the ankle and foot in the sagittal, coronal and horizontal planes.

[0017] The training module is used to take the plantar pressure data as input and the three-point force data of the sagittal, coronal and horizontal planes as output to train and verify the long short-term memory network to obtain the ankle-foot orthosis three-point force prediction model based on plantar pressure.

[0018] The prediction module is used to collect target plantar pressure data of the anatomical area of ​​the user's foot and input the target plantar pressure data into the three-point force prediction model of the ankle-foot orthosis to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes.

[0019] The second acquisition module is used to acquire three-point force data of the bare foot of the user to be predicted, corresponding to the sagittal, coronal and horizontal planes.

[0020] The judgment module is used to compare the magnitude of the predicted three-point force data with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected correction.

[0021] To achieve the above objectives, a third aspect of the present invention provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect.

[0022] To achieve the above objectives, a fourth aspect of the present invention provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the method described in the first aspect.

[0023] The present invention provides a method, device, electronic device, and storage medium for determining the correction state of an ankle-foot orthosis. This method utilizes plantar pressure data from historical users who have worn ankle-foot orthoses, along with three-point force data in the sagittal, coronal, and horizontal planes corresponding to the ankle and foot. A deep neural network based on a long short-term memory network is trained and validated to obtain a three-point force prediction model for the ankle-foot orthosis. Furthermore, the method predicts target plantar pressure data in the anatomical region of the user's foot to obtain predicted three-point force data in the sagittal, coronal, and horizontal planes. Measurements of the three-point force data in the bare foot of the user in the sagittal, coronal, and horizontal planes are collected. The predicted three-point force data is compared with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected correction state. Therefore, by establishing a mapping relationship between plantar pressure and the corrected three-point force, the ankle-foot orthosis fitting process is quantified, improving the accuracy and scientific rigor of the correction.

[0024] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0025] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0026] Figure 1 This is a flowchart illustrating a method for determining the corrective state of an ankle-foot orthosis according to an embodiment of the present invention.

[0027] Figure 2 This is a schematic diagram of a foot pressure point provided in an embodiment of the present invention;

[0028] Figure 3 This is a diagram of a long short-term memory neural network structure provided in an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of a device for determining the correction status of an ankle-foot orthosis, provided in an embodiment of the present invention. Detailed Implementation

[0030] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0031] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of this invention all comply with the relevant provisions of relevant laws and regulations.

[0032] The following describes, with reference to the accompanying drawings, a method and apparatus for determining the correction status of an ankle-foot orthosis according to an embodiment of the present invention.

[0033] Figure 1 This is a flowchart illustrating a method for determining the correction status of an ankle-foot orthosis provided in an embodiment of the present invention.

[0034] like Figure 1 As shown, the method for determining the corrective status of an ankle-foot orthosis includes the following steps:

[0035] Step 101: Obtain plantar pressure data by placing a first pressure sensor on the anatomical area of ​​the foot of a user who has historically worn an ankle-foot orthosis and the user's weight; and collect three-point force data in the sagittal, coronal, and horizontal planes by placing a second pressure sensor at three correction force points corresponding to the ankle and foot in the sagittal, coronal, and horizontal planes.

[0036] This invention provides a schematic diagram of a partial pressure point on the plantar surface, as shown in the embodiment of the invention. Figure 2 As shown, the plantar points include the first toe, third toe, first metatarsal bone, fifth metatarsal bone, anterior medial arch, anterior lateral arch, posterior medial heel, and posterior lateral heel. The selection of plantar points takes into account both the main anatomical areas of the foot and the areas where users with foot deformities are prone to lesions. Furthermore, experimental analysis shows that the pressure at these plantar points changed significantly after users wore ankle-foot orthoses, which can reflect the corrective efficacy of the ankle-foot orthoses.

[0037] Optionally, the first pressure sensor placed in the plantar anatomical region is positioned at the point where the ankle-foot orthosis corrects the function, and the three-point force system works in the sagittal, coronal, and horizontal planes.

[0038] In some possible implementations, plantar pressure data is obtained using a first pressure sensor placed on the anatomical area of ​​the foot of a user who has previously worn an ankle-foot orthosis, and the user's weight. This includes: acquiring the pressure value F of the first pressure sensor placed on the anatomical area of ​​the foot of a user who has previously worn an ankle-foot orthosis; acquiring the weight W of the user who has previously worn an ankle-foot orthosis, and using the quotient of the pressure value F divided by the weight W as the plantar pressure data F. 标 ,

[0039] For example, heavier users may have higher plantar pressure, but this does not necessarily reflect gait abnormalities or other problems. Therefore, it is necessary to standardize plantar pressure, and the standardized F... 标 Individual differences can be eliminated, and deep neural networks can make full use of datasets for training.

[0040] Step 102: Using plantar pressure data as input and three-point force data in the sagittal, coronal, and horizontal planes as output, a deep neural network is constructed from the long short-term memory network for training and validation to obtain a three-point force prediction model for ankle-foot orthotics based on plantar pressure.

[0041] In some possible implementations, Long Short-Term Memory (LSTM) networks The LSTM (Long Short-Term Memory) system consists of an input layer L0, a long short-term memory network layer L1, a fully connected layer F2, and an output layer, connected sequentially. The input layer L0 receives data sequence samples of plantar pressure data. Each data sequence sample contains plantar pressure features at a fixed time step, reflecting the plantar pressure distribution when the user is wearing an ankle-foot orthosis. The long short-term memory network layer L1 processes the long-term dependencies in the corresponding time series of the data sequence samples through a gating mechanism and outputs the hidden state of the last time step as a feature representation. The fully connected layer F2 uses the ReLU activation function to map the feature representation of the long short-term memory network layer to a set target output dimension, which is the predicted three-point force data in the sagittal, coronal, and horizontal planes. The output layer outputs the predicted three-point force data in the sagittal, coronal, and horizontal planes. The activation formula for the ReLU activation function is: f(X) = max(0,X); where X is the input quantity, which is the plantar pressure data. The activation function retains positive values ​​for the input quantity and sets negative values ​​to zero.

[0042] In the Long Short-Term Memory network, the output obtained by using the ReLU activation function as a nonlinear transformation is: output = max(0, W) T X+B), W is the weight, X is the input vector (plantar pressure data), B is the bias term, and T is the time step. The bias term is used to enhance the expressive power of the training.

[0043] Furthermore, the Long Short-Term Memory (LSTM) network layer adopts a bidirectional structure, is trained using the Adam optimizer, optimizes the LSM network using the Dropout mechanism, and dynamically enables an early stopping mechanism based on the loss curve during validation.

[0044] When training with the Adam optimizer, the learning rate can be set to 0.001, but it is not limited to this. A Dropout mechanism (with a dropout rate of 0.25) is used to prevent overfitting and optimize the Long Short-Term Memory network. Furthermore, the maximum number of training iterations can be set to 100, but it is not limited to this.

[0045] Understandably, the core idea of ​​a bidirectional LSTM network is that it considers not only the current state of the input data sequence sample but also its past states. Specifically, a bidirectional LSTM consists of a forward LSTM and a backward LSTM. The forward LSTM receives the input data sequence sample as is, processes each element sequentially starting from the first element, and outputs the forward representation of each element. The backward LSTM receives the input data sequence sample in reverse order, processes each element sequentially starting from the last element, and outputs the backward representation of each element. The outputs of these two LSTMs are concatenated, combining their respective hidden states to form a composite representation vector that integrates bidirectional temporal features. This joint feature vector can fully capture the contextual information of the data sequence sample.

[0046] Furthermore, five-fold cross-validation can be used to output the optimal ankle-foot orthosis three-point force prediction model. This is achieved by randomly dividing the original dataset (including multiple sets of plantar pressure and three-point force data) into five mutually exclusive subsets. One subset is selected sequentially as the test set, and the remaining four as the training set. A deep neural network is trained and evaluated on the test set, with this testing process repeated five times. For each test, accuracy, precision, recall, F1 score, and performance metrics (AUC) are used for comprehensive evaluation. The average of the five performance evaluation results is taken to obtain the final performance evaluation result of the deep neural network. This method reduces the impact of random sampling bias on the evaluation results through multiple data partitioning. The final deep neural network selection considers the average of the five validation results, reducing the randomness of the deep neural network performance evaluation and providing a more reliable performance estimate. Simultaneously, training and evaluating the deep neural network on multiple different training and test sets helps to better understand the generalization ability of the ankle-foot orthosis three-point force prediction model.

[0047] Step 103: Collect target plantar pressure data of the anatomical region of the user's foot to be predicted, and input the target plantar pressure data into the ankle-foot orthosis three-point force prediction model to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes.

[0048] In some possible implementations, the user selects a site on the sole of their foot that can reflect the corrective function of the ankle-foot orthosis as the plantar pressure acquisition site (the location of the first pressure sensor), and the three-point force system acts in combination, with three-point force systems in the sagittal, coronal, and horizontal planes.

[0049] Step 104: Collect three-point force data of the bare foot of the user to be predicted, corresponding to the sagittal, coronal, and horizontal planes.

[0050] In some possible implementations, a second pressure sensor can be placed at three correction force points on the user's bare foot corresponding to the sagittal, coronal, and horizontal planes to detect the three-point force and obtain the measured three-point force data in the sagittal, coronal, and horizontal planes.

[0051] Step 105: Compare the predicted three-point force data with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected correction state.

[0052] In some possible implementations, if the difference between the predicted three-point force data and the measured three-point force data is less than a set error threshold, it indicates that the predicted three-point force data represents the optimal three-point corrective force, and the ankle-foot orthosis has achieved the expected corrective state. Conversely, if the difference between the predicted three-point force data and the measured three-point force data is greater than or equal to the set error threshold, it indicates that the predicted three-point force data is not the optimal three-point corrective force and there is a deviation. The three-point corrective force of the ankle-foot orthosis is then adjusted based on the deviation until the optimal three-point corrective force is achieved, thereby promoting the rehabilitation of users with foot deformities.

[0053] The method for determining the correction status of an ankle-foot orthosis in this invention involves training and validating a deep neural network using plantar pressure data from historical users who have worn ankle-foot orthoses, as well as three-point force data in the sagittal, coronal, and horizontal planes corresponding to the ankle and foot. This results in a three-point force prediction model for the ankle-foot orthosis. The method then predicts target plantar pressure data in the anatomical region of the user's foot to obtain predicted three-point force data in the sagittal, coronal, and horizontal planes. Measurements of the three-point force data in the barefoot of the user in the sagittal, coronal, and horizontal planes are also collected. Finally, the predicted and measured three-point force data are compared to determine whether the ankle-foot orthosis has achieved the expected correction status. By establishing a mapping relationship between plantar pressure and the corrected three-point force, the method quantifies the ankle-foot orthosis fitting process, improving the accuracy and scientific rigor of the correction.

[0054] Figure 3 This is a diagram of a long short-term memory neural network structure provided in an embodiment of the present invention.

[0055] like Figure 3As shown, the Long Short-Term Memory (LSTM) network consists of a forget gate, an input gate, and an output gate, with the forget gate and input gate being the core of the LSTM unit. In the diagram, σ represents the gate activation function, typically the sigmoid function; the input and output activation functions are usually tanh functions. The forget gate's function is to select which information to retain and which to discard from the previous time step. Its calculation formula is as follows:

[0056] f t =σ(W f ·[h t-1 x t ]+b f ).

[0057] Among them, f t For the Gate of Oblivion, W f b represents the weight corresponding to the forget gate. f h is the bias corresponding to the forget gate. t-1 The output at time t-1, x t Let be the output at time t, and σ be the sigmoid activation function.

[0058] The function of the input gate is to selectively retain the current input data and the output data from the previous time point into the cell C. t In this context, the formula for calculating the input gate is:

[0059] i t =σ(W i ·[h t-1 x t ]+b i ).

[0060] Among them, W i and b i The input gate's weights and biases.

[0061] The function of the output gate is to control the output at the current time node using the current input data and the output data from the previous time point. Its calculation formula is as follows:

[0062] O t =σ(W o ·[h t-1 ,x t ]+b o ).

[0063] Among them, W o and b o The weights and biases of the output gate.

[0064] The formula for calculating the element state Ct is:

[0065] C t =ft ·C t-1 +i t ·tanh(W c ·[h t-1 ,x t ]+b c ).

[0066] Unit node output h i The calculation formula is:

[0067] h t =O t ·tanh(C t ).

[0068] During the dynamic process of human walking, plantar pressure exhibits typical periodic time-series characteristics (data sequence sample). Its variation is closely related to the phase of the gait cycle (fixed time step) and lower limb kinematic parameters. The plantar pressure distribution and the required corrective force (three electrical forces) at a given moment are not isolated but are linked to pressure signals over a period of time. This temporal information dependence requires the prediction of ankle-foot orthotic corrective force to have the ability to dynamically model data sequence samples. To address these characteristics, this invention introduces a Long Short-Term Memory (LSTM) network as the core prediction model. Through its unique gating mechanism (including input gate, forget gate, and output gate) and the synergistic effect of cell states, it can accurately capture the long-short-term dependencies in the data sequence samples.

[0069] To achieve the above embodiments, the present invention also proposes a device for judging the correction status of an ankle-foot orthosis.

[0070] Figure 4 This is a schematic diagram of a device for determining the correction status of an ankle-foot orthosis, provided in an embodiment of the present invention.

[0071] like Figure 4 As shown, the ankle-foot orthosis correction state judgment device 40 includes: a first acquisition module 41, a training module 42, a prediction module 43, a second acquisition module 44, and a judgment module 45.

[0072] The first acquisition module 41 is used to acquire plantar pressure data by using a first pressure sensor placed on the anatomical area of ​​the foot of a user who has historically worn an ankle-foot orthosis and the user's weight, and to acquire three-point force data in the sagittal, coronal and horizontal planes by using a second pressure sensor placed at three correction force points corresponding to the ankle and foot in the sagittal, coronal and horizontal planes.

[0073] Training module 42 is used to take the plantar pressure data as input and the three-point force data of sagittal, coronal and horizontal planes as output to train and verify the long short-term memory network to obtain a three-point force prediction model of ankle and foot orthosis based on plantar pressure.

[0074] Prediction module 43 is used to collect target plantar pressure data of the anatomical area of ​​the user's foot and input the target plantar pressure data into the three-point force prediction model of the ankle-foot orthosis to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes.

[0075] The second acquisition module 44 is used to acquire the force data of the bare feet of the user to be predicted at three points corresponding to the sagittal plane, coronal plane and horizontal plane.

[0076] The judgment module 45 is used to compare the magnitude of the predicted three-point force data with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected correction state.

[0077] Furthermore, in one possible implementation of this invention, the first pressure sensor placed in the plantar anatomical region is positioned at the site of the ankle-foot orthosis's corrective function.

[0078] Furthermore, in one possible implementation of this invention, the first acquisition module is specifically used for:

[0079] The pressure value F of the first pressure sensor was collected by placing a first pressure sensor on the anatomical area of ​​the sole of the foot of a user who has historically worn ankle-foot orthoses.

[0080] Obtain the weight W of users who have historically worn ankle-foot orthoses, and divide the pressure value F by the weight W to obtain the plantar pressure data F. 标 ,

[0081] Furthermore, in one possible implementation of this invention, the Long Short-Term Memory (LSTM) network consists of an input layer L0, a LTM network layer L1, a fully connected layer F2, and an output layer connected in sequence, wherein:

[0082] The input layer L0 is used to receive data sequence samples of plantar pressure data assembly. Each data sequence sample contains plantar pressure features at a fixed time step. The plantar pressure features are used to reflect the plantar pressure distribution of the user when wearing an ankle-foot orthosis.

[0083] The Long Short-Term Memory (LSI) network layer L1 processes the long-term dependencies in the time series corresponding to the data sequence samples through a gating mechanism and outputs the hidden state of the last time step as a feature representation.

[0084] The fully connected layer F2 uses the ReLU activation function to map the feature representation of the long short-term memory network layer to a set target output dimension, which is the predicted three-point force data in the sagittal, coronal and horizontal planes.

[0085] The output layer outputs the predicted three-point force data in the sagittal, coronal, and horizontal planes;

[0086] The activation formula for the ReLU activation function is: f(X) = max(0,X);

[0087] Where X is the input quantity, which is plantar pressure data. The activation function retains positive values ​​for the input quantity and sets negative values ​​to zero.

[0088] Furthermore, in one possible implementation of this invention, the long short-term memory network layer adopts a bidirectional structure, is trained using the Adam optimizer, optimizes the long short-term memory network using the Dropout mechanism, and dynamically enables an early stopping mechanism based on the loss curve during verification.

[0089] The ankle-foot orthosis correction status determination device of this invention uses plantar pressure data from historical users wearing ankle-foot orthoses and three-point force data in the sagittal, coronal, and horizontal planes corresponding to the ankle and foot to train and validate a deep neural network constructed from a long short-term memory network, resulting in an ankle-foot orthosis three-point force prediction model. It then predicts target plantar pressure data in the anatomical region of the user's foot to obtain predicted three-point force data in the sagittal, coronal, and horizontal planes. It also collects measured three-point force data of the user's bare foot in the sagittal, coronal, and horizontal planes. By comparing the predicted three-point force data with the measured three-point force data, it determines whether the ankle-foot orthosis has achieved the expected correction status. Therefore, by establishing a mapping relationship between plantar pressure and the corrected three-point force, the ankle-foot orthosis fitting process is quantified, improving the accuracy and scientific rigor of the correction.

[0090] To achieve the above embodiments, the present invention also proposes an electronic device, comprising:

[0091] At least one processor; and

[0092] A memory communicatively connected to the at least one processor; wherein,

[0093] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the aforementioned method.

[0094] To implement the above embodiments, the present invention also proposes a non-transitory computer-readable storage medium storing computer instructions for causing the computer to perform the aforementioned method.

[0095] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0096] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0097] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of the invention pertain.

[0098] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0099] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any of the following techniques known in the art, or a combination thereof: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0100] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0101] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0102] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for determining the corrective state of an ankle-foot orthosis, characterized in that, The method includes: Plantar pressure data is obtained by placing a first pressure sensor on the anatomical area of ​​the foot of a user who has historically worn ankle-foot orthoses and by measuring the user's weight. Three-point force data in the sagittal, coronal, and horizontal planes are collected by placing a second pressure sensor at three correction force points corresponding to the ankle and foot in the sagittal, coronal, and horizontal planes. Using the plantar pressure data as input and the three-point force data in the sagittal, coronal, and horizontal planes as output, a deep neural network is constructed from the long short-term memory network for training and validation to obtain a three-point force prediction model for ankle-foot orthotics based on plantar pressure. Collect target plantar pressure data of the anatomical region of the user's foot to be predicted, and input the target plantar pressure data into the three-point force prediction model of the ankle-foot orthosis to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes; Collect force data at three points on the bare feet of the user to be predicted, corresponding to the sagittal, coronal, and horizontal planes. By comparing the predicted three-point force data with the measured three-point force data, it can be determined whether the ankle-foot orthosis has achieved the expected corrective state.

2. The method according to claim 1, characterized in that, The first pressure sensor, placed in the anatomical region of the foot, is positioned at the corrective point of the ankle-foot orthosis.

3. The method according to claim 2, characterized in that, The plantar pressure data is obtained by using a first pressure sensor placed on the anatomical area of ​​the foot of a user who has historically worn ankle-foot orthoses and the user's weight, including: The pressure value F of the first pressure sensor was collected by placing a first pressure sensor on the anatomical area of ​​the sole of the foot of a user who has historically worn ankle-foot orthoses. Obtain the weight W of users who have historically worn ankle-foot orthoses, and divide the pressure value F by the weight W to obtain the plantar pressure data F. 标 , 4. The method according to claim 1, characterized in that, The Long Short-Term Memory (LSTM) network consists of an input layer L0, an LTM network layer L1, a fully connected layer F2, and an output layer connected in sequence, wherein: The input layer L0 is used to receive data sequence samples of plantar pressure data. Each data sequence sample contains plantar pressure features at a fixed time step. The plantar pressure features are used to reflect the plantar pressure distribution of the user when wearing an ankle-foot orthosis. The Long Short-Term Memory (LSI) network layer L1 processes the long-term dependencies in the time series corresponding to the data sequence samples through a gating mechanism and outputs the hidden state of the last time step as a feature representation. The fully connected layer F2 uses the ReLU activation function to map the feature representation of the long short-term memory network layer to a set target output dimension, which is the predicted three-point force data in the sagittal, coronal and horizontal planes. The output layer outputs the predicted three-point force data in the sagittal, coronal, and horizontal planes; The activation formula for the ReLU activation function is: f(X) = max(0,X); Where X is the input quantity, which is plantar pressure data. The activation function retains positive values ​​for the input quantity and sets negative values ​​to zero.

5. The method according to claim 4, characterized in that, The Long Short-Term Memory (LSTM) network layer adopts a bidirectional structure, is trained using the Adam optimizer, and is optimized using the Dropout mechanism. It also dynamically enables an early stopping mechanism based on the loss curve during validation.

6. A device for determining the corrective state of an ankle-foot orthosis, characterized in that, The device includes: The first acquisition module is used to acquire plantar pressure data by using a first pressure sensor placed on the anatomical area of ​​the foot of a user who has historically worn an ankle-foot orthosis and the user's weight, and to acquire three-point force data in the sagittal, coronal and horizontal planes by using a second pressure sensor placed at three correction force points corresponding to the ankle and foot in the sagittal, coronal and horizontal planes. The training module is used to take the plantar pressure data as input and the three-point force data of the sagittal, coronal and horizontal planes as output to train and verify the long short-term memory network to obtain the ankle-foot orthosis three-point force prediction model based on plantar pressure. The prediction module is used to collect target plantar pressure data of the anatomical area of ​​the user's foot and input the target plantar pressure data into the three-point force prediction model of the ankle-foot orthosis to obtain the predicted three-point force data in the sagittal, coronal and horizontal planes. The second acquisition module is used to acquire three-point force data of the bare foot of the user to be predicted, corresponding to the sagittal, coronal and horizontal planes. The judgment module is used to compare the magnitude of the predicted three-point force data with the measured three-point force data to determine whether the ankle-foot orthosis has achieved the expected correction state.

7. The apparatus according to claim 6, characterized in that, The first pressure sensor, placed in the anatomical region of the foot, is positioned at the corrective point of the ankle-foot orthosis.

8. The apparatus according to claim 7, characterized in that, The first acquisition module is specifically used for: The pressure value F of the first pressure sensor was collected by placing a first pressure sensor on the anatomical area of ​​the sole of the foot of a user who has historically worn ankle-foot orthoses. Obtain the weight W of users who have historically worn ankle-foot orthoses, and divide the pressure value F by the weight W to obtain the plantar pressure data F. 标 , 9. The apparatus according to claim 6, characterized in that, The Long Short-Term Memory (LSTM) network consists of an input layer L0, an LTM network layer L1, a fully connected layer F2, and an output layer connected in sequence, wherein: The input layer L0 is used to receive data sequence samples of plantar pressure data. Each data sequence sample contains plantar pressure features at a fixed time step. The plantar pressure features are used to reflect the plantar pressure distribution of the user when wearing an ankle-foot orthosis. The Long Short-Term Memory (LSI) network layer L1 processes the long-term dependencies in the time series corresponding to the data sequence samples through a gating mechanism and outputs the hidden state of the last time step as a feature representation. The fully connected layer F2 uses the ReLU activation function to map the feature representation of the long short-term memory network layer to a set target output dimension, which is the predicted three-point force data in the sagittal, coronal and horizontal planes. The output layer outputs the predicted three-point force data in the sagittal, coronal, and horizontal planes; The activation formula for the ReLU activation function is: f(X) = max(0,X); Where X is the input quantity, which is plantar pressure data. The activation function retains positive values ​​for the input quantity and sets negative values ​​to zero.

10. The apparatus according to claim 9, characterized in that, The Long Short-Term Memory (LSTM) network layer adopts a bidirectional structure, is trained using the Adam optimizer, and is optimized using the Dropout mechanism. It also dynamically enables an early stopping mechanism based on the loss curve during validation.

11. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-5.

12. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-5.

Images