Human body mechanics state evaluation system and method based on integrated double-sensing force-measuring table

Abstract

The invention discloses a human body mechanics state evaluation system and method based on an integrated dual-sensing dynamometer. The system comprises a multi-modal heterogeneous sensing fusion dynamometer platform, a data acquisition synchronization module and an upper computer, wherein the upper computer is internally provided with a data fusion processing module and a result generation module; the multi-mode heterogeneous sensing fusion force measuring platform is built by fixing pressure sensors on corresponding six-dimensional force measuring plates to form a stacked structure. The fusion processing module acquires fusion data by picking up six-dimensional force data and plantar pressure distribution data; and the result generation module is used for extracting the multi-modal biomechanical characteristics from the fused data and generating a human body biomechanical state evaluation report. According to the invention, through hardware integration and high-precision synchronous acquisition, plantar pressure and six-dimensional force data are fused, and accurate plantar shear force data are analyzed in combination with space-time registration, physical verification and reverse estimation, so that multi-dimensional features are extracted, and intelligent and quantitative evaluation of features such as a balance function, gait abnormality and a foot pathological state is realized.

Description

Technical Field

[0001] This invention relates to the field of biomechanical testing and medical rehabilitation engineering technology, specifically to a human biomechanical state assessment system and method based on an integrated dual-sensor force table. Background Technology

[0002] The interaction force between the sole of the foot and the ground is a key biomechanical signal for assessing the functional status of the human nervous and musculoskeletal systems. Foot examination is a routine method for assessing foot health, and existing foot examination techniques mainly include visual inspection, palpation, functional assessment, imaging examination, neurovascular assessment, and special examinations. Among these, the mainstream biomechanical testing techniques can be divided into two categories:

[0003] One type is plantar pressure distribution measurement, which uses an array of thin-film sensors to obtain a two-dimensional cloud map of plantar pressure distribution, visually displaying the pressure center trajectory, pressure in each area, contact area, etc. However, this technology can only measure the pressure distribution in the vertical direction and cannot measure the deeper mechanical causes (such as shear force and torque) that lead to this pressure distribution, thus limiting its application in analyzing balance strategies and joint loads.

[0004] Another type is the six-dimensional force plate test. The six-dimensional force plate can accurately measure forces (Fx, Fy, Fz) and moments (Mx, My, Mz) in three directions, making it the "gold standard" for dynamic analysis. It can calculate the center of pressure on the sole of the foot, the direction and magnitude of the resultant force vector. However, it essentially equates the complex force process of the entire sole to a single point of application (COP) and a resultant force vector, completely losing the details of the internal force distribution on the sole. It cannot accurately determine the contact state between the sole and the ground during walking. For example, in the case of Parkinson's patients, only the obvious abnormal behavior can be observed visually; the specific contact state between the arch of the foot and the ground is difficult to assess, and it is difficult to distinguish the specific contact between the sole and the force plate during walking. For instance, based solely on the data from the six-dimensional force plate, it is impossible to distinguish which area of ​​the foot generates the frictional force during walking. Furthermore, the data collected by the six-dimensional force plate is the sum of the three directions applied to it, and it cannot clearly distinguish the regional pressure on the foot.

[0005] Using these two testing devices independently has significant technical limitations. They can only obtain two independent and fragmented data streams, and a single sensor cannot fully reflect the complex biomechanical state of the foot-ground interaction, failing to achieve truly comprehensive information collection at the data and model levels. For example, it cannot answer questions such as "Why does the center of pressure shift in this way?" or "What kind of shear force is generated inside the foot when the arch collapses?" Simply using them side-by-side, however, cannot achieve a synergistic effect at the data and model levels. Therefore, it is necessary to construct a structure and create a data fusion method to realize a dual-sensor testing system. Summary of the Invention

[0006] To address the aforementioned problems, the present invention aims to propose a human biomechanical state assessment system and method based on an integrated dual-sensor force platform. Employing a sandwich-style stacked structure, the system integrates a six-dimensional force plate and a foot pressure sensor into a single integrated dual-sensor force platform. A synchronous triggering device is introduced to construct a multimodal heterogeneous sensing fusion force measurement platform, overcoming the limitations of traditional discrete equipment and fragmented data. This achieves hardware integration and synchronous acquisition of foot pressure and six-dimensional force data. A complete fusion processing chain, from spatiotemporal registration and physical verification to inverse estimation, is constructed. By establishing a constrained least-squares inverse mechanical problem model, high-precision estimation of local shear force distribution on the foot is achieved. Furthermore, multi-dimensional fusion features are extracted, and intelligent quantitative analysis and assessment of human balance, gait, and foot pathology are performed.

[0007] This was achieved through the following technical solutions:

[0008] First, a human biomechanical state assessment system based on an integrated dual-sensor force platform is proposed, including a multimodal heterogeneous sensor fusion force measurement platform, a data acquisition synchronization module, and a host computer. The host computer has a built-in data fusion processing module and a result generation module.

[0009] The multimodal heterogeneous sensing fusion force measurement platform is constructed by a matrix consisting of a single or n integrated dual-sensor force measurement stations. Any integrated dual-sensor force measurement station includes a six-dimensional force measurement plate for collecting six-dimensional force data and a pressure sensor for collecting foot pressure distribution data. Each pressure sensor is fixed on the corresponding six-dimensional force measurement plate to form a stacked structure.

[0010] The data acquisition and synchronization module includes a first acquisition unit connected to each six-dimensional force plate, a second acquisition unit connected to each pressure sensor, and a synchronization triggering device; wherein, the synchronization triggering device sends a trigger signal to each first acquisition unit and each second acquisition unit to control the synchronous start and stop of each six-dimensional force plate and each pressure sensor;

[0011] The data fusion processing module connects each first acquisition unit and each second acquisition unit, and is used to fuse the synchronously acquired six-dimensional force data and plantar pressure distribution data to obtain fused data; the result generation module is used to extract multimodal biomechanical features from the fused data and generate a human biomechanical state assessment report.

[0012] Preferably, the host computer is also equipped with a data flow management module. This module uses a double-ended queue as a data buffer to store the six-dimensional force data and plantar pressure distribution data of the entire multimodal heterogeneous sensor fusion force measurement platform. The data flow management module achieves efficient, low-latency data caching and reading / writing through the double-ended queue, effectively addressing the problem of inconsistent data rates among multimodal sensors.

[0013] Preferably, the data acquisition and synchronization module further includes a signal conditioner and an acquisition box. The signal conditioner is used to preprocess the corresponding six-dimensional force data and transmit it to the host computer; the acquisition box is used to power the corresponding pressure sensor, amplify and filter the signal, and send it to the host computer. The signal conditioner optimizes the six-dimensional force data to ensure a high signal-to-noise ratio and measurement accuracy, while the acquisition box provides dedicated power supply and filtering for the matrix array built into the pressure sensor, ensuring the integrity and stability of large-scale data.

[0014] Preferably, the data fusion processing module includes a spatiotemporal registration layer, a physical verification layer, and an inverse estimation layer. The spatiotemporal registration layer, based on an affine transformation model, maps the coordinates of each six-dimensional force data point and each plantar pressure distribution data point to a unified physical coordinate system for each six-dimensional force plate. Lagrange interpolation is used to resample data at different sampling frequencies to achieve temporal consistency. The physical verification layer verifies the spatial registration accuracy based on the vertical resultant force and pressure center trajectory of each six-dimensional force data point and each plantar pressure distribution data point. The inverse estimation layer constructs a least-squares optimization-based inverse mechanical problem model to estimate each plantar shear force distribution. Simultaneously, a Coulomb friction constraint optimization model is introduced into the inverse mechanical problem model to provide feedback on the reliability of each plantar shear force distribution. Spatiotemporal registration ensures that heterogeneous data are aligned under a unified spatiotemporal reference, resolving the issue of sampling frequency and coordinate system differences. Physical verification and inverse estimation jointly improve the physical consistency and reliability of the reconstructed shear force.

[0015] Preferably, the six-dimensional force data of any six-dimensional force measuring plate includes force F. x F y F z and torque M x M y M z The affine transformation model in the spatiotemporal registration layer is X. p =T×X i T is the transformation matrix, X pThe coordinates of the six-dimensional force measuring plate in a unified physical coordinate system, X i For the pixel coordinates of the corresponding plantar pressure distribution data, the Lagrange interpolation method is S'(t)=ΣS(t) i )·L i (t), where S represents the data at each different sampling frequency, and S' represents the resampled data. i L represents the i-th sampling time point t. i (t) is the Lagrange basis function; in the physical verification layer, through |ΣF zi - F zp | / F zp < ε and || COP i - COP p || < δ verifies spatial registration accuracy, ε is the vertical force relative error threshold, F zi F is the algebraic sum of the normal forces at each sensing point of the corresponding pressure sensor. zp COP is the sum of the normal forces on the six-dimensional force measuring plate. i COP p These represent the corresponding pressure center coordinates based on plantar pressure distribution data and six-dimensional force data, respectively, with δ being the pressure center coordinate difference threshold. In the inverse estimation layer, the mechanical inverse problem model adopts the least squares framework min‖Ax'-b‖², and the constrained optimization model is f. s =μ·P n A is the m×n force measurement matrix used in the six-dimensional force measuring plate, x' is the shear force distribution vector to be solved, b is the measured value vector, and the measured value vector is the force F. x F y and torque M x M y Total force, f s Let P be the local shear force vector, μ be the friction coefficient matrix, and P be the local shear force vector. n This represents the normal pressure distribution matrix. Affine transformation and Lagrange interpolation are used to achieve accurate coordinate mapping and temporal synchronization. Based on the error threshold verification mechanism of vertical resultant force and pressure center, abnormal registration frames can be automatically identified and eliminated, improving the reliability of fusion.

[0016] Preferably, the data fusion processing module performs fusion processing including computational mechanical coupling characteristics, local stability characteristics, functional load characteristics, and friction utilization rate. Each calculated characteristic and friction utilization rate is quantitatively evaluated using a feature-weighted fusion model to obtain a risk assessment score. Simultaneously, a fuzzy logic system is used to calculate the membership degree of each calculated characteristic relative to a corresponding set threshold. The change in membership degree within a unit time interval is used as the temporal characteristic change rate. The risk assessment score and the temporal characteristic change rate are output as mechanical coupling and stability indicators to the result generation module. This multi-dimensional biomechanical feature weighted fusion achieves a quantitative risk score for the human body's state. The introduction of fuzzy logic and temporal change rate enhances the system's sensitivity to and early warning capabilities regarding dynamic stability evolution trends.

[0017] Preferably, the multimodal biomechanical features extracted by the result generation module include plantar pressure parameters, six-dimensional mechanical parameters, mechanical coupling and stability indices, motion trajectory and balance-related parameters, and gait temporal parameters; wherein, the plantar pressure parameters include at least local and whole-foot pressure distribution, contact area and its dynamic rate of change, and plantar pressure center trajectory; the six-dimensional mechanical parameters include at least each six-dimensional force data and the frequency domain features of each six-dimensional force data; the motion trajectory and balance-related parameters include at least the pressure distribution trajectory morphology and fractal dimension; and the gait temporal parameters include at least gait stages, macroscopic gait parameters, and bipedal coordination features.

[0018] Preferably, the human biomechanical status assessment report includes foot function status assessment, balance and postural stability assessment, gait quality and athletic performance evaluation, disease screening and risk assessment, and rehabilitation progress and efficacy monitoring. The report covers core assessment dimensions in clinical and sports scenarios, meeting diverse needs such as rehabilitation, screening, and training, and enhancing the system's practical value and user decision support capabilities.

[0019] Secondly, a human biomechanical state assessment method based on an integrated dual-sensor force table is proposed. This method utilizes the aforementioned human biomechanical state assessment system and includes the following steps:

[0020] S1. Simultaneously collect the six-dimensional force data and plantar pressure distribution data of the subject's two feet on the multimodal heterogeneous sensor fusion force measurement platform as the initial signal;

[0021] S2. Perform fusion processing on the initial signal, including spatiotemporal registration, physical verification and inverse estimation in sequence, to obtain fused data;

[0022] S3. Extract multimodal biomechanical features from the fused data and generate a human biomechanical status assessment report; the human biomechanical status assessment report includes foot function status assessment, balance ability and postural stability assessment, gait quality and motor performance evaluation, disease screening and risk assessment, rehabilitation progress and efficacy monitoring.

[0023] Preferably, when generating a foot function status assessment, the calculation includes: overall pressure distribution cloud map, left and right foot load symmetry analysis, high-pressure area location and risk rating, and inference of arch morphology and function; when generating a balance ability and postural stability assessment, the calculation includes: static standing stability score, dynamic balance limit test data, sensory integration analysis, and risk alarm index; when generating a gait quality and motor performance evaluation, the calculation includes: gait cycle decomposition diagram, symmetry and coordination radar diagram, gait efficiency index, and abnormal pattern recognition results; when generating a disease screening and risk assessment, the calculation includes: foot ulcer risk, neuropathy warning indicators, and musculoskeletal abnormalities; and when generating rehabilitation progress and efficacy monitoring, the calculation includes: historical trend comparison chart, quantitative feedback on efficacy, and personalized training guidance suggestions.

[0024] The beneficial effects of this invention compared to the prior art are:

[0025] The technical solution of this invention adopts a sandwich-style layered structure, integrating a six-dimensional force measuring plate and a foot pressure sensor into a single unit and introducing a synchronous triggering device to construct a multimodal heterogeneous sensing fusion force measuring platform. This overcomes the limitations of traditional separate equipment and fragmented data, achieving hardware integration and synchronous acquisition of foot pressure and six-dimensional force data. A complete fusion processing chain from spatiotemporal registration and physical verification to inverse estimation is constructed. By establishing a constrained least-squares inverse mechanical problem model, high-precision estimation of the local shear force distribution of the foot is achieved. Furthermore, multi-dimensional fusion features are extracted, and intelligent quantitative analysis and evaluation of human balance, gait, and foot pathology are performed. Attached Figure Description

[0026] Figure 1 This is a block diagram of a human biomechanical state assessment system based on an integrated dual-sensor force table.

[0027] Figure 2 This is a flowchart of a method for assessing the biomechanical state of humans based on an integrated dual-sensor force table. Detailed Implementation

[0028] The following will be based on embodiments of the present invention. Figure 1 and Figure 2 The technical solutions in the embodiments of the present invention will be described in detail below.

[0029] like Figure 1The diagram shows a block diagram of a human biomechanical state assessment system based on an integrated dual-sensor force platform. The system includes a multimodal heterogeneous sensor fusion force measurement platform, a data acquisition and synchronization module, and a host computer. By integrating these hardware components and performing high-precision synchronous data acquisition, and then using the data fusion processing module and result generation module built into the host computer, the data is fused and multi-dimensional features are extracted, which can effectively and accurately assess the subject's balance function and gait abnormalities.

[0030] In the human biomechanical state assessment system, the main body of the multimodal heterogeneous sensor fusion force measurement platform is an integrated dual-sensor force measurement platform (hereinafter referred to as the force measurement platform) for subjects to perform balance standing or dynamic walking. This fusion force measurement platform can be built by a single force measurement platform or a matrix composed of multiple force measurement platforms. A single integrated dual-sensor force measurement platform consists of a six-dimensional force measurement plate for collecting six-dimensional force data and a pressure sensor for collecting foot pressure distribution data. Each pressure sensor is fixed on the corresponding six-dimensional force measurement plate to form a sandwich-like stacked structure.

[0031] The data acquisition and synchronization module connects two acquisition units of the integrated dual-sensor force platform: the first acquisition unit of the six-dimensional force plate, the second acquisition unit of the pressure sensor, and a synchronization triggering device. The data acquisition and synchronization module includes a single signal conditioner and an acquisition box. The signal conditioner connects to the corresponding six-dimensional force plate through the corresponding first acquisition unit, processing the corresponding six-dimensional force data (the six-dimensional force data of any six-dimensional force plate includes force F). x F y F z and torque M x M y M z The pressure sensor (where x and y are two orthogonal axes on a unified plane, and z is the vertical axis) undergoes preprocessing (noise reduction filtering, etc.). On the other hand, it can be connected to a host computer via network cable or Ethernet (UDP / TCP protocol). The pressure sensor internally contains a sensor array (potentially with tens of thousands of sensing points), therefore a data acquisition box is used to provide power and amplify and filter the large amount of data / signals transmitted by the pressure sensor before sending the entire frame of pressure image data to the host computer via USB or Ethernet. A signal conditioner optimizes the six-dimensional force data to ensure a high signal-to-noise ratio and measurement accuracy. The data acquisition box provides dedicated power and filtering for the pressure sensor's built-in matrix array, ensuring the integrity and stability of large-scale data.

[0032] To ensure synchronized acquisition of the two types of data, a synchronization triggering device (such as an NI-DAQ card, USB-6000 series, or CompactDAQ module) is used to control the synchronous activation or deactivation of the two acquisition units. Taking the NI-DAQ card as an example, a host computer can send a command to the NI-DAQ card in the test software to generate a unified TTL pulse signal (typically 0V-5V). This signal is then simultaneously transmitted via cable to the signal conditioner corresponding to the six-dimensional force plate and the "external trigger" input port of the acquisition box corresponding to the plantar pressure sensor, thus hard synchronizing the start and stop of the two acquisition devices. When both acquisition devices receive the rising edge of the TTL pulse at the same moment, they immediately begin synchronous data acquisition, achieving microsecond-level synchronization accuracy and ensuring the time alignment of the collected data.

[0033] For the testing software used in the host computer, the main control software can be developed using programming languages ​​(such as LabVIEW, C++, Python, etc.), and the acquisition parameters of the two heterogeneous sensor devices (six-dimensional force plate + pressure sensor) can be configured, such as sampling frequency and range. It supports sending instructions to the NI-DAQ card to generate trigger signals, and supports receiving and displaying data streams from the two heterogeneous sensor devices in real time. Finally, the data with a unified timestamp is stored on the hard disk.

[0034] In this embodiment, the host computer has a built-in data fusion processing module and a result generation module. The data fusion processing module is connected to the first and second acquisition units in the integrated dual-sensor force stage to achieve deep fusion of the acquired six-dimensional force data and plantar pressure distribution data. The result generation module is used to extract multimodal biomechanical features from the fused data and generate a human biomechanical state assessment report.

[0035] The data fusion processing module, as the core of the system, leverages a robust data storage system to achieve deep fusion of heterogeneous data from a six-dimensional force plate and a plantar pressure sensor. This module comprises a spatiotemporal registration layer, a physical verification layer, and an inverse estimation layer. Spatiotemporal registration ensures the alignment of heterogeneous data under a unified spatiotemporal reference, resolving differences in sampling frequency and coordinate system. Physical verification and inverse estimation jointly enhance the physical consistency and reliability of the reconstructed shear force.

[0036] The spatiotemporal registration layer, based on an affine transformation model, maps the coordinates of the six-dimensional force data and plantar pressure distribution data from a single integrated dual-sensor force stage to the physical coordinate system of the six-dimensional force plate. Lagrange interpolation is used to resample data at different sampling frequencies to achieve temporal consistency. The affine transformation model in the spatiotemporal registration layer is X. p =T×X i T is the transformation matrix (e.g., a 3×3 matrix containing affine transformation parameters such as rotation, translation, and scaling), X pThe coordinates of the six-dimensional force measuring plate in a unified physical coordinate system, X i For the corresponding plantar pressure distribution data, the Lagrange interpolation method is S'(t) = ΣS(t) i )·L i (t), where S represents the data at each different sampling frequency, and S' represents the resampled data. i L represents the i-th sampling time point t. i (t) is the Lagrange basis function.

[0037] The physical verification layer verifies spatial registration accuracy based on the vertical resultant force and pressure center trajectory of the six-dimensional force data from the integrated dual-sensor force platform and the plantar pressure distribution data. The physical verification layer uses the vertical force consistency criterion |ΣF zi -F zp | / F zp < ε verifies the accuracy of force measurement and uses the consistency criterion of center of pressure (COP) trajectory ||COP i - COP p || < δ verifies spatial registration accuracy, ε is the vertical force relative error threshold (can be taken as 0.05-0.1), F zi F is the algebraic sum of the normal forces at each sensing point of the corresponding pressure sensor. zp COP is the sum of the normal forces on the six-dimensional force measuring plate. i COP p These are the corresponding pressure center coordinates based on plantar pressure distribution data and six-dimensional force data, respectively, with δ being the pressure center coordinate difference threshold (which can be 2-5 mm).

[0038] The inverse estimation layer constructs a least-squares optimization-based inverse mechanical problem model to estimate the distribution of plantar shear force for each foot. Simultaneously, a Coulomb friction-constrained optimization model is introduced into the inverse mechanical problem model to provide feedback on the reliability of each plantar shear force distribution. In the inverse estimation layer, the inverse mechanical problem model employs the least-squares framework min‖Ax'-b‖², and the constrained optimization model is f. s = μ·P n A is the m×n force measurement matrix used in the six-dimensional force measuring plate, and x' is the shear force distribution vector to be solved (which can be decomposed into the shear force f at each point). xi and f yi (where i is a positive integer), b is the measurement value vector, and the measurement value vector is force F. x F y and torque M x M y Total force, f s Let P be the local shear force vector, μ be the friction coefficient matrix, and P be the local shear force vector. nThis represents the normal pressure distribution matrix. Affine transformation and Lagrange interpolation are used to achieve accurate coordinate mapping and temporal synchronization. Based on the error threshold verification mechanism of vertical resultant force and pressure center, abnormal registration frames can be automatically identified and eliminated, improving the reliability of fusion.

[0039] The data fusion processing module's operation mainly includes calculating mechanical coupling characteristics, local stability characteristics, functional load characteristics, and friction utilization rate. Each calculated characteristic and friction utilization rate is quantitatively evaluated using a feature-weighted fusion model to obtain a risk assessment score. Calculating mechanical coupling characteristics involves calculating the resultant force vector-pressure distribution coupling coefficient C. p = F t ·r p This reflects the correlation between the overall force and pressure distribution, where C p F is the coupling coefficient between the resultant force vector and the pressure distribution. t The total resultant force vector [F] x , F y , F z ], r p Let S be the pressure center location vector. Local stability characteristic S = P a / ‖f s The mechanical stability of various functional areas of the sole (which can be divided into multiple areas) can be quantified, where S is the local stability index and P... a For the average pressure in the local area, ||f s ‖ represents the modulus of local shear force. Functional load characteristics include the arch load torque M. a =‖r a ×F m ‖, where r a F is the position vector from the center of the arch to the center of pressure. m This is the normal force vector along the medial side of the arch of the foot. Friction utilization rate η = ||F|| s || / F z It can serve as an important parameter to assist in the assessment of foot function, among which F s It is the total shear force (F) x ,F y ), F z It refers to normal force. Based on actual clinical data, threshold data related to these characteristics can be statistically analyzed from the actual data of patients with flat feet, balance disorders, neuromuscular diseases, etc., and compared with the calculated characteristic results to generate corresponding diagnostic reports.

[0040] During the fusion process, based on each feature calculated above, a weighted fusion model R can be used. r =Σw i ·F i The quantitative assessment of pathological risk, Rr For pathological risk assessment score, w i F is the weight coefficient of the i-th feature, which can be extracted from actual clinical data. i Let be the i-th feature value. While obtaining the pathological risk assessment score, the severity level can also be calculated. Specifically, a fuzzy logic system is used to calculate the membership degree of each feature based on its relative distance to the corresponding set threshold: D = |FF| t | / σ, F is the eigenvalue, F t The clinical reference threshold for the feature is defined as σ, and the standard deviation of this feature in the healthy population is defined as σ. The change in membership degree over a unit time interval is used as the temporal feature change rate. The risk assessment score and the temporal feature change rate ΔF / Δt are used as mechanical coupling and stability indicators, and then output to the results generation module. Multi-dimensional biomechanical features are weighted and fused to achieve a quantitative risk score for the human body state. The introduction of fuzzy logic and temporal change rate enhances the system's sensitivity to and early warning capabilities regarding dynamic stability evolution trends, and can also provide objective evidence for rehabilitation effect evaluation.

[0041] In this embodiment, the multimodal biomechanical features extracted by the result generation module include plantar pressure parameters, six-dimensional mechanical parameters, mechanical coupling and stability indices, motion trajectory and balance-related parameters, and gait temporal parameters.

[0042] Plantar pressure parameters include at least the local and whole-foot pressure distribution (mean pressure, peak pressure, gradient), contact area and its dynamic rate of change, plantar pressure center trajectory and its related indicators (such as path length, 95% confidence ellipse area, movement speed, etc.).

[0043] Mean pressure P m The parameter P can be obtained from the pressure sensor on the sole of the foot. m =(Σ{i=1}^NP i ) / N, P i Let be the pressure value of the i-th sensing unit (sensing point), and N be the total number of effective sensing units. This reflects the average load intensity across the entire sole or a specific area. The peak pressure is a key indicator for assessing localized high pressure and ulcer risk; its value is the maximum pressure value on the sole or a selected area within the sole during the measurement time. Pressure gradient. In the discrete system of this scenario, the ratio of the pressure difference between adjacent sensing units to the spatial distance can be used for calculation.

[0044] Contact area A(t) = N(t) * A s N(t) represents the number of sensing units whose pressure value exceeds a preset threshold at time t, and A sThe dynamic rate of change dA / dt ≈ ΔA / Δt is the area of ​​a single sensing unit, which is the amount of change in the contact area per unit time. In gait feature analysis, this value reflects the smoothness of foot rolling and is the largest when the heel touches the ground and the toes leave the ground.

[0045] For the trajectory and related indicators (path length, 95% confidence ellipse area, movement speed) of the center of pressure (COP) in various regions of the sole, (x cop , y cop ) constitutes the COP trajectory, x cop (t) = (Σ_{i=1}^N (P_i(t) * x i )) / (Σ_{i=1}^NP i (t)), y cop (t) = (Σ_{i=1}^N (P i (t) * y i )) / (Σ_{i=1}^NP i (t)), (x i , y i () represents the spatial coordinates of the i-th sensing unit. A longer total length of the COP trajectory indicates more frequent attitude adjustments and potentially worse stability. The 95% confidence ellipse area is calculated by determining the standard deviation (σ) of the COP point in the X and Y directions. x , σ y ) and covariance, and then construct an ellipse containing 95% of the data points, with an area of ​​A. 95% =π*a*b, where a and b are the semi-major and semi-minor axes of the ellipse. But A 95% The smaller the value, the more precise the posture control and the better the balance. COP movement speed: v cop (t) = √[(dx cop / dt)² +(dy cop The mean or peak value of the instantaneous velocity of COP can reflect the sensitivity of neuromuscular control.

[0046] The six-dimensional mechanical parameters include at least the force data for each of the six dimensions (F). x F y F z M x M y M z ) and the frequency domain characteristics (dominant frequency, power spectral entropy, etc.) of each six-dimensional force data, F z Corresponding to the main supporting force, F x F y These correspond to shear forces in the forward / backward and inward / outward directions, respectively, and are related to propulsion, braking, and lateral stability; M x Corresponding inward and outward turning torque, My Corresponding flexion-extension moment, M z Corresponding to the rotational torque, these three torque parameters are important indicators for assessing a subject's fall tendency and balance stability. The dominant frequency of the force is the frequency component with the highest energy in the power spectrum after performing a Fourier transform on the force signal F(t); the power spectrum entropy H = -Σ (P k / ΣP j )*log(P k / ΣP j ), where P k It is the power of the k-th frequency range. When the entropy value is higher, the frequency components of the signal are more complex and random, which also indicates a change in control strategy or pathological state.

[0047] In this embodiment, the mechanical coupling and stability indicators include the coupling coefficient Cp between resultant force and pressure distribution, the local mechanical stability index S, the instantaneous friction utilization rate η, and the functional load. The coupling coefficient between resultant force and pressure distribution measures the degree of linear correlation between the global resultant force and the local pressure distribution, expressed as Cp = corr(F zt (t), P r (t) An examination reveals that a high coupling degree indicates efficient force transmission. The local mechanical stability index S is a comprehensive indicator. Example When pressure fluctuations (σ) p ) and gradient The larger the value, the lower the stability index S, where α is the weighting coefficient. η is the instantaneous friction utilization rate (the ratio of shear force to normal force), η(t) = √(Fx(t)² + Fy(t)²) / Fz(t). When this value approaches or exceeds 1, there is a risk of slipping. Functional load indicators, such as the arch load torque M... a It can be used to assess the support function and fatigue level of the foot arch, M a =F f *d, F f Let M be the total pressure in the forefoot region, and d be the distance from the center of pressure in the forefoot to the center of pressure in the heel. This is determined by examining M. a The size of the arch can be used to assess the stability of the foot arch under load.

[0048] The parameters related to motion trajectory and balance include at least the shape of the pressure distribution trajectory and the fractal dimension, which can be evaluated by assessing the swing area and the fractal dimension. The swing area is the sum of the instantaneous values ​​of the areas of the triangles formed by adjacent sampled points and the mean point of COP. The fractal dimension describes the complexity of the COP trajectory and its ability to fill space, and can be used to assist in feedback on age and neurological function status.

[0049] Gait temporal parameters include at least gait phases, macroscopic gait parameters, and bipedal coordination features. Gait phase identification focuses on time points such as stance, foot lift, and swing, and can be based on F...z (t) The curve is automatically identified by setting a threshold, i.e., observing F. z Curve, preliminary judgment of F z The first time the threshold is exceeded is when the heel touches the ground, F z When stable in a high position, it is a full-foot stance, F z When the toes leave the ground, the gait drops below the threshold. Macroscopic gait parameters include stride length, cadence, and speed. Stride length is the horizontal distance between two consecutive heel strikes on the same foot. Cadence is the number of steps per unit time. Speed ​​= stride length * cadence. Bipedal coordination characteristics mainly examine the balance ability of bipedal coordination, including the proportion of double support phase: (double support time / gait cycle time) * 100% and the left-right foot force-time integral symmetry ratio: (min(left, right) / max(left, right)) * 100%.

[0050] In this embodiment, the human biomechanical status assessment report includes foot function status assessment, balance and postural stability assessment, gait quality and athletic performance evaluation, disease screening and risk assessment, and rehabilitation progress and efficacy monitoring. The report covers core assessment dimensions in clinical and sports scenarios, meeting diverse needs such as rehabilitation, screening, and training, and enhancing the system's practical value and user decision support capabilities.

[0051] The assessment of foot function mainly includes calculating the overall pressure distribution cloud map (displaying the pressure of the whole foot in static standing or dynamic gait in the form of a heat map), analyzing the symmetry of load on the left and right feet (giving the percentage difference of key parameters such as peak pressure and impulse between the left and right feet), locating high-pressure areas and risk rating (combining medical knowledge base, such as the 2nd and 3rd metatarsal heads and the heel, marking areas where the pressure exceeds the clinical safety threshold and issuing red / yellow / green three-level warnings), and inferring the arch shape and function (based on the midfoot contact area and pressure, making a preliminary judgment of high arch, normal foot, low arch / flat foot, and analyzing its cushioning and propulsion functions in gait).

[0052] The assessment of balance ability and postural stability includes the calculation of: static standing stability score (based on the 95% confidence ellipse area and path length of the COP trajectory, providing a standardized score or percentile compared with age-matched norms), dynamic balance limit test data (recording the maximum deviation distance of COP and stability boundaries when the subject moves the center of gravity forward, backward, left, and right), sensory integration analysis (balance test under different sensory conditions, such as with eyes open / closed, assessing the proportion of reliance on vision, vestibular, and proprioception), and risk warning index (clearly indicating whether there is an increased risk of falling and pointing out the main problems, such as poor forward and backward control, excessive correction, etc.).

[0053] Evaluation of gait quality and athletic performance includes calculations of: gait cycle decomposition diagrams (showing the time percentage and force curves of each phase within a standard gait cycle), symmetry and coordination radar charts (displaying left-right symmetry from multiple dimensions such as stride length, stand-up time, and impulse), gait efficiency index (assessing gait efficiency by combining stride speed, cadence, and energy expenditure; energy expenditure can be reflected by the magnitude of COP fluctuations), abnormal pattern identification (drag gait—insufficient foot lift-off during the swing phase), and landing foot—insufficient control in the initial stand-up phase and rapid foot strike to the ground (analyzed by analyzing the dF at the initial heel strike). z / dt changes in magnitude to reflect the lack of propulsion (F before toes leave the ground) y The peak value of the propulsion phase is insufficient.

[0054] For disease screening and risk assessment, the main focus is on calculating the risk of foot ulcers, neuropathy warnings, and musculoskeletal abnormalities. Key indicators for foot ulcer risk include: assessing peak plantar pressure, pressure-time integral, and pressure gradient; outputting low, medium, and high risk levels based on international standards (such as the IWGDF guidelines) or machine learning models, and visualizing high-risk locations. Indicators for neuropathy warning include: plantar vibration perception threshold (if system integrated), and a significant increase in COP variability under closed-eye conditions (reflecting proprioceptive impairment). Musculoskeletal abnormalities mainly include: arthritis, manifested as abnormal pressure distribution in the corresponding joint area (such as the first metatarsophalangeal joint) and load avoidance behavior; Achilles tendinitis / plantar fasciitis, manifested as abnormal heel pressure patterns and changes in arch load torque.

[0055] For rehabilitation progress and efficacy monitoring, trend comparison charts, quantitative efficacy feedback, and personalized training guidance suggestions are generated. The trend comparison chart displays key parameters from this assessment (such as balance index, gait symmetry ratio, and peak pressure in the high-pressure area) alongside historical data in a single chart, clearly showing the changing trends. Quantitative efficacy feedback includes suggestions such as, "After 4 weeks of rehabilitation training, your static standing stability index improved by 15%, the peak pressure at the high-pressure point of your left heel decreased by 22%, and the proportion of double-support phase in your gait decreased by 5%, indicating significant improvements in balance and gait efficiency." Personalized training guidance suggestions are based on the test report and existing standard rehabilitation training guidelines, such as, "Given your asymmetry in propulsion between your left and right feet (left foot 20% weaker than right foot), it is recommended to strengthen the left gastrocnemius and soleus muscles and perform single-leg standing balance exercises." Based on actual data characteristics, corresponding suggestions can be given, such as, "Your friction utilization rate approaches the critical value of 0.8 when moving to the left rear; it is recommended to perform lateral movement and braking exercises in a safe environment to improve dynamic anti-slip ability," etc.

[0056] Furthermore, the data flow management module configured in the host computer uses a double-ended queue as a data buffer to store the six-dimensional force data and plantar pressure distribution data collected by the constructed force measurement platform. The data flow management module achieves efficient, low-latency data caching and reading / writing through the double-ended queue, effectively addressing the issue of inconsistent data rates from multimodal sensors.

[0057] like Figure 2 The diagram shows a flowchart of a human biomechanical state assessment method based on an integrated dual-sensor force table. This method uses the aforementioned human biomechanical state assessment system and includes the following steps:

[0058] S1. Simultaneously collect the six-dimensional force data and plantar pressure distribution data of the subject's two feet on the multimodal heterogeneous sensing fusion force measurement platform as the initial signal; S2. Perform fusion processing on the initial signal, including spatiotemporal registration, physical verification and inverse estimation in sequence, to obtain fused data;

[0059] S3. Extract multimodal biomechanical features from the fused data and generate a human biomechanical status assessment report. The human biomechanical status assessment report includes foot function status assessment, balance ability and postural stability assessment, gait quality and motor performance evaluation, disease screening and risk assessment, and rehabilitation progress and efficacy monitoring.

[0060] When generating foot function status assessments, the calculation includes overall pressure distribution cloud map, left and right foot load symmetry analysis, high-pressure area location and risk rating, and inference of arch morphology and function. When generating balance ability and postural stability assessments, the calculation includes static standing stability score, dynamic balance limit test data, sensory integration analysis, and risk alarm index. When generating gait quality and motor performance evaluations, the calculation includes gait cycle decomposition diagram, symmetry and coordination radar diagram, gait efficiency index, and abnormal pattern recognition results. When generating disease screening and risk assessments, the calculation includes foot ulcer risk, neuropathy warning indicators, and musculoskeletal abnormalities. When generating rehabilitation progress and efficacy monitoring, the calculation includes historical trend comparison chart, quantitative feedback on efficacy, and personalized training guidance suggestions.

[0061] It should be noted that the advantages and beneficial effects of the human biomechanical state assessment method are the same as those in the human biomechanical state assessment system mentioned above, and can be referred to the relevant descriptions in the human biomechanical state assessment system mentioned above, which will not be repeated here.

[0062] In summary, this invention employs a sandwich-style layered structure to integrate a six-dimensional force measuring plate with a foot pressure sensor and incorporates a synchronous triggering device, constructing a multimodal heterogeneous sensing fusion force measuring platform. This overcomes the limitations of traditional discrete equipment and fragmented data, achieving hardware integration and synchronous acquisition of foot pressure and six-dimensional force data. Furthermore, a complete fusion processing chain, from spatiotemporal registration and physical verification to inverse estimation, is constructed. By establishing a constrained least-squares inverse mechanical problem model, high-precision estimation of the local shear force distribution on the foot is achieved. Finally, multi-dimensional fusion features are further extracted from the fused data, and intelligent quantitative analysis and evaluation of human balance, gait, and foot pathology are performed, demonstrating significant advancements.

[0063] The above embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solutions based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A human biomechanical state assessment system based on an integrated dual-sensor force platform, characterized in that, It includes a multimodal heterogeneous sensing fusion force measurement platform, a data acquisition and synchronization module, and a host computer, which has a built-in data fusion processing module and a result generation module; The multimodal heterogeneous sensing fusion force measurement platform is constructed by a matrix consisting of a single or n integrated dual-sensor force measurement stations. Any integrated dual-sensor force measurement station includes a six-dimensional force measurement plate for collecting six-dimensional force data and a pressure sensor for collecting foot pressure distribution data. Each pressure sensor is fixed on the corresponding six-dimensional force measurement plate to form a stacked structure. The data acquisition and synchronization module includes a first acquisition unit connected to each six-dimensional force plate, a second acquisition unit connected to each pressure sensor, and a synchronization triggering device; wherein, the synchronization triggering device sends a trigger signal to each first acquisition unit and each second acquisition unit to control the synchronous start and stop of each six-dimensional force plate and each pressure sensor; The data fusion processing module connects each first acquisition unit and each second acquisition unit, and is used to fuse the synchronously acquired six-dimensional force data and plantar pressure distribution data to obtain fused data; the result generation module is used to extract multimodal biomechanical features from the fused data and generate a human biomechanical state assessment report.

2. The human biomechanical state assessment system based on an integrated dual-sensor force platform according to claim 1, characterized in that, The host computer is also equipped with a data flow management module. The data flow management module uses a double-ended queue as a data buffer to store the six-dimensional force data and plantar pressure distribution data of the entire multimodal heterogeneous sensing fusion force measurement platform.

3. The human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 1, characterized in that, The data acquisition and synchronization module also includes a signal conditioner and an acquisition box. The signal conditioner is used to preprocess the corresponding six-dimensional force data and transmit it to the host computer; the acquisition box is used to power the corresponding pressure sensor, amplify and filter the signal and send it to the host computer.

4. The human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 1, characterized in that, The data fusion processing module includes a spatiotemporal registration layer, a physical verification layer, and an inverse estimation layer; Among them, the spatiotemporal registration layer is based on an affine transformation model, which maps the coordinates of each six-dimensional force data and each plantar pressure distribution data to a unified physical coordinate system of each six-dimensional force measuring plate, and uses the Lagrange interpolation method to resample the data at each different sampling frequency to achieve temporal consistency. The physical verification layer verifies the spatial registration accuracy based on the vertical resultant force and pressure center trajectory of each six-dimensional force data and each plantar pressure distribution data. The inverse estimation layer constructs a mechanical inverse problem model based on least squares optimization to estimate the distribution of shear force on each foot. At the same time, a constrained optimization model of Coulomb friction is introduced into the mechanical inverse problem model to provide feedback on the reliability of the shear force distribution on each foot.

5. A human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 4, characterized in that, The six-dimensional force data of any one of the six-dimensional force plates includes forces F x , F y , F z and moments M x , M y , M z ; The affine transformation model in the spatiotemporal registration layer is X. p =T×X i T is the transformation matrix, X p The coordinates of the six-dimensional force measuring plate in a unified physical coordinate system, X i For the pixel coordinates of the corresponding plantar pressure distribution data, the Lagrange interpolation method is S'(t)=ΣS(t) i )·L i (t), where S represents the data at each different sampling frequency, and S' represents the resampled data. i L represents the i-th sampling time point t. i (t) is a Lagrange basis function; In the physical verification layer, through |ΣF zi - F zp | / F zp < ε and || COP i - COP p || < δ verifies spatial registration accuracy, ε is the vertical force relative error threshold, F zi F is the algebraic sum of the normal forces at each sensing point of the corresponding pressure sensor. zp COP is the sum of the normal forces on the six-dimensional force measuring plate. i COP p These are the corresponding pressure center coordinates based on plantar pressure distribution data and six-dimensional force data, respectively, with δ being the pressure center coordinate difference threshold. In the inverse estimation layer, the inverse mechanical problem model adopts the least squares framework min‖Ax'-b‖², and the constrained optimization model is f s =μ·P n A is the m×n force measurement matrix used in the six-dimensional force measuring plate, x' is the shear force distribution vector to be solved, b is the measured value vector, and the measured value vector is the force F. x F y and torque M x M y Total force, f s Let P be the local shear force vector, μ be the friction coefficient matrix, and P be the local shear force vector. n This is the normal pressure distribution matrix.

6. The human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 1, characterized in that, The data fusion processing module performs fusion processing including computational mechanical coupling characteristics, local stability characteristics, functional load characteristics, and friction utilization rate. Based on each calculated characteristic and friction utilization rate, a feature-weighted fusion model is used for quantitative evaluation to obtain a risk assessment score. At the same time, a fuzzy logic system is used to calculate the membership degree of each calculated characteristic relative to the corresponding set threshold. The change in membership degree within a unit time interval is used as the time-series characteristic change rate. The risk assessment score and the time-series characteristic change rate are used as mechanical coupling and stability indicators and output to the result generation module.

7. The human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 6, characterized in that, The results generation module extracted multimodal biomechanical features including plantar pressure parameters, six-dimensional mechanical parameters, mechanical coupling and stability indices, motion trajectory and balance-related parameters, and gait temporal parameters. Among them, the plantar pressure parameters include at least the local and whole foot pressure distribution, contact area and its dynamic rate of change, and the trajectory of the plantar pressure center; the six-dimensional mechanical parameters include at least each six-dimensional force data and the frequency domain characteristics of each six-dimensional force data; the motion trajectory and balance-related parameters include at least the pressure distribution trajectory morphology and fractal dimension; and the gait temporal parameters include at least the gait phase, macroscopic gait parameters, and bipedal coordination characteristics.

8. The human biomechanical state assessment system based on an integrated dual-sensor force table according to claim 1, characterized in that, The human biomechanical status assessment report includes foot function status assessment, balance ability and postural stability assessment, gait quality and motor performance evaluation, disease screening and risk assessment, and rehabilitation progress and efficacy monitoring.

9. A method for assessing human biomechanical status based on an integrated dual-sensor force table, wherein the method is operated using the human biomechanical status assessment system as described in any one of claims 1 to 8, characterized in that, Includes the following steps: S1. Simultaneously collect the six-dimensional force data and plantar pressure distribution data of the subject's two feet on the multimodal heterogeneous sensor fusion force measurement platform as the initial signal; S2. Perform fusion processing on the initial signal, including spatiotemporal registration, physical verification and inverse estimation in sequence, to obtain fused data; S3. Extract multimodal biomechanical features from the fused data and generate a human biomechanical status assessment report. The human biomechanical status assessment report includes foot function status assessment, balance ability and postural stability assessment, gait quality and motor performance evaluation, disease screening and risk assessment, and rehabilitation progress and efficacy monitoring.

10. A method for assessing the biomechanical state of an organism based on an integrated dual-sensor force table according to claim 9, characterized in that, When generating foot function status assessments, the calculation includes overall pressure distribution cloud map, left and right foot load symmetry analysis, high-pressure area location and risk rating, and inference of arch morphology and function; when generating balance ability and postural stability assessments, the calculation includes static standing stability score, dynamic balance limit test data, sensory integration analysis, and risk alarm index; when generating gait quality and motor performance evaluations, the calculation includes gait cycle decomposition diagram, symmetry and coordination radar diagram, gait efficiency index, and abnormal pattern recognition results; when generating disease screening and risk assessments, the calculation includes foot ulcer risk, neuropathy warning indicators, and musculoskeletal abnormalities. When generating rehabilitation progress and efficacy monitoring data, the process includes calculating historical trend comparison charts, quantitative feedback on efficacy, and personalized training guidance suggestions.

Images