Wearable plantar pressure and three-dimensional gait analysis system

Abstract

The invention discloses a wearable plantar pressure and three-dimensional gait analysis system, and relates to the technical field of gait analysis, the system comprises a wearable in-shoe plantar pressure analysis subsystem and a three-dimensional gait analysis subsystem; the wearable in-shoe plantar pressure analysis subsystem comprises a pressure sensor module, a pressure integration module, an intelligent transmission module, a detection software module, a power supply module, an instantaneous pressure and balance detection module, a standing balance detection module, a balance and gait detection module and a cloud server; the three-dimensional gait analysis subsystem comprises a hardware unit and a software unit; according to the invention, a wearable plantar pressure and three-dimensional gait analysis collaborative system is constructed, bidirectional verification compensation is realized by means of data fusion of two subsystems, and traditional single-dimensional limitation is broken; pressure sensing and a double-view reconstruction technology are fused, data are dynamically captured, and bidirectional error correction is carried out; the limitation of traditional equipment is solved, the evaluation accuracy and robustness are improved, and a quantitative basis is provided for clinical diagnosis, rehabilitation and the like.

Description

Technical Field

[0001] This invention relates to the field of gait analysis technology, specifically to a wearable plantar pressure and three-dimensional gait analysis system. Background Technology

[0002] Motor function assessment and gait analysis are core areas of research in medical rehabilitation, sports science, and biomechanics. With the increasing aging population and rising sports injury cases, the clinical need for precise quantification of human motor ability, balance function, and gait characteristics is becoming increasingly urgent. Gait abnormalities, as early key indicators of neurological diseases (such as stroke and Parkinson's disease), musculoskeletal diseases, and fall risk in the elderly, are crucial for accurate assessment in disease diagnosis, rehabilitation treatment, and efficacy evaluation. However, traditional assessment methods mainly rely on observation or single-dimensional sensors, making it difficult to comprehensively capture the coordinated changes in plantar pressure distribution and limb spatial trajectory during dynamic movement. In recent years, the rapid development of wearable sensor technology and 3D visual analysis has made multimodal motion data fusion possible. The industry urgently needs a system that can simultaneously analyze plantar pressure and 3D gait to solve problems such as data fragmentation and scenario limitations in traditional methods, and to promote the development of motor function assessment towards dynamic and quantitative approaches.

[0003] Existing technologies have significant limitations in many aspects. In terms of sensor performance, the accuracy and repeatability of pressure sensors are difficult to guarantee, and current wearable shoe pressure sensors lack fatigue resistance and are prone to breakage during gait push-off, seriously affecting the continuity and reliability of data acquisition.

[0004] In terms of scene adaptation and data acquisition, existing gait analysis devices are mostly limited to fixed paths and can only collect gait data from one side, lacking the ability to adapt to multiple movement modes; optical motion capture systems can only analyze large joints and cannot fully reflect the subtle movements of the feet, resulting in insufficient ability to detect fine joint range of motion; at the same time, it is difficult to achieve continuous natural walking monitoring in real-world environments, such as monitoring needs of a few minutes to more than ten minutes; moreover, it is difficult to collect pressure data and motion data synchronously, and it is difficult to achieve the spatiotemporal alignment of high-precision pressure measurement and postural kinematic data; lacking the ability to continuously monitor multiple steps of the whole body joints, it is impossible to support long-term natural walking with each step containing synchronous pressure and joint space data.

[0005] From a clinical and application perspective, existing solutions lack clinical practicality and reproducibility, failing to provide quantitative and reproducible indicators to support the establishment of disease diagnostic criteria. Furthermore, the high system cost, large site requirements, and complex operation limit their application in a wide range of fields, including rehabilitation assessment, diabetic foot plantar pressure risk assessment, orthopedic auxiliary diagnosis, pediatric gait development assessment, intervention efficacy evaluation, and sports science analysis. Therefore, there is an urgent need to develop a three-dimensional gait analysis system that can integrate multi-dimensional data, achieve continuous monitoring in a natural gait environment, and provide quantitative and reproducible results. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a wearable plantar pressure and three-dimensional gait analysis system. By integrating a pressure sensor module, three-dimensional visual reconstruction technology, and a cloud data analysis platform, it achieves synchronous and accurate acquisition of plantar pressure distribution and dynamic gait. The system integrates multimodal data spatiotemporal alignment algorithms and dual-view skeletal reconstruction technology, breaking through the limitations of traditional single-dimensional analysis. It supports balance function assessment in complex sports scenarios, and its cloud architecture supports real-time data storage, intelligent analysis, and visualization report generation. It provides a full-process quantitative solution for clinical rehabilitation, sports science, and scientific research, significantly improving the scientific nature and application value of sports function assessment.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a wearable plantar pressure and three-dimensional gait analysis system, the system comprising: Wearable shoe in-shoe plantar pressure analysis subsystem and three-dimensional gait analysis subsystem; The wearable in-shoe plantar pressure analysis subsystem includes a pressure sensor module, a pressure integration module, an intelligent transmission module, a detection software module, a power supply module, an instantaneous pressure and balance detection module, a standing balance detection module, a balance and gait detection module, and a cloud server. The three-dimensional gait analysis subsystem includes hardware units and software units; The wearable in-shoe plantar pressure analysis subsystem and the three-dimensional gait analysis subsystem use hardware synchronization and software algorithms to perform spatiotemporal alignment of plantar pressure data and three-dimensional gait data, multi-dimensional gait staging and event verification, and evaluation of whole-body posture stability and balance function. Furthermore, the two subsystems can perform bidirectional error correction when combined. The wearable in-shoe plantar pressure analysis subsystem and the three-dimensional gait analysis subsystem achieve collaboration with the software algorithm through the following hardware synchronization methods: hardware synchronization uses a USB-C cable to connect dual depth cameras, an RS232 interface to connect the treadmill, and Bluetooth 5.0 to transmit plantar pressure data; the software algorithm includes a dual-view spatiotemporal constraint fusion algorithm and a multimodal gait event collaborative detection algorithm to complete the spatiotemporal alignment of plantar pressure data and three-dimensional gait data, multi-dimensional gait staging and event verification, and evaluation of whole-body posture stability and balance function. Furthermore, the two are combined to achieve bidirectional error correction by constraining visual data with pressure data and correcting pressure data with visual pose. The data are linked through timestamps. Data on sudden pressure changes, pressure values, pressure change frequency, and human balance status are collected and time-stamped synchronously by the wearable insole foot pressure analysis subsystem. Image data, speed data, slope data, support force data, and joint node data are collected and time-stamped synchronously by the three-dimensional gait analysis subsystem. All data are linked one-to-one based on a unified timestamp.

[0008] Furthermore, in the wearable in-shoe plantar pressure analysis subsystem, the specific contents of each module are as follows: the pressure integration module includes a pressure testing insole and a testing shoe composed of nine-zone pressure sensors, with the pressure testing insole placed inside the testing shoe; the intelligent transmission module communicates with the detection software module via Bluetooth; the power module uses a rechargeable lithium battery pack and is equipped with a low-power power management unit, capable of automatically adjusting the power supply according to the system's operating status; the instantaneous pressure and balance detection module is used to capture data on sudden changes in plantar pressure during instantaneous human movement, and simultaneously analyzes the balance state of the human body under instantaneous movement by combining the symmetry and stability parameters of the pressure distribution. The system determines whether there is a risk of imbalance. The standing balance detection module continuously monitors the movement trajectory of the plantar pressure center, the uniformity of pressure distribution, and the frequency of pressure changes in each area while the human is standing, to assess the human's static balance ability. The balance and gait detection module simultaneously collects plantar pressure data and gait cycle-related information in dynamic gait scenarios. By analyzing the migration pattern of the pressure center with the gait cycle, the timing and duration of pressure peaks in each area, and combining gait rhythm stability parameters, it comprehensively assesses the human's dynamic balance ability and gait coordination. The cloud server includes a data communication module, a data integration module, a data calculation and analysis module, and a data storage module.

[0009] The detection modes of the detection software module include instantaneous mode, two-legged standing with eyes open mode, two-legged standing with eyes closed mode, one-legged standing with eyes open mode, one-legged standing with eyes closed mode, walking mode, treadmill mode, upstairs / downstairs mode, and uphill / downhill mode. The wearable insole plantar pressure analysis subsystem collects data such as pressure values ​​in nine regions, pressure change data, pressure center movement trajectory, pressure distribution uniformity, and pressure change frequency. All of these data carry timestamps consistent with those of the three-dimensional gait analysis subsystem and are synchronized to the three-dimensional gait analysis subsystem through the detection software module. This data serves as one of the input data for the dual-view spatiotemporal constraint fusion algorithm and the multimodal gait event collaborative detection algorithm. The data communication module employs an encrypted transmission protocol to establish a stable data connection with the detection software module, enabling bidirectional transmission of raw plantar pressure data, detection mode parameters, and device status information, ensuring data security and integrity during transmission. The data integration module receives multi-source data transmitted from the data communication module and, according to preset data classification rules, organizes and integrates plantar pressure data, subject basic information, and detection timestamps from different detection modes to form a structured data set. The data calculation and analysis module processes the structured data output by the data integration module, calculating characteristic parameters such as pressure peak value, pressure duration, and pressure center coordinates in each plantar region, generating basic analysis results reflecting the plantar pressure distribution characteristics and equilibrium state. The data storage module adopts a distributed storage architecture, allocating independent storage space for the integrated structured data and calculation and analysis results, establishing a data indexing mechanism to support rapid data query, retrieval, and long-term retention, while also providing data backup functionality to prevent data loss. The structured data integrated by the data integration module is sorted by timestamp and associated with the image data, speed data, slope data, and support force data collected by the three-dimensional gait analysis subsystem through an index, so that it can be called by the dual-view spatiotemporal constraint fusion algorithm and the multimodal gait event collaborative detection algorithm.

[0010] Furthermore, the specific contents of the three-dimensional gait analysis subsystem are as follows: the hardware unit includes a treadmill, two depth cameras, handrails, reflective markers, and a computer. The two depth cameras are located 1.5-2 meters in front of the treadmill at a height of 40-70cm, and 1.5-2 meters behind the treadmill at a height of 30-80cm, respectively. The software unit includes a data acquisition module, a dual-view calibration and synchronization module, a three-dimensional skeleton reconstruction module, a gait event detection module, a biomechanical parameter calculation module, and a visualization report generation module. The RGB images, depth images, infrared image data, treadmill speed and slope data, and handrail support force data collected by the three-dimensional gait analysis subsystem all carry timestamps. These timestamps are used to establish a one-to-one correspondence with the pressure-related data of the wearable shoe plantar pressure analysis subsystem, providing a complete set of input data for the dual-view spatiotemporal constraint fusion algorithm and the multimodal gait event collaborative detection algorithm.

[0011] Furthermore, in the three-dimensional gait analysis subsystem, the data acquisition module acquires RGB images, depth images, and infrared image data from two depth-sensing cameras; simultaneously, it acquires real-time speed and incline data of the treadmill via an RS232 interface, and acquires the subject's handrail support force data via a pressure sensor built into the handrail; the acquisition triggering methods include manual triggering and automatic triggering, with automatic triggering initiated after image recognition determines that the subject has entered the center area of ​​the treadmill platform; All data collected by the data acquisition module is stamped with high-precision timestamps. The timestamp accuracy is consistent with that of the wearable shoe plantar pressure analysis subsystem, ensuring that the pressure data is synchronized with the time sequence of image data and motion scene data.

[0012] Furthermore, the dual-view calibration and synchronization module in the 3D gait analysis subsystem employs the Zhang Zhengyou calibration method, using a 9×12 checkerboard grid with a side length of 20mm. It acquires 20-30 checkerboard images from different angles, calculates the intrinsic and extrinsic parameters of the two cameras, including focal length, principal point coordinates, and distortion coefficients, and extrinsic parameters including rotation matrix and translation vector. Data synchronization uses a combination of hardware triggering and software timestamps. Hardware-triggered synchronization is achieved by connecting the two cameras via a USB-C cable, with a synchronization error ≤1ms. Simultaneously, the software timestamps of each device are recorded for data alignment verification. The camera's internal and external parameters, as well as the synchronized image data, output by the dual-view calibration and synchronization module, along with timestamps and pressure data from the wearable shoe plantar pressure analysis subsystem, are directly input into the dual-view spatiotemporal constraint fusion algorithm.

[0013] Furthermore, the 3D skeleton reconstruction module in the 3D gait analysis subsystem: uses AI or infrared recognition technology to detect 2D human key points from RGB images acquired by two depth cameras. These 2D key points include corresponding points of key parts of the head, torso, upper limbs, and lower limbs. 3D coordinate calculation, combined with camera intrinsic and extrinsic parameters and depth image data obtained by the dual-view calibration and synchronization module, back-projects the 2D human key points to a unified world coordinate system to obtain the 3D coordinates of the key points. This is then fused using a dual-view spatiotemporal constraint fusion algorithm; the expression for the dual-view spatiotemporal constraint fusion algorithm is: ,in, The coordinates of joint i after fusion. It is a time variable. It represents the confidence weight of the original 3D coordinates of key point i captured by the camera before time t. These are the original 3D coordinates of key point i captured by the camera before time t. It is the confidence weight of the original 3D coordinates of key point i captured by the camera after time t. These are the original 3D coordinates of key point i captured by the camera after time t. It is a time continuity factor. These are the 3D coordinates of keypoint i at time t-1 after fusion. These are the 3D coordinates of keypoint i at time t-2 after fusion; The dual-view spatiotemporal constraint fusion algorithm calls the associated synchronized data, where the original 3D coordinates of the i-key points captured by the camera are combined with the plantar pressure distribution data of the corresponding timestamp to adjust the confidence weight. and The optimized 3D joint coordinates are directly calculated using algorithmic formulas, providing core data support for gait event detection and biomechanical parameter calculation.

[0014] Furthermore, the gait event detection module in the three-dimensional gait analysis subsystem: aligns the plantar pressure data transmitted by the wearable shoe plantar pressure analysis subsystem with the three-dimensional skeleton data collected by the three-dimensional gait analysis subsystem according to timestamps. It extracts the hindfoot pressure value, vertical velocity of the heel key point, vertical acceleration of the knee joint, and pressure center trajectory change features from the aligned data. A multimodal gait event collaborative detection algorithm is used, employing the following criteria: hindfoot pressure value reaching a preset threshold, vertical velocity of the heel key point changing from negative to positive, and vertical acceleration of the knee joint within a preset range. This comprehensively determines whether it is a heel-to-ground (HS) event. Combining the timing of the pressure center shifting to the hindfoot with the knee extension angle change features, a secondary verification is performed on the initially determined heel-to-ground (HS) event to correct the judgment deviation. Simultaneously, this module can also extend the toe-off (TO) event judgment based on the same algorithm logic, extracting the forefoot pressure value, vertical velocity of the toe key point, and ankle angle change features as the basis for TO event judgment. The expression of the multimodal gait event collaborative detection algorithm is: ,in, It is the result of the heel contact event at time t. =1, indicating that time t is determined to be a heel strike event; if =0 indicates that time t is determined to be a non-heel strike event. This is the hindfoot pressure value. This represents the greatest pressure on the hindfoot area in history. It is the pressure threshold coefficient. The vertical velocity of the heel. The vertical acceleration of the knee joint. It is the vertical acceleration threshold of the knee joint; The multimodal gait event collaborative detection algorithm directly calls the plantar pressure data, 3D joint coordinate data, and treadmill speed and slope data associated by timestamp. It performs threshold judgment and logical operation on the associated data through the judgment conditions in the algorithm formula, and outputs key gait events such as heel contact (HS) event and toe off-ground (TO) event to complete gait cycle segmentation.

[0015] Furthermore, the biomechanical parameter calculation module in the three-dimensional gait analysis subsystem divides biomechanical parameters into kinematic parameters and dynamic parameters. The kinematic parameter calculation targets the hip, knee, ankle joints, and foot, extracting angle change data in the sagittal, coronal, and horizontal planes respectively. It automatically calls the optimized 3D joint coordinates output by the three-dimensional skeleton reconstruction module and the plantar pressure distribution data transmitted by the wearable shoe plantar pressure analysis subsystem as the basic input for parameter calculation. It compares the parameter change trends at different times within the same gait cycle and eliminates abnormal parameter values ​​that exceed the physiological range of motion. The biomechanical parameter calculation module calls the optimized 3D joint coordinates output by the dual-view spatiotemporal constraint fusion algorithm, the gait cycle data segmented by the multimodal gait event collaborative detection algorithm, and the associated plantar pressure data, treadmill speed and slope data, and handrail support force data. Through the biomechanical analysis model, it calculates parameters such as joint angles and joint torques, forming the core parameters for assessing human gait.

[0016] Furthermore, the visualization report generation module in the three-dimensional gait analysis subsystem: receives kinematic parameters and dynamic parameters output by the biomechanical parameter calculation module and gait cycle segmentation data output by the gait event detection module, while simultaneously importing plantar pressure distribution data transmitted by the wearable in-shoe plantar pressure analysis subsystem; generates 3D model playback content, displaying human skeletal animation within the complete gait cycle, supporting viewpoint switching and playback speed adjustment; generates data charts, presenting pressure change curves, joint angle change curves, and joint torque change curves in various plantar regions; generates intervention comparison content, importing detection data from different time points of the same subject, enabling simultaneous display of data before, during, and after intervention, including parameter value comparison and trend presentation; supports custom report templates, with templates including basic subject information, detection items, detection time, various visualization contents, and parameter summaries, ultimately outputting a PDF report, which can include a signature area for the detection personnel, and supports direct printing or electronic archiving; The visualization report generation module directly receives the evaluation core parameters output by the biomechanical parameter calculation module, the 3D joint coordinate data output by the 3D skeleton reconstruction module, the gait event data output by the multimodal gait event collaborative detection algorithm, and the associated plantar pressure distribution curve data. It generates 3D skeletal animation, pressure change curve, joint angle change curve, and other content through preset visualization rendering logic, and integrates them into a visualization report according to the template.

[0017] Compared with existing technologies, this wearable plantar pressure and three-dimensional gait analysis system has the following advantages: I. This invention, through a highly integrated modular design, constructs a collaborative analysis system for wearable plantar pressure and three-dimensional gait analysis. By fusion of data from the two subsystems, it achieves a two-way data verification and compensation mechanism that is difficult to achieve with a single device, significantly improving the accuracy and reliability of the overall analysis. The system breaks through the limitations of traditional single-dimensional analysis, making the fusion analysis of plantar pressure and three-dimensional gait no longer a simple data superposition, but a synergistic relationship of mutual support and optimization, laying a more solid foundation for motor function assessment.

[0018] Second, this invention, through the deep integration of wearable in-shoe pressure sensors and dual-view 3D skeleton reconstruction technology, can dynamically capture changes in plantar pressure and joint spatial trajectories of the human body in different movement scenarios. The heterogeneous data collected by the two subsystems, after spatiotemporal alignment, can mutually verify and correct each other: when the 3D vision system experiences joint point localization problems due to limb occlusion, the high temporal resolution data from the plantar pressure subsystem can constrain and correct the visual data; and when insole displacement within the shoe causes a deviation in the pressure data baseline, the foot spatial pose of the 3D vision system can provide an absolute baseline for the pressure data, correcting the pressure center trajectory. This bidirectional error correction makes the data more accurate, effectively compensating for the shortcomings of single technologies.

[0019] This invention overcomes the inherent limitations of traditional single-dimensional analysis devices based on pure vision or pure pressure through the aforementioned bidirectional error correction and fusion mechanism. In complex real-world motion scenarios, the accuracy and robustness of balance function and gait assessment results far exceed those of single technologies. This provides unprecedented quantitative evidence for clinical gait disorder diagnosis, sports injury prevention, and rehabilitation effect evaluation. Furthermore, its multimodal data spatiotemporal alignment and dual-view fusion technology extends balance function assessment to dynamic scenarios, providing more scientific support for the development of personalized sports rehabilitation programs.

[0020] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0022] Figure 1 A schematic diagram illustrating the data interaction between the wearable shoe insole plantar pressure analysis subsystem and the three-dimensional gait analysis subsystem. Figure 2 This is a data interaction diagram of the plantar pressure and three-dimensional gait analysis subsystem in wearable shoes. Detailed Implementation

[0023] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below. Example 1

[0024] Scenario for rehabilitation assessment of children with flat feet.

[0025] For 5-year-old children with flat feet, a wearable plantar pressure and 3D gait analysis system is needed to comprehensively assess their static balance ability, dynamic gait coordination, and plantar pressure distribution characteristics. This provides data support for the development of rehabilitation training programs and the tracking of effects, and solves the problems of traditional assessments that rely on subjective experience, have limited data, and cannot dynamically monitor the situation.

[0026] Inspect the wearable shoe plantar pressure analysis subsystem to ensure that the pressure test insole, composed of nine pressure sensors, is intact, the test shoe size matches the child's foot, the rechargeable lithium battery pack is fully charged, and the low-power power management unit is working properly. This will ensure stable operation of the equipment during the testing process and prevent data loss or distortion due to equipment failure.

[0027] The hardware of the 3D gait analysis subsystem was debugged. Two depth cameras were installed 1.8 meters in front of the treadmill at a height of 50 cm and 1.8 meters behind the treadmill at a height of 60 cm. This installation position can comprehensively capture the child's full-body posture during gait. The two cameras were connected via USB-C cable, and the treadmill was connected to the computer via RS232 interface. It was ensured that the pressure sensor built into the handrail responded normally. Reflective markers were affixed to key parts of the child's head, torso, upper limbs, and lower limbs to provide clear feature point references for subsequent 3D skeleton reconstruction.

[0028] Start the cloud server and confirm that the data communication module establishes a stable connection with the detection software module through an encrypted transmission protocol to ensure that children's privacy data is not leaked during transmission. The distributed storage architecture and backup function of the data storage module are normal, ensuring that the detection data can be retained for a long time and will not be lost. The data integration module can integrate data from multiple sources according to preset rules to provide a structured data foundation for subsequent analysis.

[0029] Create a child-specific testing profile in the testing software module, enter basic information such as name, age, gender, height, and weight, and bind a unique identification code to facilitate the association and matching of data from each subsequent test, enabling long-term dynamic follow-up tracking.

[0030] Have the child wear the test shoes, ensuring the pressure test insoles fit perfectly against the soles of their feet to guarantee the accuracy of the foot pressure data collection. Activate the wearable in-shoe foot pressure analysis subsystem and select the detection mode as bipedal standing with eyes open, bipedal standing with eyes closed, single-leg standing with eyes open, single-leg standing with eyes closed, and walking mode. The multi-mode detection can comprehensively cover static and dynamic scenarios, obtaining rich balance and gait data.

[0031] The 3D gait analysis subsystem is activated, and the data acquisition module is set to automatic trigger mode. Data acquisition is initiated after the child enters the center area of ​​the treadmill platform through image recognition, avoiding time deviations caused by human triggering. Simultaneously, RGB images, depth images, and infrared image data from two depth cameras, real-time speed data of the treadmill, and handrail support force data are acquired. All data are bound to high-precision timestamps, and the timestamp accuracy is consistent with that of the plantar pressure data, ensuring that data from different sources can be accurately correlated.

[0032] In the bipedal standing mode with eyes open / closed, the standing balance detection module of the wearable in-shoe plantar pressure analysis subsystem continuously monitors the movement trajectory of the plantar pressure center, the uniformity of pressure distribution, and the frequency of pressure changes in each area. The instantaneous pressure and balance detection module captures data on sudden changes in plantar pressure that may occur during standing, and analyzes the balance status by combining pressure distribution symmetry and stability parameters to accurately determine the child's static balance ability. The three-dimensional gait analysis subsystem simultaneously collects images of the child's full-body posture and handrail support force when standing. The dual-view calibration and synchronization module completes camera parameter calibration and data synchronization with a synchronization error of ≤1ms, ensuring the temporal consistency of image data and pressure data.

[0033] In the single-leg standing mode with eyes open / closed, the above data collection process was repeated, focusing on recording the peak pressure, pressure duration, and body posture stability data of each area of ​​the sole when supporting on one foot. This further refined the dimensions of static balance ability assessment and provided a more accurate basis for judging the impact of flat feet on balance function.

[0034] In walking mode, children walk naturally on the treadmill. The balance and gait detection module of the wearable in-shoe plantar pressure analysis subsystem simultaneously collects pressure values ​​in nine plantar regions, the migration pattern of the pressure center with the gait cycle, and the timing and duration of pressure peaks in each region. Combined with gait rhythm stability parameters, it assesses dynamic balance ability and gait coordination, clearly revealing the abnormal pressure distribution in the gait of children with flat feet. The 3D skeleton reconstruction module of the 3D gait analysis subsystem uses AI technology to detect 2D human key points, and back-projects the camera's intrinsic and extrinsic parameters and depth image data to a unified world coordinate system. It then fuses the dual-view spatiotemporal constraint fusion algorithm to optimize 3D joint coordinates and improve the accuracy of joint motion trajectory capture. The expression of the dual-view spatiotemporal constraint fusion algorithm is as follows: ,in, The coordinates of joint i after fusion. It is a time variable. It represents the confidence weight of the original 3D coordinates of key point i captured by the camera before time t. These are the original 3D coordinates of key point i captured by the camera before time t. It is the confidence weight of the original 3D coordinates of key point i captured by the camera after time t. These are the original 3D coordinates of key point i captured by the camera after time t. It is a time continuity factor. These are the 3D coordinates of keypoint i at time t-1 after fusion. It represents the 3D coordinates of keypoint i at time t-2 after fusion.

[0035] All collected data is quickly synchronized to the detection software module via the intelligent transmission module (Bluetooth 5.0), and then transmitted to the cloud server. Bluetooth 5.0 transmission ensures the efficiency and stability of data transmission. Pressure data, carrying timestamps, is linked with 3D gait data through an index. The data integration module sorts and integrates the data into a structured data set according to timestamps, which facilitates subsequent algorithm calls and analysis.

[0036] The cloud server's data calculation and analysis module processes the structured data, calculates characteristic parameters such as pressure peak value, pressure duration, and pressure center coordinates in various areas of the sole, and generates basic analysis results on the distribution characteristics and equilibrium state of the sole pressure, providing quantitative data support for subsequent evaluation.

[0037] The gait event detection module of the 3D gait analysis subsystem calls the plantar pressure data and 3D joint coordinate data correlated by timestamps, and adopts a multimodal gait event collaborative detection algorithm. This algorithm combines features such as hindfoot pressure values, vertical velocity at the heel key point, and vertical acceleration of the knee joint. The expression for the multimodal gait event collaborative detection algorithm is as follows: ,in, It is the result of the heel contact event at time t. =1, indicating that time t is determined to be a heel strike event; if =0 indicates that time t is determined to be a non-heel strike event. This is the hindfoot pressure value. This represents the greatest pressure on the hindfoot area in history. It is the pressure threshold coefficient. The vertical velocity of the heel. The vertical acceleration of the knee joint. It is the vertical acceleration threshold of the knee joint, which accurately determines the heel strike (HS) event and toe lift (TO) event, completes gait cycle segmentation, and provides a clear basis for analyzing gait rhythm abnormalities.

[0038] The biomechanical parameter calculation module calls the optimized 3D joint coordinates, segmented gait cycle data, and associated plantar pressure data and treadmill speed data. Through the biomechanical analysis model, it calculates parameters such as the angle changes and joint torques of the hip, knee, and ankle joints in the sagittal, coronal, and horizontal planes, comprehensively presenting the abnormal characteristics of joint movement in children with flat feet.

[0039] By combining the big data reference range module of the three-dimensional foot and ankle data processing software, the plantar pressure parameters and gait biomechanical parameters of children are compared with the P5, P25, P50, P75 and P95 percentile reference ranges of the same age and sex. This clarifies the position of their flat foot-related indicators (such as arch index and plantar pressure distribution uniformity) in the same age group, scientifically assesses the degree of developmental deviation, and provides targeted directions for subsequent rehabilitation interventions.

[0040] The visualization report generation module generates PDF reports, which include 3D skeletal animation playback, pressure change curves in various areas of the sole, joint angle change curves, gait cycle segmentation results, and comparison charts with big data reference ranges, intuitively presenting various abnormal data related to children's flat feet.

[0041] Based on the report, the rehabilitation physician assesses the specific manifestations of the child's static balance deficiency, the areas of uneven foot pressure distribution in dynamic gait, and the abnormalities in joint movement, and develops a targeted rehabilitation training plan to ensure that the rehabilitation plan is tailored to the child's actual situation.

[0042] Follow-up tests are conducted every 3 months. The individual development follow-up module generates follow-up curves for each characteristic indicator, compares the changes in plantar pressure distribution, balance ability and gait parameters before and after intervention, quantifies the rehabilitation effect, dynamically adjusts the rehabilitation plan, and ensures the effectiveness and accuracy of rehabilitation training.

[0043] In summary, this embodiment utilizes a wearable plantar pressure and 3D gait analysis system to complete the entire process of equipment debugging, multi-mode data acquisition, precise data processing, and result application according to specifications. Figure 1 As shown, by leveraging core technologies such as dual-view spatiotemporal constraint fusion algorithms and multimodal gait event collaborative detection algorithms, combined with the big data reference and follow-up modules of 3D foot and ankle data processing software, the static balance, dynamic gait, and plantar pressure characteristics of children with flat feet are comprehensively quantified. This not only solves the problems of subjective assessment and limited data in traditional methods, but also enables dynamic monitoring of rehabilitation effects, providing a scientific basis for the formulation and adjustment of personalized rehabilitation plans, ultimately helping children improve foot function and achieve the expected rehabilitation goals. Example 2

[0044] Fall risk screening scenarios for the elderly.

[0045] For individuals aged 70 and older, a wearable plantar pressure and 3D gait analysis system is used to comprehensively assess their standing balance, gait stability, and plantar pressure response, screening for fall risk factors. This addresses the limitation of traditional screening methods in quantifying the synergistic relationship between overall postural stability and plantar pressure, providing a scientific basis for fall prevention interventions. Specific steps are as follows: Figure 2 As shown.

[0046] The wearable shoe's plantar pressure analysis subsystem was inspected to ensure that the nine-zone pressure sensors in the pressure test insole were functioning correctly and could accurately capture subtle changes in plantar pressure in the elderly. The test shoes were comfortable to wear without squeezing, avoiding any impact on the naturalness of gait due to discomfort. The power module had sufficient battery life, and the testing software module could switch normally between instantaneous mode, walking mode, treadmill mode, etc., to meet the testing needs of multiple scenarios.

[0047] The 3D gait analysis subsystem was debugged. Two depth cameras were installed 1.5 meters in front of the treadmill at a height of 40 cm and 2 meters behind the treadmill at a height of 70 cm, respectively, to adapt to the height and gait characteristics of the elderly and ensure that the full-body movement trajectory could be clearly captured. Hardware triggering synchronization was achieved through a USB-C cable, and an RS232 interface ensured stable transmission of treadmill speed data. The handrails were firmly installed, and reflective markers were accurately affixed to the key joints of the elderly's body to provide reliable feature points for 3D skeleton reconstruction.

[0048] Start the cloud server to confirm that the encrypted transmission function of the data communication module is normal, ensuring the security of the elderly's privacy data, and that the distributed storage and backup function of the data storage module is reliable, ensuring that the test data can be retained and traced for a long time, providing data support for subsequent follow-up.

[0049] The basic information of the elderly is entered into the detection software module and a unique identification code is bound to it, which makes it easy to associate multiple detection data and realize long-term dynamic monitoring.

[0050] After the elderly wear the test shoes, the wearable in-shoe foot pressure analysis subsystem is activated. The detection mode is selected as instantaneous mode, bipedal standing with eyes open / closed mode, and treadmill mode, which is adapted to the elderly's exercise ability and avoids safety risks caused by excessive exercise intensity. At the same time, it comprehensively covers scenarios such as instantaneous movements, static standing, and dynamic walking.

[0051] The 3D gait analysis subsystem is activated, and the data acquisition module is set to manual trigger mode. Data acquisition is initiated after the testing personnel confirm that the elderly person is ready, ensuring the safety of the elderly person during the testing process. Simultaneously, image data from two depth cameras, treadmill speed data, and handrail support force data are acquired. All data are bound to a high-precision timestamp consistent with the plantar pressure data to ensure data synchronization.

[0052] In instantaneous mode, the elderly perform instantaneous movements such as standing up and turning around. The instantaneous pressure and balance detection module of the wearable in-shoe plantar pressure analysis subsystem captures the sudden change data of plantar pressure during the movement. Combined with the pressure distribution symmetry and stability parameters, it analyzes whether there is a risk of imbalance and accurately identifies the potential for falls during instantaneous movements. The three-dimensional gait analysis subsystem simultaneously collects images of the changes in the whole body posture during the movement and records the movement trajectory of joint nodes, providing a basis for analyzing movement coordination.

[0053] In the bipedal standing mode with eyes open / closed, the standing balance detection module continuously monitors the movement trajectory of the center of pressure on the sole of the foot, the uniformity of pressure distribution, and the frequency of pressure changes in each area to assess static balance ability and identify the balance deficiencies of the elderly when standing. The three-dimensional gait analysis subsystem collects images of the whole body static posture and handrail support force data when standing to determine whether the elderly rely on handrails to maintain balance, thus supplementing the fall risk assessment dimension.

[0054] In treadmill mode, as elderly people walk slowly on the treadmill, the balance and gait detection module of the wearable in-shoe plantar pressure analysis subsystem simultaneously collects pressure data from nine areas of the sole and the migration pattern of the pressure center with the gait cycle. Combined with gait rhythm stability parameters, it assesses dynamic balance ability and captures abnormal distribution of plantar pressure during walking. The three-dimensional skeleton reconstruction module of the three-dimensional gait analysis subsystem detects 2D human key points through infrared recognition technology and optimizes 3D joint coordinates through a dual-view spatiotemporal constraint fusion algorithm, accurately capturing subtle abnormalities in joint movement during the elderly person's gait.

[0055] The collected data is transmitted to the detection software module via Bluetooth 5.0 and then synchronized to the cloud server. The data integration module integrates the plantar pressure data and 3D gait data into structured data according to timestamps, establishes index associations, and provides efficient data retrieval support for subsequent algorithm analysis.

[0056] The cloud server's data computing and analysis module processes structured data to generate analysis results such as plantar pressure distribution characteristics, response speed to sudden pressure changes, and range of pressure center movement. It identifies risk factors such as delayed plantar pressure response and uneven pressure distribution, providing quantitative basis for fall risk assessment.

[0057] The gait event detection module of the three-dimensional gait analysis subsystem adopts a multimodal gait event collaborative detection algorithm, which combines plantar pressure data and 3D joint motion data to accurately determine heel strike and toe lift events, analyze whether the gait cycle is regular and whether the stride length is symmetrical, and identify fall risk points related to gait stability.

[0058] The biomechanical parameter calculation module calculates the angle changes and joint torques of the hip, knee, and ankle joints, assesses whether the range of motion of the joints is normal and whether the force is coordinated, and determines whether there is gait instability due to joint function decline, further improving the analysis of fall risk factors.

[0059] By utilizing the big data reference range module of the three-dimensional foot and ankle data processing software, the balance parameters and gait parameters of the elderly are compared with the reference range of the same age and gender. Indicators that deviate significantly from the normal range (such as gait cycle variation coefficient higher than P95, plantar pressure center movement range greater than P95) are identified, and high-risk factors for falls are clarified, providing precise direction for the formulation of intervention measures.

[0060] The visualization report generation module generates reports that include 3D skeletal animation, pressure change curves, joint angle change curves, fall risk assessment indicators, and big data reference comparisons. These reports clearly demonstrate the specific manifestations of insufficient balance function and abnormal gait in the elderly, making it easier for medical staff to quickly grasp the core risk points.

[0061] Based on the reports, medical staff develop intervention measures, such as recommending custom-made orthotic insoles for uneven foot pressure distribution, developing home balance training plans for poor balance, and suggesting the use of walking aids for gait instability, ensuring that the intervention measures are targeted and feasible.

[0062] A retest is conducted every 6 months. Follow-up curves are generated through the individual development follow-up module to dynamically monitor the improvement of various indicators after intervention, adjust intervention measures in a timely manner, continuously reduce the risk of falls, and ensure the safety of elderly people when traveling.

[0063] In summary, this embodiment relies on a wearable plantar pressure and 3D gait analysis system, optimizing detection settings for the physical characteristics of the elderly. Through multi-scenario data collection, high-precision data correlation, and professional algorithm analysis, it comprehensively identifies fall risk factors related to standing balance, gait stability, and plantar pressure response. By combining big data reference ranges to clarify risk levels and dynamically tracking intervention effects through a follow-up module, it effectively compensates for the shortcomings of traditional screening methods that cannot quantify the synergistic relationship between overall posture and plantar pressure. This provides accurate and objective detection data and intervention directions for fall prevention in the elderly, effectively reducing fall risk and ensuring their safety in daily life.

[0064] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A wearable plantar pressure and three-dimensional gait analysis system, characterized in that, The system includes: a wearable in-shoe plantar pressure analysis subsystem and a three-dimensional gait analysis subsystem; The wearable in-shoe plantar pressure analysis subsystem includes a pressure sensor module, a pressure integration module, an intelligent transmission module, a detection software module, a power supply module, an instantaneous pressure and balance detection module, a standing balance detection module, a balance and gait detection module, and a cloud server. The three-dimensional gait analysis subsystem includes hardware units and software units; The wearable in-shoe plantar pressure analysis subsystem and the three-dimensional gait analysis subsystem use hardware synchronization and software algorithms to perform spatiotemporal alignment of plantar pressure data and three-dimensional gait data, multi-dimensional gait staging and event verification, and evaluation of whole-body posture stability and balance function. Furthermore, the two subsystems can perform bidirectional error correction when combined.

2. The wearable plantar pressure and three-dimensional gait analysis system according to claim 1, characterized in that, The wearable in-shoe plantar pressure analysis subsystem comprises the following modules: the pressure integration module includes a pressure testing insole and a testing shoe composed of nine pressure sensors, with the insole placed inside the shoe; the intelligent transmission module communicates with the detection software module via Bluetooth; the power module uses a rechargeable lithium battery pack and is equipped with a low-power power management unit, capable of automatically adjusting power supply according to system operating status; the instantaneous pressure and balance detection module captures sudden changes in plantar pressure during instantaneous movements, and, combined with the symmetry and stability parameters of pressure distribution, analyzes the balance state of the human body during instantaneous movements to determine whether... The system assesses whether there is a risk of imbalance. The standing balance detection module continuously monitors the movement trajectory of the plantar pressure center, the uniformity of pressure distribution, and the frequency of pressure changes in each area while the human is standing, evaluating the static balance ability of the human body. The balance and gait detection module simultaneously collects plantar pressure data and gait cycle-related information in dynamic gait scenarios. By analyzing the migration pattern of the pressure center with the gait cycle, the timing and duration of pressure peaks in each area, and combining gait rhythm stability parameters, it comprehensively evaluates the human body's dynamic balance ability and gait coordination. The cloud server includes a data communication module, a data integration module, a data calculation and analysis module, and a data storage module.

3. The wearable plantar pressure and three-dimensional gait analysis system according to claim 1, characterized in that, The specific contents of the three-dimensional gait analysis subsystem are as follows: The hardware unit includes a treadmill, two depth cameras, handrails, reflective markers, and a computer. The two depth cameras are located 1.5-2 meters in front of the treadmill at a height of 40-70cm and 1.5-2 meters behind the treadmill at a height of 30-80cm, respectively. The software unit includes a data acquisition module, a dual-view calibration and synchronization module, a three-dimensional skeleton reconstruction module, a gait event detection module, a biomechanical parameter calculation module, and a visualization report generation module.

4. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, In the three-dimensional gait analysis subsystem, the data acquisition module acquires RGB images, depth images, and infrared image data from two depth cameras; simultaneously, it acquires real-time speed and incline data of the treadmill via an RS232 interface, and acquires the subject's handrail support force data via a pressure sensor built into the handrail; the acquisition triggering methods include manual triggering and automatic triggering, with automatic triggering starting after image recognition determines that the subject has entered the center area of ​​the treadmill platform.

5. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, The dual-view calibration and synchronization module in the three-dimensional gait analysis subsystem adopts the Zhang Zhengyou calibration method, uses a 9×12 chessboard grid with a side length of 20mm, acquires 20-30 chessboard grid images from different angles, and calculates the intrinsic and extrinsic parameters of the two cameras. The intrinsic parameters include focal length, principal point coordinates, and distortion coefficients, while the extrinsic parameters include rotation matrix and translation vector. Data synchronization employs a combination of hardware triggering and software timestamps. Hardware-triggered synchronization is achieved by connecting two cameras via a USB-C cable, with a synchronization error of ≤1ms. Simultaneously, the software timestamps of each device are recorded for data alignment verification.

6. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, The 3D skeleton reconstruction module in the 3D gait analysis subsystem employs AI or infrared recognition technology to detect 2D human key points from RGB images acquired by two depth cameras. These 2D key points include corresponding points on the head, torso, upper limbs, and lower limbs. 3D coordinate calculation, combined with camera intrinsic and extrinsic parameters and depth image data obtained from the dual-view calibration and synchronization module, back-projects the 2D human key points to a unified world coordinate system to obtain their 3D coordinates. These coordinates are then fused using a dual-view spatiotemporal constraint fusion algorithm. The expression for this algorithm is as follows: ,in, The coordinates of joint i after fusion. It is a time variable. It represents the confidence weight of the original 3D coordinates of key point i captured by the camera before time t. These are the original 3D coordinates of key point i captured by the camera before time t. It is the confidence weight of the original 3D coordinates of key point i captured by the camera after time t. These are the original 3D coordinates of key point i captured by the camera after time t. It is a time continuity factor. These are the 3D coordinates of keypoint i at time t-1 after fusion. It represents the 3D coordinates of keypoint i at time t-2 after fusion.

7. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, The gait event detection module in the three-dimensional gait analysis subsystem aligns the plantar pressure data transmitted by the wearable shoe plantar pressure analysis subsystem with the three-dimensional skeleton data collected by the three-dimensional gait analysis subsystem according to timestamps. It extracts the hindfoot pressure value, vertical velocity of the heel key point, vertical acceleration of the knee joint, and pressure center trajectory change features from the aligned data. A multimodal gait event collaborative detection algorithm is employed, using the following criteria: hindfoot pressure value reaching a preset threshold, vertical velocity of the heel key point changing from negative to positive, and vertical acceleration of the knee joint within a preset range. This comprehensive assessment determines whether it is a heel-to-ground (HS) event. The algorithm further verifies the initially determined HS events by combining the timing of the pressure center shifting to the hindfoot with the knee extension angle change features, correcting any errors. Simultaneously, this module can also extend the toe-off (TO) event determination based on the same algorithm logic, extracting forefoot pressure value, vertical velocity of the toe key point, and ankle angle change features as the basis for TO event determination. The expression for the multimodal gait event collaborative detection algorithm is as follows: ,in, It is the result of the heel contact event at time t. =1, indicating that time t is determined to be a heel strike event; if =0 indicates that time t is determined to be a non-heel strike event. This is the hindfoot pressure value. This represents the greatest pressure on the hindfoot area in history. It is the pressure threshold coefficient. The vertical velocity of the heel. The vertical acceleration of the knee joint. It is the vertical acceleration threshold of the knee joint.

8. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, The biomechanical parameter calculation module in the three-dimensional gait analysis subsystem divides biomechanical parameters into kinematic parameters and dynamic parameters. The kinematic parameter calculation targets the hip, knee, ankle joints, and foot, extracting angle change data in the sagittal, coronal, and horizontal planes respectively. It automatically calls the optimized 3D joint coordinates output by the three-dimensional skeleton reconstruction module and the plantar pressure distribution data transmitted by the wearable shoe plantar pressure analysis subsystem as the basic input for parameter calculation. It compares the parameter change trends at different times within the same gait cycle and eliminates abnormal parameter values ​​that exceed the physiological range of motion.

9. The wearable plantar pressure and three-dimensional gait analysis system according to claim 3, characterized in that, The visualization report generation module in the 3D gait analysis subsystem receives kinematic parameters and dynamic parameters from the biomechanical parameter calculation module and gait cycle segmentation data from the gait event detection module. It also imports plantar pressure distribution data transmitted from the wearable in-shoe plantar pressure analysis subsystem. The module generates 3D model playback content, displaying human skeletal animation throughout the complete gait cycle, supporting viewpoint switching and playback speed adjustment. It generates data charts, presenting pressure change curves, joint angle change curves, and joint torque change curves in various plantar regions. It generates intervention comparison content, importing detection data from different time points for the same subject, enabling simultaneous display of data before, during, and after intervention, including parameter value comparisons and trend presentations. The module supports custom report templates, which include basic subject information, detection items, detection time, various visualization contents, and parameter summaries. The final output is a PDF report, which can include a signature area for the testing personnel and supports direct printing or electronic archiving.

Images