Parkinson's patient gait evaluation method and related equipment

Abstract

The invention belongs to the technical field of gait assessment, and discloses a Parkinson's patient gait assessment method and related equipment, and the method comprises the steps: obtaining a radar point cloud and a micro-Doppler time-frequency spectrum when a user walks in a test area through a millimeter wave radar, and carrying out the preprocessing of the radar point cloud; a human body point cloud cluster is extracted from each frame of radar point cloud after preprocessing through a clustering method, and mass center position tracks of left and right feet are recognized in combination with Kalman filtering and a Hungary algorithm; calculating the speeds of the left and right feet according to the centroid position tracks of the left and right feet, and detecting step events of the left and right feet in combination with the micro-Doppler time-frequency spectrum; calculating gait evaluation parameters according to the time sequence of the gait events and the centroid point position of the human body point cloud cluster; comparing the gait assessment parameters with pre-stored reference data to generate a Parkinson disease assessment result of the user; therefore, continuous and convenient rehabilitation monitoring can be provided for the Parkinson's disease, and the accuracy and practicability of Parkinson's disease assessment are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of gait assessment, in particular, to a Parkinson's patient gait assessment method and related equipment. BACKGROUND

[0002] Parkinson's disease is a common neurodegenerative disease, and its motor symptoms such as tremor and bradykinesia have subtle and variable characteristics, making it difficult to fully and accurately reflect the true condition of the disease during clinical consultation. In the disease assessment system of Parkinson's disease, gait assessment is an extremely important part and has an important reference function for the judgment of the rehabilitation status of patients.

[0003] Traditional gait analysis systems mainly rely on three-dimensional motion capture systems, force platforms, electromyography devices and other equipment in hospital laboratories. These devices are not only expensive and complex to operate, but also require professional personnel to operate and maintain, making it difficult to popularize their use in daily home environments, thereby limiting the continuous and convenient home rehabilitation monitoring of Parkinson's patients.

[0004] In recent years, the development of wearable sensing technology has provided a new solution for gait monitoring. However, contact sensors, such as the gait analysis device based on inertial sensors disclosed in Chinese patent CN120345891A, placed on the foot bottom, will to some extent affect the natural gait of patients, thereby affecting the accuracy of gait assessment. This contact monitoring method may cause the patient's gait to change when wearing the sensor, which cannot truly reflect their daily gait characteristics, thereby reducing the reliability of the assessment results. Therefore, the existing technology still faces problems such as expensive equipment, complex operation, difficulty in home application, and the influence of contact sensors on assessment accuracy in the assessment of Parkinson's patient gait.

[0005] In view of the above problems, the existing technology needs to be improved. SUMMARY

[0006] The purpose of the present application is to provide a Parkinson's patient gait assessment method and related equipment, aiming to solve the problems of expensive equipment, complex operation, difficulty in home application, and the influence of contact sensors on assessment accuracy in the assessment of Parkinson's patient gait in the prior art.

[0007] In a first aspect, the present application provides a Parkinson's patient gait assessment method, which comprises the following steps: A1. Using a millimeter wave radar to obtain radar echo signals of a user walking in a test area, generating a radar point cloud and a micro-Doppler time-frequency spectrum; A2. Coordinate conversion and normalization processing of the radar point cloud to obtain a pre-processed radar point cloud; A3. Extracting human point cloud clusters from each frame of pre-processed radar point cloud by clustering method, and combining Kalman filtering and Hungarian algorithm to identify the centroid position trajectory of left and right feet; A4. Calculating the speed of left and right feet according to the centroid position trajectory of left and right feet, and combining the micro-Doppler time-frequency spectrum to detect the step event of left and right feet; A5. Calculating gait assessment parameters according to the time sequence of the step event and the centroid point position of the human point cloud cluster; the gait assessment parameters include at least one of step frequency, step frequency mutation parameter, step length, walking speed, and gait symmetry index; A6. Comparing the calculated gait assessment parameters with pre-stored reference data to generate a Parkinson's disease assessment result of the user.

[0008] In a second aspect, the present application provides a gait assessment system for Parkinson's patients, which comprises a millimeter wave radar and a processing device; The millimeter wave radar is used to emit radar beams to a test area and receive radar echo signals, which are uploaded to the processing device; The processing device is used to obtain radar echo signals of a user walking in the test area by the millimeter wave radar, generate a radar point cloud and a micro-Doppler time-frequency spectrum; perform coordinate conversion and normalization processing on the radar point cloud to obtain a pre-processed radar point cloud; extract human point cloud clusters from each frame of pre-processed radar point cloud by clustering method, and combine Kalman filtering and Hungarian algorithm to identify the centroid position trajectory of left and right feet; calculate the speed of left and right feet according to the centroid position trajectory of left and right feet, and combine the micro-Doppler time-frequency spectrum to detect the step event of left and right feet; calculate gait assessment parameters according to the time sequence of the step event and the centroid point position of the human point cloud cluster; the gait assessment parameters include at least one of step frequency, step frequency mutation parameter, step length, walking speed, and gait symmetry index; compare the calculated gait assessment parameters with pre-stored reference data to generate a Parkinson's disease assessment result of the user.

[0009] In a third aspect, the present application provides an electronic device comprising a processor and a memory, wherein the memory stores a computer program executable by the processor, and when the processor executes the computer program, the steps of the gait assessment method for Parkinson's patients are run.

[0010] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the gait assessment method for Parkinson's patients are run.

[0011] Beneficial effects: The Parkinson patient gait assessment method and related device provided by the application provide a gait assessment method with low cost, simple operation and suitable for home environment by using non-contact millimeter wave radar technology, which can provide continuous and convenient rehabilitation monitoring for Parkinson patients, and significantly improve the accuracy and practicality of Parkinson disease assessment. Effectively solve the problems of expensive equipment, complex operation, difficult to apply at home and the influence of contact sensor on assessment accuracy in the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0012] Figure 1 The flowchart of the Parkinson patient gait assessment method provided by the application.

[0013] Figure 2 The schematic diagram of the Parkinson patient gait assessment system provided by the application.

[0014] Figure 3 The structural schematic diagram of the electronic device provided by the application.

[0015] Label explanation: 1, millimeter wave radar; 2, processing device; 301, processor; 302, memory; 303, communication bus. DETAILED DESCRIPTION

[0016] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. The components of the embodiments of the application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the application provided in the drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0017] It should be noted that: similar labels and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the application, the terms "first", "second" and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0018] Please refer to Figure 1 The Parkinson patient gait assessment method in some embodiments of the application includes the following steps: A1. Use millimeter wave radar to obtain radar echo signals of the user walking in the test area, generate radar point cloud and micro-Doppler time-frequency spectrum; A2. Coordinate transformation and normalization processing is performed on the radar point cloud to obtain a pre-processed radar point cloud; A3. Through a clustering method, a human point cloud cluster is extracted from each frame of the pre-processed radar point cloud, and a Kalman filter and a Hungarian algorithm are combined to identify the centroid position trajectories of the left and right feet; A4. The speeds of the left and right feet are calculated according to the centroid position trajectories of the left and right feet, and the step events of the left and right feet are detected in combination with a micro-Doppler time-frequency spectrum; A5. According to the time sequence of the step events and the centroid point positions of the human point cloud cluster, gait evaluation parameters are calculated; the gait evaluation parameters include at least one of step frequency, step frequency mutation parameter, step length, walking speed, and gait symmetry index; A6. The calculated gait evaluation parameters are compared with pre-stored reference data to generate a Parkinson's disease condition evaluation result of the user.

[0019] The application overcomes the interference of traditional contact sensors on the natural gait, avoids the high cost and complexity of large medical devices, and enables gait evaluation of Parkinson's patients in a daily home environment, thereby providing continuous and convenient rehabilitation monitoring for patients.

[0020] The millimeter wave radar is a radar system that uses millimeter wave frequency band electromagnetic waves for detection, has high resolution, the ability to penetrate non-metallic materials, and is not easily affected by environmental light, and is suitable for non-contact monitoring of human motion.

[0021] The radar echo signal is the signal reflected back by the electromagnetic wave emitted by the millimeter wave radar after encountering a target (such as a human body), which carries information such as the distance, speed, and angle of the target.

[0022] The radar point cloud is a three-dimensional point set generated after processing the radar echo signal, with each point representing a reflection point on the surface of the target and containing the spatial coordinates and speed information of the point.

[0023] The micro-Doppler time-frequency spectrum is a kind of frequency spectrum obtained by time-frequency analysis of the radar echo signal, which can reflect the frequency change caused by the small movement of the target (such as limb swinging), and is commonly used to identify and analyze fine movements in human gait.

[0024] The human point cloud cluster is a point set belonging to the human body identified by a clustering algorithm in the radar point cloud, which represents the body contour of the user at a certain time.

[0025] The centroid position trajectory refers to the continuous path of the geometric center of the human point cloud cluster or the foot point cloud cluster changing over time, which is used to describe the movement trend of the human body or the foot.

[0026] Step events refer to key events in the gait cycle, such as heel strike, toe off, etc., which are important time nodes for gait analysis.

[0027] Gait assessment parameters are indicators used to quantify and describe gait characteristics, such as step frequency, step length (distance per step), walking speed, gait symmetry, etc., which are of great significance for evaluating the motor function of Parkinson's patients.

[0028] The present application provides a method for evaluating the gait of Parkinson's patients, which is based on the non-contact sensing capability of millimeter wave radar, combined with advanced signal processing and data analysis algorithms, to achieve accurate evaluation of the gait of Parkinson's patients.

[0029] In step A1, the radar echo signals of the user walking in the test area are obtained by using the millimeter wave radar, and the radar point cloud and micro-Doppler time-frequency spectrum are generated. For example, the millimeter wave radar can be placed on one side of the test area, and when the user walks in the area, the radar will continuously emit millimeter wave beams and receive the reflected signals. These raw radar echo signals are then sent to the signal processing module, and through a series of digital signal processing techniques, radar point cloud data containing real-time three-dimensional position information of each part of the user's body can be extracted. At the same time, time-frequency analysis such as short-time Fourier transform is performed on the radar echo signals to generate a micro-Doppler time-frequency spectrum, which can capture the Doppler frequency changes caused by the user's limb micro-movements (such as foot swing), providing fine motion features for subsequent step event detection.

[0030] In step A2, the radar point cloud is subjected to coordinate transformation and normalization processing to obtain the preprocessed radar point cloud. For example, the coordinate system of the original radar point cloud is usually taken as the origin of the radar itself, in order to facilitate subsequent human body recognition and gait parameter calculation, the coordinates of the point cloud data can be uniformly converted to the ground coordinate system, so that the ground is the XY plane and the height is the Z axis. In addition, normalization processing can be performed on the point cloud data, such as adjusting the intensity or distance values of the point cloud to a unified range, to eliminate the differences caused by different radars or environmental factors, and ensure the consistency and comparability of the data.

[0031] In step A3, the human point cloud cluster is extracted from each pre-processed radar point cloud by clustering method, and the centroid position trajectory of left and right feet is identified by combining Kalman filter and Hungarian algorithm. Specifically, clustering algorithms such as DBSCAN, K-means, etc. can be used to process each pre-processed radar point cloud, and the point set belonging to the human body is clustered into a human point cloud cluster. For example, according to the density and spatial distribution characteristics of the point cloud, the densely connected points can be identified as a cluster. Once the human point cloud cluster is identified, the centroid point position can be calculated, which represents the overall position of the human body. In order to further identify the left and right feet, local sub-clustering can be performed within the human point cloud cluster to filter out the sub-cluster that meets the foot feature as a candidate foot point cloud cluster. Then, according to the relative position relationship between the candidate foot point cloud cluster and the centroid point position of the human point cloud cluster, they are distinguished as left foot point cloud cluster and right foot point cloud cluster. Finally, Kalman filter is used to smooth and predict the centroid position of left and right feet to reduce the influence of measurement noise, and Hungarian algorithm is combined for data association, so as to identify the centroid position trajectory of left and right feet, ensuring the continuity and accuracy of the trajectory.

[0032] In step A4, the speed of left and right feet is calculated according to the centroid position trajectory of left and right feet, and the step event of left and right feet is detected by combining the micro-Doppler time-frequency spectrum. For example, the moving speed of left and right feet at each time can be calculated by differentiating the centroid position trajectory of left and right feet. In order to more accurately detect the step event, the speed component of the moving speed of left and right feet in the forward direction of the user can be further calculated to obtain the forward speed of left and right feet. When the forward speed of the left foot or the right foot is close to zero (for example, less than a preset threshold), it can be preliminarily determined that the landing event occurs, and the time of the event is recorded. In order to verify whether the landing event is a real step event, the micro-Doppler time-frequency spectrum can be used for auxiliary judgment. By calculating the correlation degree between the peak time sequence of the reference micro-Doppler time-frequency spectrum and the peak time sequence of the control speed data, and comparing it with a preset threshold, it is determined whether the landing event is a step event.

[0033] In step A5, the gait evaluation parameters are calculated according to the time sequence of the step event and the centroid point position of the human point cloud cluster. These gait evaluation parameters include at least one of step frequency, step frequency mutation parameter, step length, walking speed, gait symmetry index. For example, the step frequency can be calculated by taking the inverse of the average step period of the time sequence of the step event. The walking speed can be obtained by calculating the average speed of the centroid point of each frame of human point cloud cluster. The step length can be obtained by dividing the walking speed by the step frequency. The step frequency mutation parameter can be obtained by analyzing the coefficient of variation of the step period in the time sequence of the step event. When the coefficient of variation exceeds a preset threshold, it is determined that a step frequency mutation event occurs. The gait symmetry index can be obtained by calculating the average step length difference between left and right feet.

[0034] In step A6, the calculated gait evaluation parameters are compared with pre-stored reference data to generate the Parkinson's disease assessment result of the user. For example, the calculated step frequency, step length, walking speed, gait symmetry index and other parameters can be compared with the reference data of healthy people or Parkinson's patients with different disease severity. These reference data can be the normal range, abnormal threshold or typical parameter values at different disease stages obtained by a large amount of clinical data statistics in advance. Through comparison, it can be determined whether the user's current gait characteristics deviate from the normal range and the degree of deviation, thereby generating the Parkinson's disease assessment result of the user, for example, the severity of the user's Parkinson's disease can be determined, or specific gait abnormality prompts can be given.

[0035] The Parkinson's patient gait evaluation method of the present application effectively solves the problem of interference of traditional contact sensors with the natural gait of patients through non-contact millimeter wave radar technology, improving the accuracy and reliability of the evaluation. Compared with traditional large medical equipment, the method has the advantages of low cost, simple operation and easy home deployment, so that Parkinson's patients can continuously and conveniently monitor their gait at home.

[0036] Specifically, the core innovation of the present application is: Firstly, the millimeter wave radar is used to obtain radar point cloud and micro-Doppler time-frequency spectrum, realizing non-contact and high-precision perception of user gait. The radar point cloud provides accurate three-dimensional position information of each part of the user's body, and the micro-Doppler time-frequency spectrum captures the fine features of small limb movements, which together provide a rich and comprehensive data basis for subsequent gait analysis. Compared with traditional gait analysis devices based on inertial sensors, the potential impact of sensor wearing on the patient's gait is avoided, ensuring the authenticity of the evaluation results.

[0037] Secondly, the present application realizes accurate identification of left and right foot center of mass position trajectories and reliable detection of step events through fine data processing and algorithms. Human point cloud clusters and foot point cloud clusters are extracted by clustering method, and Kalman filtering and Hungarian algorithm are combined to smooth and associate the foot trajectory, effectively overcoming the noise and uncertainty in radar data. In addition, the left and right foot speeds are combined with the micro-Doppler time-frequency spectrum to verify the step events, further improving the accuracy of step event detection and laying a foundation for accurate calculation of subsequent gait parameters. Compared with methods that rely solely on a single data source (such as speed threshold) to determine step events, the present application has higher robustness and accuracy.

[0038] Finally, the application can calculate multiple key gait assessment parameters and compare them with pre-stored reference data to generate an evaluation result of the user's Parkinson's disease. These parameters include step frequency, step frequency jump parameter, step length, walking speed, gait symmetry index, etc., which comprehensively reflect the gait characteristics of Parkinson's patients. By comparing with the reference data, the severity of the patient's condition can be objectively and quantitatively evaluated, providing an important reference for clinical diagnosis and rehabilitation treatment. This comprehensive evaluation method, compared with the evaluation method focusing on only one or a few gait parameters, can more comprehensively and accurately reflect the true condition of the patient.

[0039] In summary, the Parkinson's patient gait assessment method of the application has made significant progress in non-contact, high-precision, and easy-to-family application by introducing millimeter wave radar technology and advanced data processing algorithms, providing effective technical support for long-term monitoring and rehabilitation management of Parkinson's patients.

[0040] Preferably, step A1 comprises: A101. Obtain radar echo signals of the user walking towards or away from the millimeter wave radar in the test area using the millimeter wave radar; A102. Generate multiple frames of radar point clouds according to the radar echo signals; A103. Perform short-time Fourier transform on the radar echo signals to obtain micro-Doppler time-frequency spectrum.

[0041] Step A101 aims to ensure the effectiveness and consistency of radar signal acquisition. By guiding the user to walk towards or away from the millimeter wave radar in the test area (i.e. walking in the direction of the radar beam), the Doppler effect related to gait can be maximized to capture high-quality radar echo signals.

[0042] Further, in step A102, the obtained radar echo signals are processed to generate multiple frames of radar point clouds. The radar point cloud is a data set composed of a series of discrete points, each point representing a reflector detected by the radar beam in space. These points usually contain information such as distance, angle and velocity. By processing the continuous radar echo signals, multiple frames of point cloud data changing over time can be generated, reflecting the dynamic position and shape of the user in the test area.

[0043] Further, in step A103, a short-time Fourier transform (STFT) is performed on the radar echo signal to obtain the micro-Doppler time-frequency spectrum. Micro-Doppler effect refers to the slight change in frequency of the radar echo signal caused by internal movement or vibration of the target. For human gait, the swing of limbs such as feet and legs produces unique micro-Doppler characteristics. Short-time Fourier transform is a time-frequency analysis method that can convert time-domain signals into time-frequency domain representation, clearly showing the change of micro-Doppler information contained in the radar echo signal over time, forming a micro-Doppler time-frequency spectrum.

[0044] The scheme of the present application realizes systematic acquisition and preliminary processing of the original radar echo signal by refining step A1 into A101, A102 and A103. First, by limiting the walking direction of the user (towards or away from the millimeter wave radar), it ensures that the radar echo signal can effectively capture the key information related to gait, providing high-quality raw data for subsequent analysis, and also facilitating the subsequent differentiation of left and right foot point cloud clusters from candidate foot point cloud clusters. Second, converting the radar echo signal into multiple frames of radar point cloud makes the spatial position and motion trajectory of the user in the test area intuitively represented, laying a foundation for subsequent human recognition and foot centroid trajectory extraction. Finally, by performing a short-time Fourier transform on the radar echo signal to generate a micro-Doppler time-frequency spectrum, the fine features of limb movement can be captured from the frequency dimension, which is crucial for subsequent gait event detection, as the micro-Doppler time-frequency spectrum can reflect the unique frequency change pattern when the foot contacts the ground.

[0045] Through the above technical scheme, the steps of millimeter wave radar data acquisition and preliminary processing are clearly and refined. This refinement not only improves the standardization and effectiveness of data acquisition, ensuring that the acquired radar echo signal can fully reflect the gait characteristics of the user, but also provides two complementary and key data sources for subsequent gait parameter calculation by generating radar point cloud and micro-Doppler time-frequency spectrum respectively. Radar point cloud provides spatial position information, while micro-Doppler time-frequency spectrum provides fine motion frequency information, and the combination of the two significantly improves the accuracy and reliability of gait assessment.

[0046] In some embodiments, step A3 includes: A301. Extracting a human point cloud cluster from each pre-processed frame of radar point cloud by a clustering method, and obtaining the centroid point position of each frame of human point cloud cluster; A302. Performing local sub-clustering in each frame of human point cloud cluster, screening out sub-clusters that meet the pre-set height range and point cloud number threshold as candidate foot point cloud clusters, and obtaining the centroid point position of the candidate foot point cloud clusters; A303. According to the relative position relationship between the position of the centroid point of the candidate foot point cloud cluster and the position of the centroid point of the human body point cloud cluster, the candidate foot point cloud cluster is distinguished as a left foot point cloud cluster and a right foot point cloud cluster; A304. According to the position of the centroid point of the left foot point cloud cluster of each frame and the position of the centroid point of the right foot point cloud cluster of each frame, the Kalman filter and the Hungarian algorithm are used to identify the centroid position trajectory of the left and right feet.

[0047] In step A301, the clustering method can use DBSCAN, K-means or Mean-shift algorithm, the purpose of which is to group the points in the radar point cloud that belong to the same target, so as to separate the human body from the background noise or other sundries. The position of the centroid point of the human body point cloud cluster is usually obtained by calculating the average coordinates of all points in the cluster, and the centroid point can be used as an approximate position of the human body in the current frame.

[0048] Specifically, in step A302, since the feet are usually located at a lower position in the human body point cloud cluster and have certain point cloud quantity characteristics, a height range (for example, 0.05 meters to 0.3 meters from the ground) and a point cloud quantity threshold can be set to identify potential foot regions, i.e., for the identified sub-clusters, only when the height of the sub-cluster is within the preset height range and the number of points contained in the sub-cluster is not less than the point cloud quantity threshold, the sub-cluster is identified as a candidate foot point cloud cluster. Through local sub-clustering, the foot point cloud can be more finely separated from the leg or other parts of the body, and the centroid point positions of these candidate foot point cloud clusters are calculated.

[0049] In step A303, the centroid point of the human body point cloud cluster can be used as a reference to determine whether the candidate foot point cloud cluster is located on the left side or the right side. When the user walks towards or away from the millimeter wave radar, by comparing the lateral coordinates of the candidate foot point cloud cluster and the lateral coordinates of the centroid point of the human body point cloud cluster, the left and right feet can be effectively distinguished.

[0050] Specifically, in step A304, the Kalman filter is used to smooth and predict the centroid point positions of the left and right feet to reduce the influence of radar measurement noise and handle data loss or temporary occlusion, so as to obtain a more stable and continuous trajectory. The Hungarian algorithm is used to associate data between consecutive frames, ensuring that the left foot point cloud cluster and the right foot point cloud cluster identified in the current frame can be correctly matched with the left and right foot trajectories of the previous frame, avoiding confusion of the left and right foot identities due to instantaneous changes in the point cloud cluster.

[0051] The scheme of the present application realizes accurate identification of the left and right foot center of mass position trajectories in gait assessment of Parkinson patients by refining step A3 into the sub-steps A301 to A304 described above. First, by preliminary clustering of the radar point cloud (A301), the human body target can be effectively separated from the complex environment. On this basis, by local sub-clustering and combining height and point cloud quantity threshold (A302), the foot region can be accurately extracted from the human body point cloud, avoiding misidentification of other parts of the body as feet. Further, by analyzing the relative position between the foot point cloud cluster and the human body center of mass point (A303), the left foot and the right foot can be reliably distinguished, which is crucial for subsequent gait symmetry analysis. Finally, Kalman filtering is used to smooth and predict the center of mass position, effectively suppressing radar measurement noise and ensuring the continuity of the trajectory; while the Hungarian algorithm ensures correct data association in multi-target (left and right feet) tracking, avoiding trajectory crossing or identity confusion, thus ensuring the accuracy and robustness of left and right foot center of mass position trajectory identification.

[0052] Through the above technical scheme, the accuracy and stability of extracting human body point cloud clusters from radar point cloud data and identifying left and right foot center of mass position trajectories can be significantly improved. This detailed foot point cloud identification and trajectory tracking method effectively overcomes the problems of traditional methods, such as being easily disturbed by noise in complex environments, difficulty in distinguishing left and right feet, and trajectory jumping, laying a solid foundation for accurate calculation of subsequent gait assessment parameters. Thus, the present application can provide more reliable and detailed Parkinson disease assessment results, which is helpful for doctors to accurately judge the patient's condition and develop treatment plans.

[0053] In practical applications, it may be difficult to accurately distinguish between true step events and non-step events, such as brief pauses, body swaying or foot sliding, etc., only by the simple combination of speed and micro-Doppler time-frequency spectrum, thereby affecting the accuracy and reliability of gait assessment. If the above problem is not solved, it may lead to misjudgment of step events, thereby affecting the calculation accuracy of subsequent gait assessment parameters, and ultimately affecting the accuracy of Parkinson disease assessment results. In view of this, the present application further proposes a scheme to refine and optimize the above step of detecting step events to improve the accuracy and reliability of step event detection. Specifically, step A4 includes: A401. Calculate the moving speed of the left and right feet according to the center of mass position trajectories of the left and right feet; A402. Calculate the speed component of the moving speed of the left and right feet in the forward direction of the user to obtain the forward speed of the left and right feet; A403. When the forward speed of the left foot or the right foot is zero, determine that a landing event occurs, and record the landing event occurrence time; A404. Verify whether the landing event is a step event using the micro-Doppler time-frequency spectrum; A405.If the landing event is a step event, the corresponding landing event occurrence time is added to the time sequence of the step event.

[0054] In step A401, the centroid position trajectories of the left and right feet can be identified by the above-mentioned step A3. Based on these trajectory data, the instantaneous moving speed of the left and right feet at each time can be calculated by difference or filtering method. Specifically, the moving speed can be understood as the displacement of the centroid point in three-dimensional space per unit time.

[0055] Further, in step A402, in order to more accurately reflect the actual situation of the user's gait, it is necessary to decompose the moving speed of the left and right feet into the forward direction of the user. The forward direction of the user can be determined according to the overall moving direction of the human body point cloud cluster or the preset test area direction. By calculating the projection of the speed vector in the forward direction, the forward speed of the left and right feet can be obtained.

[0056] In step A403, when the forward speed of the left foot or the right foot approaches zero, it usually means that the foot is in contact with the ground and is in a relatively static state, i.e. a landing event occurs. In order to avoid noise interference, "zero speed" can be defined as a speed less than a preset minimum threshold. Once it is determined that a landing event occurs, the time point of its occurrence will be accurately recorded.

[0057] As a preferred embodiment, in step A404, in order to ensure the accuracy of the landing event, the present application introduces a verification mechanism of micro-Doppler time-frequency spectrum. The micro-Doppler time-frequency spectrum can reflect the fine features of human motion, especially the slight vibration and speed change when the foot is in contact with the ground. By analyzing the micro-Doppler time-frequency spectrum characteristics at the time of landing event, it can be judged whether the event meets the typical characteristics of step event, so as to exclude the misjudgment caused by other non-gait movements.

[0058] Therefore, in step A405, only the landing event that is verified by the micro-Doppler time-frequency spectrum as a step event will have its occurrence time added to the time sequence of the step event. In addition, in order to facilitate the subsequent calculation of gait symmetry index, the mark of whether the step event is produced by the left foot or the right foot can also be recorded at the same time.

[0059] The scheme of the present application effectively solves the problem of insufficient accuracy of step event detection in traditional methods by refining the detection process of step events into multiple logical steps and introducing a verification mechanism of micro-Doppler time-frequency spectrum. Specifically, first, by calculating the speed of the left and right foot centroid position trajectories and extracting the forward direction component (steps A401, A402), the landing event of the foot contacting the ground (step A403) can be preliminarily identified. This zero-speed-based judgment method can more intuitively capture the moment of foot landing. However, relying solely on speed information may not completely rule out false positives caused by non-gait movements such as swaying in place or foot sliding. To address this limitation, the present application further uses micro-Doppler time-frequency spectrum to verify the preliminarily determined landing event (step A404). Micro-Doppler time-frequency spectrum can provide fine features of foot movement, such as impact, slight sliding, etc., which will exhibit specific patterns on the time-frequency spectrum. By comparing the micro-Doppler time-frequency spectrum corresponding to the landing event with the preset step event characteristics, the true step event and non-step event can be effectively distinguished. Thus, only the landing event verified by the micro-Doppler time-frequency spectrum is confirmed as a step event and recorded (step A405), greatly improving the accuracy and reliability of step event detection and avoiding deviations in subsequent gait evaluation parameters caused by false positives.

[0060] Through the above technical scheme, the present application can provide a more accurate and reliable step event detection method. Compared with schemes that rely solely on speed or simply combine micro-Doppler time-frequency spectrum, the present application effectively eliminates the interference of non-gait movements on step event determination by introducing forward speed component analysis and a verification mechanism of micro-Doppler time-frequency spectrum, significantly reducing the false positive rate. Thus, the detected step event time sequence is more accurate, laying a solid foundation for the accurate calculation of subsequent gait evaluation parameters such as step frequency, step length, walking speed, and gait symmetry indicators, ultimately improving the accuracy and clinical reference value of the evaluation results of Parkinson's patients.

[0061] Preferably, step A404 can include: extracting the micro-Doppler time-frequency spectrum within a preset length of time window centered on the landing event occurrence time as a reference micro-Doppler time-frequency spectrum; extracting the peak time sequence of the reference micro-Doppler time-frequency spectrum, denoted as the first peak time sequence; extracting the moving speed of the foot corresponding to the landing event within the time window as the control speed data; extracting the peak time sequence of the control speed data, denoted as the second peak time sequence; calculating the correlation degree of the first peak time sequence and the second peak time sequence; if the correlation degree is greater than a preset correlation degree threshold, the landing event is determined as a step event.

[0062] Specifically, the preset length of the time window can be set according to the actual application scenario and the radar data sampling rate, for example, it can be set to 0.5-2 seconds, the purpose of which is to capture sufficient micro-Doppler features and foot speed change information before and after the landing event. The peak time sequence of the micro-Doppler time-frequency spectrum refers to the sequence of time points at which the energy or amplitude in the micro-Doppler time-frequency spectrum reaches a local maximum within the time window. These peaks usually correspond to specific phases of foot movement, such as toe-off, heel strike, etc. The contrast speed data refers to the foot movement speed data obtained by differentiating the foot center of mass position trajectory within the same time window. The peak time sequence of the contrast speed data refers to the sequence of time points at which the foot movement speed reaches a local maximum within the time window, these peaks reflect the speed changes of the foot in the swing or support phase. The correlation can be calculated using various statistical methods such as Pearson correlation coefficient, cross-correlation function, etc., the purpose of which is to quantify the consistency or synchronization between the two time sequences. The preset correlation threshold is a discriminant criterion for determining whether the landing event is a step event, this threshold can be adjusted according to experimental data and expert experience to balance the sensitivity and specificity of detection.

[0063] The scheme of the present application can more accurately verify whether the landing event is a step event by performing correlation analysis on the peak time sequence of the micro-Doppler time-frequency spectrum and the peak time sequence of the foot movement speed. This is because when a real step event occurs, there is a high degree of synchronization and internal correlation between the movement state of the foot (including speed change) and the micro-Doppler effect it produces. For example, the foot reaches a peak speed during the swing process, and at the same time, the micro-Doppler time-frequency spectrum will also have a characteristic peak at the corresponding time point. By calculating the correlation of the two peak time sequences, this internal correlation can be effectively captured. If the correlation is high, it indicates that the foot movement and the micro-Doppler features are highly consistent, thereby enhancing the confidence of step event determination and avoiding misjudgment of non-gait-related ground contact as a step event.

[0064] Through the above technical scheme, the present application can significantly improve the accuracy and robustness of step event detection. By combining micro-Doppler time-frequency spectrum and foot movement speed data in two different dimensions for cross-validation, false positives caused by other interference factors (such as body sway, ground friction, etc.) can be effectively excluded, ensuring that the detected step event is a real event closely related to the gait cycle. This is crucial for the accurate calculation of subsequent gait evaluation parameters (such as step frequency, step length, etc.), thereby providing more reliable and accurate data support for the condition assessment of Parkinson's patients.

[0065] In step A5, at least one of the step frequency, the step frequency mutation parameter, the step length, the walking speed, and the gait symmetry index needs to be calculated. The gait evaluation parameters can be calculated in the following manner.

[0066] The step of calculating the step frequency can include: The average step cycle is calculated according to the time sequence of the step events, and the reciprocal of the average step cycle is calculated as the step frequency.

[0067] By analyzing the time sequence of the step events, the time intervals between adjacent step events can be determined, and the average of these time intervals is calculated to obtain the average step cycle. The step frequency can be understood as the number of steps occurring per unit time, and its calculation method is the reciprocal of the average step cycle, which aims to quantify the user's walking rhythm.

[0068] The step of calculating the step frequency mutation parameter can include: A continuous segment of a preset duration (which can be set according to actual needs) is extracted from the time sequence of the step events as a reference time sequence; A window of a preset number of steps (which can be set according to actual needs, for example, 10 steps) is used to extract multiple groups of window data in the reference time sequence; The step cycle standard deviation of each group of window data is calculated, and the coefficient of variation corresponding to each group of window data is obtained by dividing the step cycle standard deviation by the step frequency; When the coefficient of variation exceeds a preset coefficient threshold (which can be set according to actual needs), it is determined that a step frequency mutation event has occurred, and the number of occurrences of the step frequency mutation event is counted as the step frequency mutation parameter.

[0069] The step frequency mutation parameter is used to evaluate the stability of the user's gait. By sliding the window in the reference time sequence and calculating the step cycle standard deviation and the coefficient of variation, the abnormal fluctuations of the gait rhythm can be detected. When the coefficient of variation, i.e., the ratio of the step cycle standard deviation to the step frequency, exceeds the preset coefficient threshold, it indicates that the stability of the gait has decreased significantly, i.e., a step frequency mutation event has occurred. Counting the number of occurrences of these events (the number of occurrences within a preset duration) can quantify the degree of instability of the user's gait.

[0070] The step of calculating the walking speed can include: According to the positions of the centroid points of each frame of human point cloud clusters, the average speed of the centroid points of the human point cloud clusters is calculated as the walking speed.

[0071] Specifically, the positions of the centroid points of the human point cloud clusters are extracted in step A3, which represent the overall position of the user's body. By tracking the displacement of these centroid points between consecutive frames and combining the frame rate (the reciprocal of the frame rate is the inter-frame time interval), the average moving speed of the user in the test area, i.e., the walking speed, can be calculated.

[0072] The step of calculating the step length can include: The step length is obtained by dividing the walking speed by the step frequency.

[0073] In practical applications, the step length refers to the average forward distance of each step of the user. In the case of calculating the walking speed and the step frequency, the step length can be directly calculated by dividing the walking speed by the step frequency, and the purpose is to reflect the amplitude of the user's gait.

[0074] The step of calculating the gait symmetry index can include: According to the positions of the center points of the human point cloud clusters and the time sequence of the step events, the single-step length corresponding to each step event is calculated; for example, the moving distance of the center point positions of the human point cloud clusters within the time interval of adjacent step events is obtained, and the moving distance is divided by the time interval of the adjacent step events to obtain the single-step length corresponding to the corresponding step event; According to the single-step length corresponding to each step event, the left foot average step length and the right foot average step length are calculated; for example, according to the marks of each step event produced by the left foot or the right foot recorded in advance, the average value of the single-step length corresponding to the left foot is calculated to obtain the left foot average step length, and the average value of the single-step length corresponding to the right foot is calculated to obtain the right foot average step length; The gait symmetry index is calculated according to the following formula: ; Wherein, is the gait symmetry index, is the left foot average step length, is the right foot average step length.

[0075] The gait symmetry index SLA is used to evaluate the coordination of the left and right foot gaits. By calculating the average step length of the left foot and the right foot respectively and , and substituting into the above formula, the difference between the left and right foot step lengths can be quantified. The calculation of this index helps to identify the common gait asymmetry problem of Parkinson's patients.

[0076] Through the above technical solutions, the present application can provide a set of comprehensive and accurate gait evaluation parameter calculation method. These parameters not only include the basic step frequency, step length and walking speed, but also introduce the step frequency mutation parameter which has important indicative significance for gait stability, and the gait symmetry index which reflects the coordination of the left and right feet. These detailed and quantitative parameters can provide more rich and objective data support for the gait evaluation of Parkinson's patients, thereby helping doctors to more accurately judge the severity and progression of the patient's condition, and providing a scientific basis for subsequent treatment and rehabilitation.

[0077] In some embodiments, step A6 includes: The gait evaluation parameters are weighted and integrated to obtain a comprehensive score; The comprehensive score is compared with reference data of Parkinson's patients with different disease severity to determine the user's Parkinson's disease severity as the Parkinson's disease evaluation result; the reference data is a reference range of the comprehensive score.

[0078] Specifically, after calculating a plurality of gait evaluation parameters such as step frequency, step frequency mutation parameter, step length, walking speed, gait symmetry index, etc., these parameters need to be weighted and integrated. The weighted integration refers to assigning a weight coefficient to each parameter according to its relative importance in reflecting the severity of Parkinson's disease. For example, some parameters may be more closely related to specific symptoms of Parkinson's disease (such as bradykinesia, gait instability), and therefore can be given a higher weight. By multiplying the standardized value of each parameter by its corresponding weight and adding up all the weighted parameter values, a comprehensive score can be obtained. This comprehensive score can more comprehensively reflect the overall gait condition of the patient.

[0079] Further, the obtained comprehensive score will be used for comparison with pre-stored reference data. The reference data is not a threshold for a single gait parameter, but a reference range for the comprehensive score, which corresponds to Parkinson's patients with different disease severity. For example, a comprehensive score range can be pre-set to correspond to "mild Parkinson's disease", another range to correspond to "moderate Parkinson's disease", and another range to correspond to "severe Parkinson's disease". By falling into a specific reference range, the user's comprehensive score can determine the user's Parkinson's disease severity and be used as the final Parkinson's disease evaluation result. The reference data is usually established by statistical analysis of gait data of a large number of Parkinson's patients with known disease severity.

[0080] The scheme of the present application effectively solves the limitations of single or independent parameter evaluation in traditional methods by introducing weighted integration of gait evaluation parameters. Specifically, since Parkinson's disease has a multi-faceted impact on gait, a single parameter often cannot fully characterize the disease. By weighting and integrating multiple parameters such as step frequency, step length, symmetry, etc., the abnormal features of the patient's gait can be more comprehensively captured, and appropriate weights can be assigned according to their different contributions to disease evaluation, thereby generating a comprehensive score that can more accurately reflect the overall condition of the patient. In addition, comparing the comprehensive score with reference ranges for different disease severity, rather than only with a single threshold, allows the evaluation result to more finely distinguish the disease stage of the patient, avoiding evaluation bias caused by mutual influence between parameters or individual differences.

[0081] By means of the technical solutions, the Parkinson's disease assessment method provided by the application is more comprehensive, accurate and refined. Compared with the traditional method which only relies on a single or simple comparison of multiple gait parameters, the application generates a more representative comprehensive score by weighting and synthesizing each gait evaluation parameter, thereby significantly improving the reliability and clinical practicability of the evaluation result. In addition, by comparing the comprehensive score with the reference range of different disease severity, the disease stage of the Parkinson's patient can be more effectively identified and distinguished, which provides a more accurate basis for clinical diagnosis, treatment plan formulation and efficacy evaluation, and helps to realize early detection and dynamic management of the disease of the Parkinson's patient.

[0082] Reference Figure 2 The application provides a gait assessment system for Parkinson's patients, which comprises a millimeter wave radar 1 and a processing device 2. The millimeter wave radar 1 is used to emit radar beams to the test area and receive radar echo signals, which are uploaded to the processing device 2. The processing device 2 is used to obtain the radar echo signals of the user when walking in the test area by means of the millimeter wave radar 1, generate a radar point cloud and a micro-Doppler time-frequency spectrum, perform coordinate conversion and normalization processing on the radar point cloud to obtain a preprocessed radar point cloud, extract a human point cloud cluster from each frame of the preprocessed radar point cloud by means of a clustering method, and identify the centroid position trajectories of the left and right feet in combination with Kalman filtering and the Hungarian algorithm. The speed of the left and right feet is calculated according to the centroid position trajectories of the left and right feet, and the step events of the left and right feet are detected in combination with the micro-Doppler time-frequency spectrum. The gait evaluation parameters are calculated according to the time sequence of the step events and the centroid point positions of the human point cloud cluster. The gait evaluation parameters include at least one of the step frequency, the step frequency mutation parameter, the step length, the walking speed and the gait symmetry index. The calculated gait evaluation parameters are compared with the pre-stored reference data to generate the Parkinson's disease assessment result of the user (for details, refer to steps A1-A6 in the foregoing).

[0083] The millimeter wave radar 1 can be a separate radar sensor module, which has a working frequency of, for example, 60GHz-70GHz and a maximum detection distance of, for example, 5m-10m, and can provide high-resolution distance, speed and angle information. In actual application, the millimeter wave radar 1 can be fixedly installed on the wall or support of the test area to ensure stable monitoring of the user's gait. For example, the millimeter wave radar 1 can be configured as a single-transmit multi-receive (MIMO) array to improve the spatial resolution and target recognition capability. In addition, the millimeter wave radar 1 can also be a module integrated in a smart home device, such as a smart speaker or a smart display, to realize more convenient deployment and use.

[0084] The processing device 2 is used to receive the radar echo signal uploaded by the millimeter wave radar 1, and perform a series of complex signal processing and data analysis tasks. The processing device 2 can be a high-performance central processing unit (CPU), such as an embedded processor, a dedicated digital signal processor (DSP), or a programmable logic gate array (FPGA) to meet the real-time data processing requirements. The processing device 2 can be configured as a local computing unit, directly integrated in the evaluation device, or connected to a cloud server through a network for data processing and storage.

[0085] Please refer to Figure 3 A structure schematic diagram of an electronic device provided by the embodiment of the present application, the present application provides an electronic device, comprising: a processor 301 and a memory 302, the processor 301 and the memory 302 are interconnected and communicate with each other through a communication bus 303 and / or other forms of connection mechanism (not marked), the memory 302 stores a computer program executable by the processor 301, when the electronic device runs, the processor 301 executes the computer program to execute the gait assessment method for Parkinson's patients in any optional implementation manner of the above-mentioned embodiment, to realize the following functions: using millimeter wave radar to obtain radar echo signal of user walking in test area, generating radar point cloud and micro-Doppler time-frequency spectrum; coordinate conversion and normalization processing are carried out on the radar point cloud, and the preprocessed radar point cloud is obtained; through clustering method, human point cloud cluster is extracted from each frame of preprocessed radar point cloud, and the centroid position trajectory of left and right feet is identified by combining Kalman filtering and Hungarian algorithm; the speed of left and right feet is calculated according to the centroid position trajectory of left and right feet, and the step event of left and right feet is detected by combining micro-Doppler time-frequency spectrum; according to the time sequence of step event and the centroid point position of human point cloud cluster, the gait assessment parameter is calculated; the gait assessment parameter includes at least one of step frequency, step frequency mutation parameter, step length, walking speed and gait symmetry index; the calculated gait assessment parameter is compared with the prestored reference data, and the Parkinson's disease assessment result of the user is generated.

[0086] The embodiment of the application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to execute the Parkinson patient gait evaluation method in any optional implementation manner of the above embodiment, so as to realize the following functions: radar echo signals of a user walking in a test area are acquired by using a millimeter wave radar, a radar point cloud and a micro-Doppler time-frequency spectrum are generated; coordinate conversion and normalization processing are performed on the radar point cloud, and a preprocessed radar point cloud is obtained; a human point cloud cluster is extracted from each frame of the preprocessed radar point cloud by using a clustering method, and a centroid position trajectory of a left foot and a right foot is identified by combining Kalman filtering and a Hungarian algorithm; the speed of the left foot and the right foot is calculated according to the centroid position trajectory of the left foot and the right foot, and a step event of the left foot and the right foot is detected by combining the micro-Doppler time-frequency spectrum; gait evaluation parameters are calculated according to a time sequence of the step event and a centroid point position of the human point cloud cluster; the gait evaluation parameters include at least one of a step frequency, a step frequency mutation parameter, a step length, a walking speed and a gait symmetry index; the calculated gait evaluation parameters are compared with prestored reference data, and a Parkinson disease evaluation result of the user is generated. The computer readable storage medium can be realized by any type of volatile or nonvolatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk.

[0087] The above merely describes the embodiments of the application, but does not serve to limit the protection scope of the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. A method of gait assessment in Parkinson's patients, characterized in that, The method comprises the following steps: A1. acquiring radar echo signals of a user walking in a test area by using a millimeter wave radar, generating a radar point cloud and a micro-Doppler time-frequency spectrum; A2. performing coordinate conversion and normalization processing on the radar point cloud to obtain a preprocessed radar point cloud; A3. extracting a human point cloud cluster from each frame of the preprocessed radar point cloud by using a clustering method, and identifying the centroid position trajectory of the left and right feet by combining Kalman filtering and the Hungarian algorithm; A4. calculating the speed of the left and right feet according to the centroid position trajectory of the left and right feet, and detecting the step event of the left and right feet in combination with the micro-Doppler time-frequency spectrum; A5. calculating gait evaluation parameters according to the time sequence of the step event and the centroid point position of the human point cloud cluster; the gait evaluation parameters include at least one of step frequency, step frequency mutation parameter, step length, walking speed, and gait symmetry index; A6. comparing the calculated gait evaluation parameters with pre-stored reference data to generate a Parkinson's disease evaluation result of the user.

2. The method for gait assessment of Parkinson's patients according to claim 1, characterized in that, Step A1 comprises: A101. acquiring radar echo signals of a user walking towards or away from the millimeter wave radar in a test area by using a millimeter wave radar; A102. generating multiple frames of radar point clouds according to the radar echo signals; A103. performing short-time Fourier transform on the radar echo signals to obtain the micro-Doppler time-frequency spectrum.

3. The method of gait assessment of a Parkinson's patient according to claim 1, characterized in that, Step A3 comprises: A301. extracting a human point cloud cluster from each frame of the preprocessed radar point cloud by using a clustering method, and acquiring the centroid point position of each frame of the human point cloud cluster; A302. performing local sub-clustering in each frame of the human point cloud cluster, screening out sub-clusters meeting a pre-set height range and point cloud quantity threshold as candidate foot point cloud clusters, and acquiring the centroid point position of the candidate foot point cloud clusters; A303. according to the relative position relationship between the centroid point positions of the candidate foot point cloud clusters and the human point cloud cluster, distinguishing the candidate foot point cloud clusters into left foot point cloud clusters and right foot point cloud clusters; A304. according to the centroid point positions of each frame of the left foot point cloud cluster and each frame of the right foot point cloud cluster, identifying the centroid position trajectory of the left and right feet by using Kalman filtering and the Hungarian algorithm.

4. The method of gait assessment of a Parkinson's patient according to claim 1, characterized in that, Step A4 comprises: A401. calculating the moving speed of the left and right feet according to the centroid position trajectory of the left and right feet; A402. calculating the speed component of the moving speed of the left and right feet in the forward direction of the user to obtain the forward speed of the left and right feet; A403. when the forward speed of the left foot or the right foot is zero, determining that a landing event occurs, and recording the landing event occurrence time; A404. verifying whether the landing event is a step event by using the micro-Doppler time-frequency spectrum; A405. if the landing event is a step event, adding the landing event occurrence time corresponding to the step event to the time sequence of the step event.

5. The method of gait assessment of a Parkinson's patient according to claim 4, characterized in that, Step A404 comprises: extracting the micro-Doppler time-frequency spectrum in a preset length of time window centered on the landing event occurrence time as a reference micro-Doppler time-frequency spectrum; extracting the peak time sequence of the reference micro-Doppler time-frequency spectrum, denoted as a first peak time sequence; extracting the moving speed of the foot corresponding to the landing event in the time window as control speed data; extract a peak time sequence of the control speed data, denoted as a second peak time sequence; calculate a correlation degree of the first peak time sequence and the second peak time sequence; if the correlation degree is greater than a preset correlation degree threshold, determine that the landing event is a step event.

6. The method of gait assessment of a Parkinson's patient according to claim 1, characterized in that, The step of calculating the step frequency comprises: calculating an average step period according to a time sequence of the step event, and calculating a reciprocal of the average step period as the step frequency; The step of calculating the step frequency mutation parameter comprises: extracting a continuous segment of a preset time length from the time sequence of the step event as a reference time sequence; extracting a plurality of groups of window data in a preset step number window in the reference time sequence by sliding window extraction; calculating a step period standard deviation of each group of window data, and dividing the step period standard deviation by the step frequency to obtain a coefficient of variation corresponding to each group of window data; when the coefficient of variation exceeds a preset coefficient threshold, determining that a step frequency mutation event occurs, and counting the number of step frequency mutation events as the step frequency mutation parameter; The step of calculating the walking speed comprises: calculating a center of mass point average speed of the human body point cloud cluster according to the center of mass point position of each frame of the human body point cloud cluster, as the walking speed; The step of calculating the step length comprises: dividing the walking speed by the step frequency to obtain the step length; The step of calculating the gait symmetry index comprises: calculating a single step length corresponding to each step event according to the center of mass point position of each frame of the human body point cloud cluster and the time sequence of the step event; calculating a left foot average step length and a right foot average step length according to the single step length corresponding to each step event; calculating the gait symmetry index according to the following formula: ； wherein, is the gait symmetry index, is the left foot average stride length, is the right foot average stride length.

7. The method of gait assessment of a Parkinson's patient according to claim 1, characterized in that, Step A6 comprises: weighting and synthesizing each gait evaluation parameter to obtain a comprehensive score; comparing the comprehensive score with reference data of Parkinson's patients with different disease severity stored in advance to determine the disease severity of the user as a Parkinson's disease severity evaluation result; the reference data is a reference range of the comprehensive score.

8. A system for gait assessment of Parkinson's patients, characterized by, The system comprises a millimeter wave radar and a processing device; The millimeter wave radar is used to emit radar beams to a test area and receive radar echo signals uploaded to the processing device; The processing device is used to obtain radar echo signals of the user walking in the test area by the millimeter wave radar, generate a radar point cloud and a micro-Doppler time-frequency spectrum; coordinate conversion and normalization processing are performed on the radar point cloud to obtain a preprocessed radar point cloud; human body point cloud clusters are extracted from each frame of the preprocessed radar point cloud by a clustering method, and the center of mass position trajectories of the left and right feet are identified by combining Kalman filtering and the Hungarian algorithm; the speeds of the left and right feet are calculated according to the center of mass position trajectories of the left and right feet, and the step events of the left and right feet are detected by combining the micro-Doppler time-frequency spectrum; According to the time sequence of the step event and the center of mass point position of the human body point cloud cluster, gait evaluation parameters are calculated; the gait evaluation parameters include at least one of the step frequency, the step frequency mutation parameter, the step length, the walking speed, and the gait symmetry index; the calculated gait evaluation parameters are compared with the pre-stored reference data to generate a Parkinson's disease severity evaluation result of the user.

9. An electronic device, comprising: A computer program product comprising a computer readable medium having stored thereon computer program means which, when executed by a processor, cause the processor to perform the steps of the method of gait assessment for Parkinson's patients according to any of claims 1-7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to run the steps of the method of gait assessment for Parkinson's patients according to any of claims 1-7.

Images