An elderly exercise fall risk assessment method, system, device and medium

Abstract

This invention discloses a method, system, device, and medium for assessing the risk of falls during exercise in the elderly, relating to the field of risk assessment technology. The steps include: collecting raw signals from an IMU, plantar pressure sensor, and sEMG during the elderly person's movement; mapping the raw signals onto the spatial coordinates of key points on the human body, performing ellipse fitting on a discrete set of points at specific locations to obtain a geometric representation of the movement posture; constructing a fall behavior feature vector by integrating the spatiotemporal information of the key points; scoring each indicator in the fall behavior feature vector according to a preset dynamic gait index scoring rule, and combining the scores to obtain a risk score and outputting the corresponding risk level. This invention, by comprehensively analyzing subtle changes in head height, minor disturbances in the movement trajectories of the knees / ankles, and geometric distortions of the movement envelope ellipse, can quantitatively score the elderly person before they actually lose balance, providing precise intervention guidance.

Description

Technical Field

[0001] This invention relates to the field of risk assessment technology, and in particular to a method, system, device and medium for assessing the risk of falls during exercise in the elderly. Background Technology

[0002] With the increasing aging of the global population, falls among the elderly have become a major public health problem leading to serious injuries, disability, and even death. While exercise can slow functional decline, it also increases the risk of falls. Therefore, real-time and accurate fall risk assessments during exercise are crucial for older adults.

[0003] Existing assessment methods include subjective assessment and objective assessment based on single-modal sensors. Subjective assessment methods, such as those relying on the Berg Balance Scale, depend on the therapist's experience and observation, resulting in low efficiency, high subjectivity, and difficulty in achieving dynamic continuous monitoring. Objective assessment methods based on single-modal sensors, such as those using only accelerometers for gait analysis or only cameras for posture estimation, are insufficient for capturing crucial "micro-vibration signals" (such as subtle muscle tremors and postural swaying) caused by the degeneration of the proprioceptive system and the attenuation of peripheral nerve conduction in older adults, as described in spatiotemporal graph convolutional networks and vision-based posture estimation. The fundamental reason is that physiological degeneration leads to a significant decrease in the signal-to-noise ratio (SNR) of these crucial signals. Single-modal sensors cannot effectively separate these weak features indicating imbalance from noise, resulting in "input-level systematic errors" in risk assessment and limiting accuracy.

[0004] Existing technologies include fall risk assessment systems for the elderly based on single or dual-modal wearable sensors. These systems use inertial measurement units (IMUs) strapped to the waist or ankle to assess falls by analyzing macroscopic parameters such as gait cycles and trunk tilt angles. However, relying solely on single or dual-modal sensors limits the kinematic information acquired, making it difficult to comprehensively reflect the complex neuromuscular control and balance regulation mechanisms preceding falls. Furthermore, due to the limited signal dimension, the system is insensitive to early subtle imbalances caused by neuromuscular function decline in the elderly (such as muscle tremors and abnormal foot pressure distribution), resulting in delayed risk assessment. Alarms are often only triggered when there is significant tilting or large-scale stride imbalance, lacking true pre-fall prediction capabilities. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by providing a method, system, device, and medium for assessing the risk of falls during exercise in the elderly, thereby solving the problems in the prior art.

[0006] The present invention specifically provides the following technical solution: A method for assessing the risk of falls during exercise in older adults includes: Raw signals from an inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor were collected during the movement of elderly individuals. The preprocessed original signal is mapped onto the spatial coordinates of key points on the human body. Based on the spatial coordinates of the key points, the least squares method is used to fit an ellipse to the discrete point set corresponding to the motion envelope of the torso and lower limbs to obtain a geometric representation of the motion posture. The geometric representation includes the ellipse center, major axis, minor axis and tilt angle parameters. A fall behavior feature vector is constructed by integrating spatiotemporal information of key points on the human body; the fall behavior feature vector includes changes in head height, changes in lower limb joint height, and geometric representation of movement posture; wherein changes in head height are obtained by the vertical distance between the center of the head key point and the ground, and changes in lower limb joint height are obtained by the instantaneous changes or standard deviations in the height of the two knees and two ankles; According to the preset dynamic gait index scoring rules, each indicator in the fall behavior feature vector is scored, and the risk score is obtained by combining the scores. The risk score is compared with the preset threshold to output the corresponding risk level.

[0007] Preferably, the step of using the least squares method to perform ellipse fitting on the discrete point set corresponding to the motion envelope of the trunk and lower limbs to obtain a geometric representation of the motion posture specifically involves: The least squares method is used to fit an ellipse to the discrete point set corresponding to the motion envelope of the trunk and lower limbs, obtaining the parameters of the ellipse center (x_c, y_c), major axis a, minor axis b, and tilt angle θ. All parameters are used as a geometric representation of the motion posture. The specific expression for the ellipse fitting is as follows: Ax² + Bxy + Cy² + Dx + Ey + F = 0; Where A, B, C, D, E, and F are the correlation coefficients, and x and y represent the discrete points corresponding to the trunk and lower limb motion envelopes.

[0008] Preferably, before mapping the preprocessed original signal onto the spatial coordinates of key points on the human body, the method further includes: The original signal is preprocessed, and the preprocessed multimodal signal is regarded as a whole. The signal processing model is used for fusion analysis to obtain the enhanced motion micro-vibration characteristics from electromyographic noise, instantaneous pressure fluctuations and inertial micro-vibrations.

[0009] Preferably, when constructing the fall behavior feature vector by integrating the spatiotemporal information of key human body points, the method further includes: extracting time-frequency domain features from the original signal and combining them with the spatiotemporal information of key human body points to form the fall behavior feature vector.

[0010] Preferably, the preset dynamic gait index scoring rule is as follows: The assessment criteria for multiple tasks in the Dynamic Gait Index (DGI) scale are simulated, and corresponding scoring rules are set for each indicator in the fall behavior feature vector; the multiple tasks include walking on flat ground, walking with head turned, and crossing obstacles.

[0011] Preferably, the step of comparing the risk score with a preset threshold and outputting the corresponding risk level specifically involves: Set the total risk score to m points and set n preset thresholds. When the risk score falls within different preset threshold ranges, output the corresponding risk level.

[0012] This invention provides a fall risk assessment system for the elderly during exercise, comprising: The data acquisition module is used to collect raw signals from the inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor during the elderly's exercise process; The fitting module is used to map the preprocessed original signal onto the spatial coordinates of human body key points, and based on the spatial coordinates of human body key points, to perform ellipse fitting on the discrete point set corresponding to the motion envelope of the trunk and lower limbs using the least squares method to obtain a geometric representation of the motion posture; the geometric representation includes the ellipse center, major axis, minor axis and tilt angle parameters. The vector construction module is used to construct a fall behavior feature vector by integrating the spatiotemporal information of key points of the human body. The fall behavior feature vector includes changes in head height, changes in lower limb joint height, and geometric representation of movement posture. The changes in head height are obtained by the vertical distance between the center of the head key point and the ground, and the changes in lower limb joint height are obtained by the instantaneous changes or standard deviations of the height of the knees and ankles. The assessment module is used to score each indicator in the fall behavior feature vector according to the preset dynamic gait index scoring rules, and to obtain a risk score by combining the scores. The risk score is then compared with a preset threshold to output the corresponding risk level.

[0013] The present invention provides a computer device, including a memory and a processor. The memory stores a program, and when the program is executed by the processor, the processor performs the steps of the above-described method for assessing the risk of falls during exercise in the elderly.

[0014] The present invention provides a storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method for assessing the risk of falls during exercise in the elderly.

[0015] Compared with the prior art, the present invention has the following significant advantages: This invention maps inertia, pressure, and electromyography (EMG) signals to unified spatial coordinates of key points on the human body. Based on these spatial coordinates, it transforms dispersed physiological signals into spatial geometric features with clear physical meaning. This enables the quantification of subtle shifts in muscle coordination abnormalities and plantar pressure distribution implied in EMG signals into deformations of the trunk and lower limb motion envelope ellipse. By obtaining these deformations, this invention no longer relies solely on significant body tilt (i.e., significant kinematic changes) as an assessment criterion. Through comprehensive analysis of slight decreases in head height, minor disturbances in the movement trajectories of the knees and ankles, and geometric distortions of the motion envelope ellipse, the system can quantify and score the declining trend of neuromuscular control ability in elderly individuals before they actually lose balance. It can identify "soft" signs of imbalance before a fall earlier, thus truly realizing the transformation from "post-event / in-event alarm" to "pre-event risk scoring and early warning." Furthermore, risk level assessment is performed based on scoring rules, providing precise guidance for subsequent personalized interventions (such as targeted muscle training or balance training). Attached Figure Description

[0016] Figure 1 This is a flowchart of a method for assessing the risk of falls during exercise for the elderly, provided in an example of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0018] The system includes a wearable component and a data processing component.

[0019] Wearable components: Inertial Measurement Unit (IMU): Employing a six-axis sensor such as the MPU-6050, it is worn on the limbs and torso (e.g., waist) of the human body to acquire three-dimensional acceleration and three-dimensional angular velocity signals in real time, capturing macroscopic and microscopic motion postures and vibrations.

[0020] Foot pressure distribution sensor array: Employing high-resolution flexible thin-film sensors such as Tekscan FlexiForce A201, embedded in the insole, it is used to collect dynamic pressure distribution signals in multiple areas of the sole (such as 9 zones) to reflect the balance regulation mechanism.

[0021] Surface electromyography (sEMG): Using sensors such as NORAXON Ultium, it is attached to key muscle groups of the lower limb (such as the tibialis anterior and gastrocnemius muscles) to collect muscle electrical activity signals and directly reflect the neuromuscular control state.

[0022] Data processing section: All sensors achieve precise time synchronization for data acquisition and processing with the central processing unit (such as a mobile phone or embedded computer) via wireless or wired means.

[0023] This invention discloses a method for assessing the risk of falls during exercise in the elderly, such as... Figure 1 As shown, it includes the following steps: Step 1: Synchronously trigger and acquire raw signals from the inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor during the elderly person's movement.

[0024] Preprocessing of the original signal, such as filtering (e.g., low-pass, band-pass filtering (20–500Hz)) and noise reduction.

[0025] Micro-vibration feature enhancement and extraction based on multimodal fusion: This step is crucial for improving the signal-to-noise ratio. The original signal is preprocessed, and the preprocessed multimodal signal is treated as a whole, undergoing fusion analysis using a signal processing model (such as the Gaussian pulse signal modeling and frequency conversion reconstruction method mentioned in the background section of the manual). This process effectively suppresses noise in individual modes and obtains enhanced "motion micro-vibration characteristics" jointly characterized by electromyographic fremitus, instantaneous pressure fluctuations, and inertial micro-vibrations.

[0026] Step 2: Map the preprocessed raw signal onto the spatial coordinates of human body key points, and based on the spatial coordinates of human body key points, use the least squares method to fit the discrete point set corresponding to the motion envelope of the trunk and lower limbs to an ellipse to obtain a geometric representation of the motion posture; the geometric representation includes the ellipse center, major axis, minor axis and tilt angle parameters.

[0027] The least squares method is used to fit an ellipse to the discrete point set corresponding to the motion envelope of the trunk and lower limbs, obtaining the parameters of the ellipse center (x_c, y_c), major axis a, minor axis b, and tilt angle θ. These parameters dynamically characterize the stability and symmetry of the motion posture, and all parameters are used as a geometric representation of the motion posture. The specific expression for the ellipse fitting is as follows: Ax² + Bxy + Cy² + Dx + Ey + F = 0; Where A, B, C, D, E, and F are the correlation coefficients, and x and y represent the discrete points corresponding to the trunk and lower limb motion envelopes.

[0028] Step 3: Construct a fall behavior feature vector by integrating spatiotemporal information of key points on the human body. The fall behavior feature vector includes changes in head height, changes in lower limb joint height, and geometric representations of movement posture. Changes in head height are obtained by the vertical distance between the center of the head key point and the ground, and changes in lower limb joint height are obtained by the instantaneous changes or standard deviations in the height of the knees and ankles. These parameters together constitute a multidimensional geometric and kinematic comprehensive index set, namely the fall behavior feature vector, which is used to quantify human posture stability and map it to a risk scoring system.

[0029] Construct a feature vector within each time window: V = [H_head, H_limb, a, b, θ].

[0030] Among them, the head height change H_head: the head center is obtained based on the coordinates of the head key points (both eyes, both ears, nose), and then its vertical distance from the ground is calculated; the time series change rate is extracted.

[0031] Lower limb joint height change H_limb: Calculate the instantaneous change or standard deviation of the height of both knees and ankles; extract the instantaneous change amplitude.

[0032] Elliptical posture parameters (a, b, θ): reflect the dynamic changes in the range and direction of trunk and limb swing; a: amplitude of forward and backward swing; b: lateral stability; θ: direction of body tilt.

[0033] It also includes: extracting time-frequency domain features from the original signal and combining them with the spatiotemporal information of key points on the human body to form a fall behavior feature vector; such as the instantaneous frequency entropy of the signal calculated in step S2.

[0034] Step 4: Based on the preset dynamic gait index scoring rules, score each indicator in the fall behavior feature vector, and combine the scores to obtain a risk score. Compare the risk score with a preset threshold and output the corresponding risk level. The quantitative logic based on the dynamic gait index sets corresponding scoring rules for each indicator of the fall behavior feature vector, obtains the corresponding score, and outputs the risk level according to the preset threshold.

[0035] The constructed feature vector V is input into a risk assessment model. This model is constructed based on the quantization logic of the Dynamic Gait Index (DGI):

[0036] The assessment criteria for multiple tasks (such as walking on flat ground, turning around, and crossing obstacles) in the Dynamic Gait Index (DGI) scale are simulated, and corresponding scoring rules are set for each indicator in the fall behavior feature vector V.

[0037] For example, if the rate of descent of H_head exceeds the threshold and the elliptic parameter b (minor axis, reflecting lateral stability) fluctuates drastically, points will be deducted for that "test item".

[0038] The system automatically calculates the total score S (with a maximum score of m points simulating the DGI total score), sets n preset thresholds, and outputs the risk level based on the preset thresholds. In one embodiment, m=24 points and n=4, that is: S<19 points: high risk; 19≤S≤21 points: medium risk; 22≤S≤23 points: low risk; S=24 points: normal.

[0039] Example 1: Experimental Subjects and Testing Environment To verify the effectiveness of the method of the present invention, elderly subjects were selected to conduct a comparative experiment.

[0040] 1. Experimental subjects: Several elderly people aged 65 and above were selected as test subjects. They were able to walk independently and had no severe cognitive impairment or acute sports injury.

[0041] 2. Testing motion scenarios: Six movement modes were set up: walking forward with eyes open, walking with eyes open and turning head left and right, walking with eyes open and turning head up and down, walking forward with eyes closed, walking backward with eyes open, and walking backward with eyes closed. Each mode required one minute of adaptive walking before testing. The sensor configuration parameters are detailed in Table 1.

[0042] Table 1 Sensor Configuration Parameters When the sensor is collecting data, all signals are collected synchronously.

[0043] The experimental results and data support the following: 1. Instantaneous frequency entropy test results: The instantaneous frequency entropy of this invention is stably maintained at 0.10 ± 0.02. In contrast, the entropy fluctuation range of the comparative method is 0.30–0.70. This demonstrates that the motion control system of this method has higher stability.

[0044] 2. The results of the multi-joint coordination entropy are shown in Table 2.

[0045] Table 2 Results of five tests: The entropy value range of this invention is 0.110 – 0.145, which is significantly lower than that of the comparative method.

[0046] 3. ROC curve analysis: In six motion scenarios: the AUC of this invention is approximately 0.95–0.98, while the AUC of the comparison method is approximately 0.78–0.85, indicating that the recognition accuracy is higher.

[0047] Through multimodal collaborative perception: effectively improve the signal-to-noise ratio, suppress single-modal drift error, identify "soft imbalance" characteristics before a fall in advance, and output risk level stratification results. Original innovation, significantly improved signal-to-noise ratio: Through a specific combination and synchronous operation of inertial, pressure, and electromyographic sensors, the system achieves cross-sensing of the same physiological event (such as muscle contraction) from different dimensions of mechanics, electricity, and kinematics. This multi-source information complementarity fundamentally overcomes the problem of low single-signal signal-to-noise ratio caused by physiological degeneration in the elderly, and can capture micro-vibration characteristics that were previously submerged by noise.

[0048] High accuracy and good stability: As shown in the experimental data, the evaluation indicators output by this invention (such as instantaneous frequency entropy stabilizing at around 0.1, and multi-joint coordination entropy fluctuating within a narrow range of 0.11-0.145) exhibit excellent stability. The AUC values ​​of the ROC curves under different motion forms are closer to 1, demonstrating its high sensitivity and specificity, and ensuring accurate and reliable evaluation results.

[0049] It is highly systematic and easy to implement: It provides a complete technical solution from hardware selection, data fusion, feature extraction to risk classification, which can be easily integrated into a portable intelligent assessment device or system for real-time monitoring in communities, homes or rehabilitation centers.

[0050] This invention proposes a fall risk assessment system for the elderly during exercise, comprising: The system comprises several modules: a data acquisition module for collecting raw signals from an inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor during elderly individuals' movement; a fitting module for mapping the preprocessed raw signals onto the spatial coordinates of key human body points, and using the least squares method to fit an ellipse to the discrete point set corresponding to the trunk and lower limb movement envelopes based on these coordinates, obtaining a geometric representation of the movement posture; the geometric representation includes the ellipse center, major axis, minor axis, and tilt angle parameters; a vector construction module for constructing a fall behavior feature vector by integrating the spatiotemporal information of key human body points; the fall behavior feature vector includes changes in head height, changes in lower limb joint height, and a geometric representation of the movement posture; changes in head height are obtained through the vertical distance between the center of the head key point and the ground, and changes in lower limb joint height are obtained through the instantaneous changes or standard deviations of the heights of the knees and ankles; and an evaluation module for scoring each indicator in the fall behavior feature vector according to a preset dynamic gait index scoring rule, combining the scores to obtain a risk score, comparing the risk score with a preset threshold, and outputting the corresponding risk level.

[0051] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a program, and when the program is executed by the processor, the processor performs the steps of a method for assessing the risk of falls during exercise in the elderly.

[0052] According to the disclosed embodiments, the computer device can communicate with one or more external devices (e.g., keyboard, pointing device, Bluetooth communication, etc.) or with any device that enables the computing device to communicate with one or more other computing devices (e.g., router, demodulator, etc.).

[0053] The present invention also provides a storage medium storing a computer program thereon, characterized in that the computer program, when executed by a processor, implements the steps of a method for assessing the risk of falls during exercise in the elderly.

[0054] According to the disclosed embodiments, the storage medium can be a non-volatile computer-readable storage medium, such as, but not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, the storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0055] The above description, in conjunction with specific preferred embodiments, provides a more detailed explanation of the present invention. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such deductions or substitutions should be considered to fall within the scope of protection of the present invention.

Claims

1. A method for assessing the risk of falls during exercise in the elderly, characterized in that, include: Raw signals from an inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor were collected during the movement of elderly individuals. The preprocessed original signal is mapped onto the spatial coordinates of key points on the human body. Based on the spatial coordinates of the key points, the least squares method is used to fit an ellipse to the discrete point set corresponding to the motion envelope of the torso and lower limbs to obtain a geometric representation of the motion posture. The geometric representation includes the ellipse center, major axis, minor axis and tilt angle parameters. A fall behavior feature vector is constructed by integrating spatiotemporal information of key points on the human body; the fall behavior feature vector includes changes in head height, changes in lower limb joint height, and geometric representation of movement posture; wherein changes in head height are obtained by the vertical distance between the center of the head key point and the ground, and changes in lower limb joint height are obtained by the instantaneous changes or standard deviations in the height of the two knees and two ankles; According to the preset dynamic gait index scoring rules, each indicator in the fall behavior feature vector is scored, and the risk score is obtained by combining the scores. The risk score is compared with the preset threshold to output the corresponding risk level.

2. The method for assessing the risk of falls during exercise in the elderly as described in claim 1, characterized in that, The method of using least squares to fit an ellipse to the discrete point set corresponding to the motion envelope of the trunk and lower limbs to obtain a geometric representation of the motion posture is as follows: The least squares method is used to fit an ellipse to the discrete point set corresponding to the motion envelope of the trunk and lower limbs, obtaining the parameters of the ellipse center (x_c, y_c), major axis a, minor axis b, and tilt angle θ. All parameters are used as a geometric representation of the motion posture. The specific expression for the ellipse fitting is as follows: Ax² + Bxy + Cy² + Dx + Ey + F = 0; Where A, B, C, D, E, and F are the correlation coefficients, and x and y represent the discrete points corresponding to the trunk and lower limb motion envelopes.

3. The method for assessing the risk of falls during exercise in the elderly as described in claim 1, characterized in that, Before mapping the preprocessed raw signal onto the spatial coordinates of key points on the human body, the process also includes: The original signal is preprocessed, and the preprocessed multimodal signal is regarded as a whole. The signal processing model is used for fusion analysis to obtain the enhanced motion micro-vibration characteristics from electromyographic noise, instantaneous pressure fluctuations and inertial micro-vibrations.

4. The method for assessing the risk of falls during exercise in the elderly as described in claim 1, characterized in that, When constructing a fall behavior feature vector by integrating the spatiotemporal information of key human body points, the method further includes: extracting time-frequency domain features from the original signal and combining them with the spatiotemporal information of key human body points to form a fall behavior feature vector.

5. The method for assessing the risk of falls during exercise in the elderly as described in claim 1, characterized in that, The preset dynamic gait index scoring rule is as follows: The assessment criteria for multiple tasks in the Dynamic Gait Index (DGI) scale are simulated, and corresponding scoring rules are set for each indicator in the fall behavior feature vector; the multiple tasks include walking on flat ground, walking with head turned, and crossing obstacles.

6. The method for assessing the risk of falls during exercise in the elderly as described in claim 1, characterized in that, The step of comparing the risk score with a preset threshold and outputting the corresponding risk level is as follows: Set the total risk score to m points and set n preset thresholds. When the risk score falls within different preset threshold ranges, output the corresponding risk level.

7. A fall risk assessment system for the elderly during exercise, characterized in that, include: The data acquisition module is used to collect raw signals from the inertial measurement unit (IMU), plantar pressure sensor, and surface electromyography (sEMG) sensor during the elderly's exercise process; The fitting module is used to map the preprocessed original signal onto the spatial coordinates of human body key points, and based on the spatial coordinates of human body key points, to perform ellipse fitting on the discrete point set corresponding to the motion envelope of the trunk and lower limbs using the least squares method to obtain a geometric representation of the motion posture; the geometric representation includes the ellipse center, major axis, minor axis and tilt angle parameters. The vector construction module is used to construct a fall behavior feature vector by integrating the spatiotemporal information of key points of the human body. The fall behavior feature vector includes changes in head height, changes in lower limb joint height, and geometric representation of movement posture. The changes in head height are obtained by the vertical distance between the center of the head key point and the ground, and the changes in lower limb joint height are obtained by the instantaneous changes or standard deviations of the height of the knees and ankles. The assessment module is used to score each indicator in the fall behavior feature vector according to the preset dynamic gait index scoring rules, and to obtain a risk score by combining the scores. The risk score is then compared with a preset threshold to output the corresponding risk level.

8. A computer device, characterized in that, The device includes a memory and a processor, wherein the memory stores a program that, when executed by the processor, causes the processor to perform the steps of a method for assessing the risk of falls during exercise in older adults as described in any one of claims 1 to 6.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for assessing the risk of falls during exercise for the elderly as described in any one of claims 1 to 6.

Images