Intelligent fall prevention system based on biological characteristics of the elderly

Abstract

The application discloses an intelligent fall prevention system based on biological characteristics of the old, which comprises a wearable photoelectric plethysmography sensor, a human body sensing sensor, a gait sensor and an evaluation module. The system can judge the wake state of a user after the user prepares to get up. When the user is in the wake state, the user is allowed to walk. When the user is in the un-wake state, the user is prompted to postpone getting up. If the user starts to walk regardless of the prompt, the system further issues a warning in the form of sound and light and the like to wake up the user, so that the user can be prevented from falling.

Description

Technical Field

[0001] This invention relates to fall prevention and early warning technology for the elderly, and more particularly to an intelligent fall prevention system based on the biometrics of the elderly. Background Technology

[0002] Fall prevention in the elderly is an important health issue. Medical practice shows that falls in the elderly can cause soft tissue injuries, fractures, internal bleeding, and in severe cases, death. At the same time, falls often cause significant psychological stress for the elderly, such as fear of falling again, anxiety, and emotional depression. Post-injury mobility limitations often lead to social isolation, all of which greatly affect the quality of life of the elderly.

[0003] Current fall prevention measures using technology often include: age-friendly environmental modifications, such as adding handrails, increasing floor friction, and personal protective equipment such as fall airbags. While these technologies can reduce the risk of falls to some extent, they cannot accurately provide real-time behavioral guidance, hazard warnings, or automatic emergency assistance based on individual circumstances.

[0004] In addition, research has found that getting up at night is a high-risk scenario for accidental falls among the elderly. The main reason is that they have just woken up, are not fully awake, and have insufficient control over their bodies. Therefore, it is necessary to provide reliable fall prevention solutions for this special scenario.

[0005] Based on the above problems, the inventors have simultaneously utilized wearable sensors and intelligent environmental modifications to coordinate human biometrics with external fall prevention facilities via the Internet of Things, thereby achieving human-machine integrated fall risk warning and fall assistance services. Summary of the Invention

[0006] To overcome the aforementioned problems, the inventors conducted intensive research and designed an intelligent fall prevention system based on the biometrics of the elderly. This system includes a wearable photoplethysmography sensor, a human body sensor, a gait sensor, and an assessment module. It can determine the user's awakeness after the user prepares to get up. When the system determines that the user is awake, it allows the user to walk. When the system determines that the user is not awake, it prompts the user to postpone getting up. If the user ignores the reminder and begins to walk, it further issues a warning in the form of sound and light to wake the user up and prevent the user from falling, thus completing the invention.

[0007] Specifically, the purpose of this invention is to provide an intelligent fall prevention system based on the biometrics of the elderly, the system comprising:

[0008] Wearable photoplethysmography sensor 1, which is mounted on a wearable device worn by the user for a long time, is used to acquire the user's pulse interval sequence in real time.

[0009] Human body sensor 2, which is installed at the foot of the user's bed, is used to monitor whether the user is ready to get up and to send an assessment command when the user is ready to get up;

[0010] The assessment module 3 is used to retrieve the pulse interval sequence of the most recent time unit after receiving the assessment instruction, and to determine the user's autonomic nervous system arousal level. When the autonomic nervous system arousal level is lower than the set suitable walking value, an alarm state is entered.

[0011] The most recent 5 time units are calculated from the moment the user is detected preparing to get up, and then going back 5 time units.

[0012] The preferred time unit is 1 minute.

[0013] Among them, based on the pulse interval sequence data for each time unit, the autonomic arousal level of a user is obtained through the following formula (i):

[0014]

[0015] Wherein, AL represents the level of autonomic nervous system arousal;

[0016] This represents the average value of the sai sequence.

[0017] This represents the average value of the pai sequence.

[0018] A represents a constant.

[0019] In the evaluation module 3, after obtaining 5 autonomic arousal levels, each level is compared with the suitable walking value. When 3 or more autonomic arousal levels are greater than the suitable walking value, the user is considered to be awake and no additional operation is required. When 3 or more autonomic arousal levels are lower than the set suitable walking value, an alarm state is entered.

[0020] The suitable walking value is pre-stored in the evaluation module 3.

[0021] Preferably, the suitable walking value is variable, and the average value of similar users is collected as the initial value.

[0022] The method for updating the appropriate walk value is as follows:

[0023] The wearable photoplethysmography sensor 1 collects the pulse interval sequence of the user in a normal waking state in real time.

[0024] The data was divided into multiple groups based on time units, and the corresponding level of autonomic arousal was obtained for each group.

[0025] The obtained autonomic arousal levels were sorted from high to low, and the value ranked 95th was selected as the appropriate value for walking.

[0026] When the assessment module 3 determines that the level of autonomic nerve arousal is greater than or equal to the set suitable walking value, the indicator displays a "can get up" signal.

[0027] The "you can get up" signal includes a green light, a sound effect or voice indicating that you can walk.

[0028] In the alarm state, an indicator displays a "postpone waking up" signal.

[0029] In the alarm state, when the indicator detects that the user has started walking, it displays a "stop walking" signal;

[0030] The "stop walking" signal includes vibration, red light, sound effects or voice indicating that you should stop walking.

[0031] The system also includes gait sensors 4, which are installed on both of the user's shoes;

[0032] The gait sensor 4 is used to detect and record the user's foot movement information, including the foot's spatial movement trajectory and foot rotation angle.

[0033] The beneficial effects of this invention include:

[0034] (1) The intelligent fall prevention system based on the biometrics of the elderly provided by the present invention can assess whether the mental state of the user meets the safe walking standards when getting up, and has a fall warning function, and finally provides key prevention for high-risk fall scenarios such as getting up at night and getting up in bed;

[0035] (2) The intelligent fall prevention system based on the biometrics of the elderly provided by the present invention uses user standing detection, real-time detection of alertness level and guidance for standing and waiting before walking to enable users to complete the mental state preparation before walking through a set of guidance processes, thereby basically eliminating the risk of falls in high-risk scenarios such as getting up at night and getting out of bed. Attached Figure Description

[0036] Figure 1 The diagram shows the overall logic of the intelligent fall prevention system based on the biometrics of the elderly, as described in this application.

[0037] Explanation of reference numerals in the attached figures

[0038] 1-Wearable photoplethysmography sensor

[0039] 2-Human body sensor

[0040] 3-Evaluation Module

[0041] 4-Gait Sensor Detailed Implementation

[0042] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Through these descriptions, the features and advantages of the present invention will become clearer and more apparent.

[0043] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0044] This invention provides an intelligent fall prevention system based on the biometrics of the elderly. The system includes a wearable photoplethysmography sensor 1, a human body sensing sensor 2, an assessment module 3, and a gait sensor 4.

[0045] The wearable photoplethysmography sensor 1 is mounted on a wearable device worn by the user for extended periods of time, and is used to acquire the user's pulse interval sequence in real time; the wearable device is a smart bracelet or watch; the signal generated by the wearable photoplethysmography sensor is filtered and peak extracted by the bracelet or watch to obtain the pulse interval sequence.

[0046] The human body sensor 2 is installed at the foot of the user's bed to monitor whether the user is preparing to get up and sends an assessment command when the user is ready to get up. The human body sensor 2 can be a gravity sensor, which detects changes in weight at the foot of the bed to determine whether the user's feet are on the ground. When the user's feet are on the ground, it can be considered that the user is ready to get up. The human body sensor 2 can also be a lidar sensor or an infrared sensor, capable of real-time detection of whether the user's legs / feet have fallen into a detection height area. The detection height area can be set according to actual conditions, for example, an area below 30cm from the ground. When the user's legs / feet fall into the detection height area, it can be considered that the user is ready to get up. In this application, due to the different working principles of the human body sensor 2, the criteria for determining whether a user is ready to get up are not entirely consistent, but all can determine that the user is transitioning from a sleeping state to a walking activity state, requiring targeted analysis and early warning.

[0047] The assessment module 3 is used to retrieve the pulse interval sequence of the most recent 5 time units after receiving the assessment instruction, and to determine the user's autonomic nervous system arousal level based on this. When the autonomic nervous system arousal level is lower than the set suitable walking value, an alarm state is entered.

[0048] Preferably, the most recent 5 time units are counted backwards from when the user is detected preparing to get up, which is 5 time units.

[0049] The preferred time unit is 1 minute. Based on the pulse interval sequence data of each time unit, the autonomic nervous system arousal level of a user is obtained.

[0050] Preferably, the user's autonomic arousal level is obtained by the following formula (a):

[0051]

[0052] Wherein, AL represents the level of autonomic nervous system arousal; in this application, the level of autonomic nervous system arousal can also be referred to as the degree of wakefulness;

[0053] The SAI sequence represents the average value of the SAI sequence. The pulse interval sequence includes the SAI sequence, meaning that the SAI sequence is a portion of the data in the pulse interval sequence. The specific meaning of the SAI sequence is the sympathetic nerve index sequence obtained by inputting the intracardiac cycle sequence within the time window into the Laguerre regression model, which is used to characterize the strength of sympathetic nerve activity.

[0054] The PAI sequence represents the average value of the PAI sequence. The PAI sequence is included in the interpulse sequence, meaning that the PAI sequence is a portion of the data in the interpulse sequence. The PAI sequence specifically means the parasympathetic index sequence obtained by inputting the intracardiac cycle sequence within the time window into the Laguerre regression model, which is used to characterize the strength of parasympathetic activity. The Laguerre regression model is the model from Gaetano V, Luca C, Philip SJ, et al. Measures of Sympathetic and Parasympathetic Autonomic Outflow from Heartbeat Dynamics.[J]. Journal of Applied Physiology, 2018, 125(1):19-39.

[0055] A represents a constant, and its preferred value is 10000.

[0056] In assessment module 3, after obtaining 5 autonomic arousal levels, each level is compared with the suitable walking value. When 3 or more autonomic arousal levels are greater than the suitable walking value, the user is considered to be awake and no additional operation is required. When 3 or more autonomic arousal levels are lower than the set suitable walking value, an alarm state is entered.

[0057] Preferably, the suitable walking value is pre-stored in the evaluation module 3.

[0058] Preferably, the suitable walking value is variable, and is initially determined by collecting the average value of a similar group of users. The similar group of users refers to people of the same age and gender as the user, whose weight differs from the user's by no more than 2 kg.

[0059] Preferably, the method for updating the appropriate walk value is as follows:

[0060] The wearable photoplethysmography sensor 1 collects the pulse interval sequence of the user in a normal waking state in real time, and the waking state is determined by a gait sensor and the current time.

[0061] The data is divided into multiple groups according to time units. The corresponding level of autonomic nervous system arousal is obtained based on each group of data. That is, each minute corresponds to a group of data. After the user wears the intelligent fall prevention system, a large amount of data can be generated within a period of time, and the suitable walking value can be updated once an hour.

[0062] At least 1000 sets of data were collected from individuals, sorted from smallest to largest (daytime activity data, not sleep data). The obtained autonomic arousal levels were then sorted from highest to lowest and normalized (maximum and minimum values ​​were standardized), with the 20th percentile for the maximum and the 80th percentile for the minimum. The value at the 75th percentile was selected as the suitable walking value. This method selects a suitable walking value that is essentially at the minimum level for the user to remain awake, ensuring that the user is alert and less likely to fall, while also minimizing the user's waiting time. In most cases, the user reaches this level of alertness when preparing to get out of bed, allowing for subsequent actions such as getting out of bed and walking.

[0063] Preferably, when the evaluation module 3 determines that the level of autonomic nerve arousal is greater than or equal to the set suitable walking value, the indicator displays a "can get up" signal;

[0064] The "you can get up" signal includes a green light, a sound effect or voice indicating that you can walk.

[0065] Preferably, in the alarm state, an "delay waking up" signal is presented via an indicator.

[0066] In the alarm state, when the indicator detects that the user has started walking, it displays a "stop walking" signal;

[0067] The "stop walking" signal includes vibration, red light, sound effects or voice indicating that you should stop walking.

[0068] The gait sensor 4 is installed on both of the user's shoes. It detects and records the user's foot movement information, including the spatial trajectory of the foot and the foot rotation angle. The detected data can be converted into a series of indicators in real time, such as cadence, stride length, landing posture, ground contact time, foot eversion angle, swing angle, impact load, vertical stride ratio, etc. Based on the detected data, it can determine whether the user is walking, and thus, in conjunction with the evaluation module, provide alerts to users who are walking while not fully conscious. It can also assess the user's fall risk based on the detected data.

[0069] Furthermore, the system can also include service nodes to receive and store data detected by the gait sensor 4 in real time. These service nodes accumulate user walking indicator data, and when the data volume exceeds a certain amount, such as indoor walking data for two calendar days, the service nodes establish distribution information for each individual's walking indicator. The service nodes can then use this distribution information to establish fall risk warning standards. Specifically, during subsequent monitoring, a moderate risk warning is triggered when a user's indicator exceeds a certain limit in their data distribution. This limit can be 1.5 standard deviations centered on the mean. The user's own indicator can also be compared with pre-stored standardized fall-prone pattern status data for the general population. When the data matches, the service node will trigger a high fall risk warning.

[0070] Furthermore, the system can also incorporate information repeaters, which handle communication between sensors, indicators, and service nodes, replacing direct communication between them. This enhances networking flexibility, reduces costs, and improves the system's environmental adaptability. Repeaters acquire sensor data wirelessly and via wired connections. For example, information repeaters connect to sensors and indicators using short-range wireless transmission protocols, such as via Wi-Fi, Bluetooth, or satellite connections. Repeaters communicate with service nodes via a wide area network. In wireless connections, spatial limitations such as distance and wall obstruction exist. When the required transmission range exceeds the effective range of a single repeater, multiple repeaters can be bridged to extend the transmission space. For instance, in a multi-room environment, repeaters can be deployed in each room and corridor and bridged together.

[0071] Example

[0072] Sixty-eight elderly residents with normal cognitive function and who are able to get out of bed and walk independently were selected from a nursing home. Among the 68 residents, there were 32 men and 36 women, all of whom were over 65 years old.

[0073] Each elderly person is provided with a set of the intelligent fall prevention system based on the biometrics of the elderly as described in this application. After teaching the elderly person how to use it, each elderly person wears the intelligent fall prevention system based on the biometrics of the elderly for an extended period of time.

[0074] This intelligent fall prevention system, such as Figure 1 As shown, it includes a wearable photoplethysmography sensor 1, a human body sensing sensor 2, an evaluation module 3, and a gait sensor 4;

[0075] The wearable photoplethysmography sensor acquires the user's pulse interval sequence in real time.

[0076] The system uses human body sensors to monitor whether the user is ready to get up and sends an assessment command when the user is ready to get up.

[0077] After receiving an assessment instruction, the assessment module retrieves the pulse interval sequence of the most recent five time units to determine the user's autonomic nervous system arousal level. When the autonomic nervous system arousal level is lower than the set suitable walking value, an alarm state is entered.

[0078] During the alarm state, an "Wait to wake up" signal is displayed via an indicator.

[0079] In the alarm state, when the gait sensor 4 detects that the user has started walking, the indicator displays a "stop walking" signal; the "stop walking" signal includes vibration, red light, sound effect or voice indicating that walking should be stopped.

[0080] The most recent 5 time units are counted backward from when the user is found to be getting up, and the time unit is 1 minute. Based on the pulse interval sequence data of each time unit, the autonomic nervous arousal level of a user is obtained.

[0081] The user's autonomic arousal level is obtained by the following formula (a):

[0082]

[0083] Wherein, AL represents the level of autonomic nervous system arousal;

[0084] Represents the average value of the sai sequence;

[0085] This represents the average value of the pai sequence;

[0086] A represents a constant with a value of 10000.

[0087] After obtaining 5 autonomic arousal levels, each is compared with the suitable walking value. When 3 or more autonomic arousal levels are greater than the suitable walking value, the user is considered to be awake and no additional operation is required. When 3 or more autonomic arousal levels are lower than the set suitable walking value, an alarm state is entered.

[0088] The method for updating the appropriate step value is as follows:

[0089] The wearable photoplethysmography sensor 1 collects the pulse interval sequence of the user in a normal waking state in real time, and the waking state is determined by a gait sensor and the current time.

[0090] The data is divided into multiple groups according to time units, and the corresponding autonomic nervous system arousal level is obtained based on each group of data, that is, one group of data per minute;

[0091] The obtained autonomic arousal levels were sorted from high to low, and the value ranked 75th was selected as the appropriate value for walking.

[0092] The intelligent fall prevention system and the elderly person's living conditions will be continuously monitored over a period of 3 months.

[0093] During the testing period, the alarm was activated 122 times, and a stop-walking signal was issued 18 times. After the stop-walking signal was issued, the elderly person stopped walking and sat back on the bed. Throughout the entire testing period, no elderly people wearing the smart fall prevention system based on their biometrics experienced a fall at night.

[0094] According to the nursing home's records, before the introduction of the intelligent fall prevention system based on the biometrics of the elderly proposed in this application, an average of at least two elderly people fell at night every month.

[0095] Therefore, the intelligent fall prevention system based on the biometrics of the elderly provided in this application can effectively prevent nighttime falls among the elderly and has extremely high practical value.

[0096] The present invention has been described above with reference to preferred embodiments; however, these embodiments are merely exemplary and illustrative. Various substitutions and modifications can be made to the present invention based on these embodiments, all of which fall within the scope of protection of the present invention.

Claims

1. An intelligent fall prevention system based on the biometrics of the elderly, characterized in that, The system includes: Wearable photoplethysmography sensor (1), which is mounted on a wearable device worn by the user for a long time, is used to acquire the user's pulse interval sequence in real time; Human body sensor (2), which is installed at the foot of the user's bed, is used to monitor whether the user is ready to get up and to send an assessment command when the user is ready to get up; The assessment module (3) is used to retrieve the pulse interval sequence of the last 5 time units after receiving the assessment instruction, and to determine the user's autonomic nervous system arousal level. When the autonomic nervous system arousal level is lower than the set suitable walking value, an alarm state is entered.

2. The intelligent fall prevention system based on the biometrics of the elderly according to claim 1, characterized in that, The most recent 5 time units are counted from when the user was found to be getting up, going back 5 time units.

3. The intelligent fall prevention system based on the biometrics of the elderly according to claim 2, characterized in that, The time unit is 1 minute.

4. The intelligent fall prevention system based on the biometrics of the elderly according to claim 1, characterized in that, Based on the pulse interval sequence data for each time unit, the autonomic arousal level of a user is obtained using the following formula (i): (one) in, Indicates the level of autonomic nervous system arousal; This represents the average value of the sai sequence. This represents the average value of the pai sequence. Represents a constant.

5. The intelligent fall prevention system based on the biometrics of the elderly according to claim 4, characterized in that, In the evaluation module (3), after obtaining 5 autonomic arousal levels, they are compared with the suitable walking value one by one. When 3 or more autonomic arousal levels are greater than the suitable walking value, it is considered that the user is awake and no additional operation is required. When 3 or more autonomic arousal levels are lower than the set suitable walking value, an alarm state is entered.

6. The intelligent fall prevention system based on the biometrics of the elderly according to claim 1, characterized in that, The suitable walking value is pre-stored in the evaluation module (3). The suitable walking value is variable and is initially obtained by collecting the average value of similar users.

7. The intelligent fall prevention system based on the biometrics of the elderly according to claim 6, characterized in that, The method for updating the appropriate step value is as follows: The pulse interval sequence of the user in a normal waking state is collected in real time by a wearable photoplethysmography sensor (1). The data was divided into multiple groups based on time units, and the corresponding level of autonomic arousal was obtained for each group. The obtained autonomic arousal levels were sorted from high to low, and the value ranked 95th was selected as the appropriate walking value.

8. The intelligent fall prevention system based on the biometrics of the elderly according to claim 1, characterized in that, When the assessment module (3) determines that the level of autonomic nervous arousal is greater than or equal to the set suitable walking value, the indicator displays a "can get up" signal; The "you can get up" signal includes a green light and a sound effect or voice indicating that you can walk.

9. The intelligent fall prevention system based on the biometrics of the elderly according to claim 1, characterized in that, During the alarm state, an "undelay" signal is displayed via an indicator. In the alarm state, when the indicator detects that the user has started walking, it displays a "stop walking" signal; The "stop walking" signal includes vibration, red light, sound effects or voice indicating that you should stop walking.

10. The intelligent fall prevention system based on the biometrics of the elderly according to claim 9, characterized in that, The system also includes gait sensors (4), which are mounted on the user's two shoes; The gait sensor (4) is used to detect and record the user's foot movement information, including the foot spatial movement trajectory and foot rotation angle.

Images