Fall prediction wearable device based on transversus abdominis muscle electrical signals

Abstract

The utility model provides a kind of wearable device based on abdominal transverse muscle myoelectric signal, it is related to wearable health monitoring equipment technical field, solve the current fall monitoring technology existing's hysteresis high and misjudgment rate higher Limitation problem.The device is sequentially provided with upper shell (1), bottom shell (2) and electrode piece (4) from top to bottom, and accommodating cavity is formed in bottom shell (2), and microprocessor (5) and motion pose sensor (7) are arranged in accommodating cavity;Microprocessor (5) is electrically connected with electrode piece (4), and microprocessor (5) is electrically connected with motion pose sensor (7);The lower surface of electrode piece (4) is attached to the abdominal transverse muscle of user, for collecting the abdominal transverse muscle myoelectric signal of user;Motion pose sensor (7) is used to collect the physical motion characteristics of user.The utility model integrates the comprehensive monitoring of abdominal transverse muscle myoelectric and wearer's own motion characteristics, and can realize the early prediction and accurate early warning of fall risk.

Description

Technical Field

[0001] This utility model relates to the field of wearable health monitoring equipment technology, and is applied to the real-time health monitoring of the elderly, sports rehabilitation patients and high-risk workers. Specifically, it relates to a wearable device for fall prediction based on the electromyographic signal of the transverse abdominis muscle. Background Technology

[0002] With the accelerating aging of society, the health management of the elderly is becoming increasingly prominent. Besides underlying chronic diseases, the decline in motor balance due to physiological function deterioration has become a key factor threatening their quality of life. Clinical data shows that accidental falls, a major cause of death and disability among the elderly, often stem from the progressive decline in gait stability and postural control. Fractures and subsequent bedridden complications caused by falls can create a vicious cycle, significantly increasing the medical burden and endangering lives. Therefore, establishing an effective fall risk early warning mechanism will have significant social value.

[0003] Current fall detection technologies primarily rely on single-modal sensing solutions such as accelerometers, gyroscopes, or barometric pressure sensors. These solutions detect events by identifying the kinematic characteristics of fall movements. While they have some application value in the alarm process after a fall occurs, they suffer from significant technical limitations: passive response mechanisms struggle to effectively intervene before a fall occurs, failing to meet preventative medical needs; and due to the limited dimensions of sensor data, everyday activities such as squatting, bending over, etc., are easily misinterpreted as falls, resulting in insufficient system specificity. Furthermore, existing technologies generally neglect the core physiological mechanisms of human motor control, failing to incorporate muscle function assessment into the risk prediction system.

[0004] Medical research has confirmed a close relationship between the stability of the core muscles, particularly the transverse abdominis, and the body's dynamic balance. As a core muscle group maintaining postural stability, the transverse abdominis not only participates in respiratory regulation but also directly affects the biomechanical balance of the spine and pelvis through intra-abdominal pressure regulation. When this core muscle group experiences muscle weakness or abnormal motor coordination, it leads to a decline in the ability to adjust the center of gravity, significantly increasing the risk of falls. However, currently, there is no technology that integrates core muscle function assessment with motion sensor data, which has become a key bottleneck restricting the development of fall warning technology. Therefore, developing intelligent monitoring devices that can simultaneously monitor muscle function and motor characteristics has become a key research direction for improving the effectiveness of health management for groups such as the elderly. Utility Model Content

[0005] The purpose of this invention is to introduce the acquisition of transverse abdominis muscle electromyography (EMG) signals to address the limitations of current fall detection technologies, such as high lag and high false alarm rates. Therefore, a wearable fall prediction device based on transverse abdominis muscle EMG signals is proposed. This device integrates comprehensive monitoring of transverse abdominis muscle EMG and the wearer's own movement characteristics, enabling early prediction and accurate warning of fall risks.

[0006] The present invention employs the following technical solution to achieve its objective: A wearable device for fall prediction based on electromyographic signals of the transverse abdominis muscle is disclosed. The device comprises, from top to bottom, an upper shell, a lower shell, and electrode pads. A receiving cavity is formed within the lower shell, and a microprocessor and a motion posture sensor are disposed within the receiving cavity. The microprocessor is electrically connected to the electrode pads and the motion posture sensor. The lower surface of the electrode pads is attached to the user's transverse abdominis muscle to collect the user's electromyographic signals of the transverse abdominis muscle. The motion posture sensor is used to collect the user's physical movement characteristics.

[0007] Specifically, an AD acquisition module is also installed inside the cavity, and the microprocessor is electrically connected to the electrode pads through the AD acquisition module. The microprocessor and the AD acquisition module are connected by a four-wire connection using the SPI protocol, and the AD acquisition module and the electrode pads are connected by positive and negative electrode lines. The AD acquisition module is used to convert the electromyographic signals of the user's transverse abdominis muscle acquired by the electrode pads into digital signals and then transmit them to the microprocessor.

[0008] Furthermore, the motion posture sensor has an inertial measurement unit, which consists of a three-axis accelerometer and a gyroscope. The three-axis accelerometer is used to acquire the static tilt angle and acceleration value of the user's posture at the device wearing point, and the gyroscope is used to acquire the instantaneous attitude angle of the user's posture at the device wearing point. The user's physical motion characteristics are composed of the data acquired by the inertial measurement unit. The microprocessor and the motion posture sensor are connected by a four-wire connection using the SPI protocol.

[0009] Furthermore, a wireless communication module is also installed inside the cavity. The microprocessor and the wireless communication module are connected via a four-wire UART serial port protocol. The wireless communication module is used for wireless communication between its own device and other fall prediction wearable devices worn by the same user, as well as for wireless communication between its own device and the user's personal terminal device.

[0010] Specifically, a power management module is also installed inside the cavity. The microprocessor and the power management module are connected by positive and negative power lines. The power management module has a built-in power management chip and battery. The power management module is used to power all the electrical components in the fall prediction wearable device.

[0011] Specifically, a standard interface is provided on the bottom shell corresponding to the location of the power management module; the standard interface connects the microprocessor and the power management module respectively, and is used to provide battery charging input and external data input.

[0012] Preferably, both the upper and lower shells are made of flexible TPU material; a highly breathable lightweight textile fabric is also provided between the lower shell and the electrode sheet; the highly breathable lightweight textile fabric has a preset thickness value to provide breathable isolation between the lower shell and the electrode sheet.

[0013] Specifically, the upper shell is also equipped with function buttons, which are electrically connected to the microprocessor. The function buttons are used to feed back operation instructions from the user to the microprocessor.

[0014] Preferably, a buzzer is also provided inside the cavity, and the microprocessor is electrically connected to the buzzer; a speaker hole is provided on the upper shell at the location corresponding to the buzzer.

[0015] Preferably, multiple fall prediction wearable devices are worn on the transverse abdominis muscle of the same user, and the various fall prediction wearable devices are connected to each other via wireless communication.

[0016] In summary, due to the adoption of this technical solution, the beneficial effects of this utility model are as follows: This invention introduces electromyography (EMG) signal monitoring of the transverse abdominis muscle and motion sensing monitoring. The integration of these two technologies significantly improves the real-time performance and accuracy of fall risk warnings. Compared to traditional single-modal sensing devices, this invention can effectively capture abnormal states of core muscle groups minutes to hours before a fall, while maintaining long-term monitoring stability, thus overcoming the limitations of existing technologies in terms of warning timeliness.

[0017] At the hardware architecture level, this invention enables the simultaneous acquisition of electromyographic signals of the transverse abdominis muscle and human motion parameters. This dual-channel physical acquisition mechanism not only enhances the ability to identify pre-fall warning signs but also effectively eliminates interference signals caused by everyday movements such as squatting and bending over through hardware-level fusion of multi-source sensor data. In practical applications, the device significantly reduces the false trigger rate for non-fall-related movements, improving the reliability of device operation and user compliance.

[0018] The flexible electrode pads' adhesive design allows this invention to balance monitoring performance and wearability comfort, while the low-power circuitry enables extended wear. Through innovative combinations at the physical sensing level, this invention provides the elderly and other groups with a fall risk monitoring tool that combines early warning capabilities with practical value, demonstrating its advantages in improving the quality of elderly care and reducing the burden on healthcare. Attached Figure Description

[0019] Figure 1 This is an exploded view of the composition and structure of the fall prediction wearable device of this utility model; Figure 2 This is a schematic diagram of the overall shape of the fall prediction wearable device of this utility model; Figure 3 This is an example diagram showing the present invention worn by a user; Figure 4 This is a schematic diagram of the working communication interaction logic of this utility model.

[0020] The meanings of the markings in the attached diagram are as follows: 1-Upper shell, 2-Bottom shell, 3-Highly breathable lightweight textile fabric, 4-Electrode sheet, 5-Microprocessor, 6-AD acquisition module, 7-Motion posture sensor, 8-Wireless communication module, 9-Power supply management module, 10-Standard interface, 11-Function button. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this utility model clearer, the technical solutions of the embodiments of this utility model will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this utility model, not all embodiments. The components of the embodiments of this utility model described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0022] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0023] Example 1 like Figure 1 As shown, a wearable device for fall prediction based on transverse abdominis muscle electromyography (EMG) signals is provided. The device consists of an upper shell 1, a bottom shell 2, and an electrode plate 4 arranged sequentially from top to bottom. A receiving cavity is formed inside the bottom shell 2, and a microprocessor 5 and a motion posture sensor 7 are disposed inside the receiving cavity. The microprocessor 5 is electrically connected to the electrode plate 4 and the motion posture sensor 7. The lower surface of the electrode plate 4 is attached to the user's transverse abdominis muscle to collect the user's transverse abdominis muscle EMG signals. The motion posture sensor 7 is used to collect the user's physical movement characteristics.

[0024] The overall shape of the device can be seen in the image. Figure 2This is a schematic diagram. Based on the structural characteristics of the device, mature predictive response algorithm logic can be configured to realize a fall prediction response mechanism based on the electromyographic signal of the transverse abdominis muscle and physical motion characteristics. Therefore, the device in this embodiment can be preset with different response logics, which can be deployed according to functional requirements in actual applications. Detailed limitations are not provided here; only some preferred response methods are described in other embodiments. This embodiment describes the details of the key components in the device.

[0025] like Figure 1 As shown, an AD acquisition module 6 is also provided in the receiving cavity of the bottom shell 2. The AD acquisition module 6 is selected as a low-power ADC such as AD8232 or AD8233. The microprocessor 5 is electrically connected to the electrode plate 4 through the AD acquisition module 6. The microprocessor 5 is selected as STM32H7 or GD32H7, which are all 32-bit microprocessor units. Such microprocessor units have high performance characteristics and can be competent for both direct operation of conventional logic algorithms and deployment of small and medium-sized convolutional neural networks. Therefore, they are suitable for the application scenarios of the device in this embodiment.

[0026] In this embodiment, the microprocessor 5 and the AD acquisition module 6 are connected by a four-wire SPI protocol, namely VCC, SDIO, SCLK and GND, the principle of which will not be described in detail. The AD acquisition module 6 and the electrode 4 are connected by positive and negative electrode lines S+ and S-. The AD acquisition module 6 is used to convert the electromyographic signal of the user's transverse abdominis muscle acquired by the electrode 4 into a digital signal that can be recognized by the microprocessor 5, and then transmit it to the microprocessor 5 through the connection line.

[0027] In this embodiment, the motion posture sensor 7 has an inertial measurement unit (IMU), which consists of a triaxial accelerometer and a gyroscope. The triaxial accelerometer acquires the static tilt angle and acceleration value of the user's posture at the device's wearing point. By detecting the component of gravitational acceleration in three-dimensional space, the static tilt angle of the device relative to the Earth's coordinate system can be determined. The combined application and interactive communication of multiple devices will make the detection of the static tilt angle more accurate. The gyroscope acquires the instantaneous attitude angle of the user's posture at the device's wearing point. The user's physical motion characteristics are composed of data acquired by the IMU. Similar to the connection method of the AD acquisition module 6, the microprocessor 5 and the motion posture sensor 7 are connected via a four-wire SPI protocol, with VCC, SDIO, SCLK, and GND as the wires. Based on the relevant data of physical motion characteristics, the device can determine the user's posture and center of gravity, serving as part of the basis for the fall prediction response mechanism.

[0028] like Figure 1As shown, a wireless communication module 8 is also installed inside the cavity of the bottom shell 2. The microprocessor 5 and the wireless communication module 8 are connected via a four-wire UART serial port protocol, with VCC, TX, RX, and GND as the wires. The principle is not described in detail here. The wireless communication module 8 is used for wireless communication between its own device and other fall prediction wearable devices worn by the same user, as well as for wireless communication between its own device and the user's personal terminal device. The specific wireless communication technology can adopt Internet of Things (IoT) technologies such as Bluetooth or NB-IoT, thereby realizing interconnection and communication between multiple devices, as well as the interaction of fall prediction response mechanism-related information to the user's mobile phone, tablet, and other terminal devices.

[0029] like Figure 1 As shown, a power management module 9 is also housed within the receiving cavity of the bottom shell 2. The microprocessor 5 is connected to the power management module 9 via positive and negative power lines. The power management module 9 contains a power management chip and a battery, and is used to power all electrical components in the fall prediction wearable device. The device can use a high-density lithium battery, and the power management chip is selected as TP4056, which can protect the lithium battery from overcharging or over-discharging, and can charge the lithium battery to the target voltage. During charging, it can provide variable current to extend the lithium battery's lifespan.

[0030] The microprocessor 5, acting as the management and interaction hub for all components in the device, processes various digital signals from these components. These signals are then used to determine the transverse abdominis muscle electromyography (EMG) and posture of the user, and subsequently trigger various alarms through a fall prediction response mechanism. The device is used simply by ensuring that the lower surface of the electrode pads 4 are attached to the user's transverse abdominis muscle. Specific wearing accessories or methods can be selected based on existing mature technologies, such as fixing the device at a preset position on a waist belt, or using a direct adhesive method similar to a medicated patch.

[0031] In this embodiment, the preferred configuration of the device is a combination of multiple devices arranged on the user's transverse abdominis muscle, as shown in the reference. Figure 3 The image shows the device after being worn. Two of these devices are placed on each of the left and right sides. Since the various fall prediction wearable devices are connected via wireless communication, the optimal balance between wearability and data collection effect can be achieved, thereby obtaining accurate early warning and alarm effects.

[0032] Example 2 Based on Example 1, this example provides an optimized description of the auxiliary structural features of the fall prediction wearable device, aiming to improve wearing comfort and ease of use.

[0033] like Figure 1 As shown, a standard interface 10 is provided on the bottom shell 2 at the location corresponding to the power management module 9. Figure 2 This interface is not shown; please refer to it only. Figure 1 That's it. The standard interface 10 can directly use the mature Type-C interface in this field, which is compatible with charging and data interaction functions; therefore, the standard interface 10 is connected to the microprocessor 5 and the power management module 9 respectively, and the standard interface 10 is used to provide battery charging input and external data input.

[0034] like Figure 1 , Figure 2 As shown, both the upper shell 1 and the bottom shell 2 are made of flexible TPU material, which maximizes the wearer's comfort. The upper shell 1 and the bottom shell 2 are connected by multiple self-tapping screws. A highly breathable lightweight textile fabric 3 is also provided between the bottom shell 2 and the electrode plate 4. This fabric has a preset thickness to provide breathable insulation between the bottom shell 2 and the electrode plate 4, preventing the TPU shell from directly and completely pressing against the transverse abdominis muscle and causing discomfort.

[0035] like Figure 1 , Figure 2 As shown, the upper shell 1 is also equipped with function buttons 11. The microprocessor 5 is electrically connected to the function buttons 11, which are used to feed back user operation instructions to the microprocessor 5. The function buttons 11 can realize basic functions such as powering on / off and mode switching, making it convenient for users to operate the device directly and simply. Due to the device's wireless communication function, users can also operate the device's related functions through mobile phone apps. Feedback from the device's functions to the user can be achieved by setting corresponding indicator lights or small displays (not shown in the figure) on the upper shell 1. The principles and technologies are simple and mature, and will not be elaborated here.

[0036] Finally, as a preferred embodiment and also a local emergency warning / alarm method under conditions of poor communication such as no network, this embodiment also includes a buzzer (not shown in the figure) installed in the receiving cavity of the bottom shell 2, and the microprocessor 5 is electrically connected to the buzzer; a speaker hole is provided on the upper shell 1 corresponding to the location of the buzzer. When the microprocessor 5 determines that a warning / alarm is required based on the fall prediction response mechanism, it can directly control the buzzer 5 to play a help-seeking audio, thereby realizing a local warning / alarm reminder.

[0037] Example 3 Based on any of the above embodiments, this embodiment provides a preferred description of the working principle of the fall prediction wearable device, specifically by employing a superior fall prediction response mechanism to maximize the hardware functionality of the device. In practical applications, other logically configured response mechanisms can also be used to provide fall warnings / alarms with specific target needs, thus catering to different user groups.

[0038] In summary, the working principle of this device and its achievable fall prediction capabilities can be found in [link to relevant documentation]. Figure 4 The flowchart illustrates the process. Based on a comprehensive assessment of the transverse abdominis muscle electromyography (EAG) signals and physical movement characteristics, if an external warning / alarm is required, the device's local buzzer can play a distress audio message. After the device communicates with the user's mobile phone, the user's phone can also simultaneously play the distress audio message. Based on the development of related apps and other functions on the user's mobile phone, after triggering a warning / alarm, the device can send WeChat notifications, make phone calls for help, and send location information to the user's family members' phones. It can also make phone calls for help and send location information to medical institutions providing related services. Simultaneously, the user's mobile phone can establish communication with the cloud service backend of the relevant system via the MQTT protocol. When direct phone calls are difficult, the cloud service backend can push warning / alarm messages to family members' phones and medical institutions.

[0039] Next, this embodiment introduces a superior fall prediction response mechanism applied to this device. This response mechanism is based on the microprocessor 5 using high-performance chip units such as STM32H7 and GD32H7, and implements the fall prediction process by deploying a lightweight convolutional neural network model.

[0040] Flexible, high-precision electrode pads from several devices are attached to the surface of the transverse abdominis muscle to achieve continuous acquisition of the transverse abdominis muscle electromyography (EMG) signal. After acquiring the user's transverse abdominis muscle EMG signal and digitizing it through the low-power AD acquisition module 6, the EMG signal feature values ​​of the transverse abdominis muscle are extracted by wavelet transform and input into a pre-deployed convolutional neural network model, which then outputs the transverse abdominis muscle stability score parameter. When the transverse abdominis muscle stability score parameter is less than a certain threshold, a potential instability alarm W1 for the EMG part will be triggered.

[0041] Furthermore, corresponding to the motion posture sensor 7 in the device, the physical motion characteristics it collects can be used as a basis for detecting the wearer's center of gravity and posture angle. Under normal circumstances, when the user is in a normal standing posture, the human body's center of gravity is approximately located at the second sacral vertebra. In this embodiment, it is suggested that when the motion posture sensor 7 detects that the wearer's center of gravity has deviated from the standard position by more than specific thresholds of 15° and 45°, and when the three-axis accelerometer shows a sudden and significant change, such as greater than 8G, as a threshold, posture instability alarms W2 and W3 will be triggered respectively. These two alarms can correspond to a slower, dependent fall and a violent, direct fall, respectively, serving as a basis for classifying alarm urgency.

[0042] In this embodiment, the transverse abdominis muscle electromyography (EMG) signal is processed using wavelet transform to extract its root mean square (RMS), median frequency (MF), and temporal entropy. These feature values ​​are then input into a lightweight convolutional neural network (CNN) model. The lightweight CNN model can employ a 1D-CNN+LSTM algorithm, which combines a one-dimensional convolutional neural network and a long short-term memory (LSTM) network to obtain the transverse abdominis muscle stability scoring parameters.

[0043] In the fall prediction and response mechanism used by the device, the relevant alarm logic is as follows: when the transverse abdominis stability score parameter is less than 0.7 (with a score of 1 as the maximum), the device triggers a W1 alarm locally, but only records it in the microprocessor 5; when the local W1 alarm accumulates for one hour between 0:00 and 24:00 on a single day, if the device does not trigger a W2 or W3 alarm, it will only send a warning of the potential fall possibility to the wearer and their related family members through a buzzer and wireless communication module 8. The warning level increases from medium to high with the duration of the W1 alarm, so that relevant personnel can pay close attention to the wearer's core stability ability to prevent falls.

[0044] Conversely, if the transverse abdominis stability score is greater than or equal to 0.7, the device determines the result to be normal. If the device does not trigger alarms W2 or W3, the device continues to operate normally in the routine monitoring state, and the microprocessor 5 performs continuous monitoring.

[0045] If the device has triggered a local W1 alarm, and the accumulated duration of the local W1 alarm reaches 30 minutes in a single day, and a W3 alarm has also been triggered, it is determined that the user has fallen. The device will directly report the wearer's real-time location information to the hospital and the user's family through the wireless communication module 8, and control the playback of a distress audio message to attract the attention of relevant personnel so that the user can receive rescue as soon as possible. Furthermore, if there is no local W1 alarm but a W3 alarm is triggered, it may indicate an accidental fall due to other reasons. In this case, the device will first alert the wearer. If the fall is severe and the user cannot turn off the alarm by resetting the device, the device will then execute the same process as the subsequent alarms mentioned above. Otherwise, the triggering of the W3 alarm may be due to movement under the patient's control, rather than a fall. In this case, the user can reset the device within a preset time to allow the device to continue to operate normally in the normal monitoring state.

[0046] If the device has triggered the local W1 alarm, and the local W1 alarm + W2 alarm have accumulated for more than one hour in a single day, then a warning about the potential fall will be sent to the wearer and their associated family members via a buzzer and wireless communication module 8. The fall risk corresponding to this warning is high, and the warning level is directly set to high level, thereby strongly reminding the wearer and their associated personnel to pay attention to their physical condition and seek medical treatment in a timely manner if necessary.

[0047] In addition to the aforementioned fall prediction and response mechanism and its logical design, the device in this embodiment can also perform its intended functions under other mechanism logics, so they will not be described in detail here. Through the application of the aforementioned fall prediction and response mechanism in the device of this embodiment, an early warning can be issued minutes or even hours before the wearer actually falls, significantly improving timeliness compared to existing technologies. When an actual fall occurs, the device's alarm timeliness is no less than that of any existing single-modal device. Furthermore, this embodiment combines a dual-modal data judgment process using transverse abdominis muscle electromyography signals and physical motion characteristics, which can filter out false alarms caused by special situations such as squatting to retrieve objects or tying shoelaces, significantly improving the accuracy of fall prediction and greatly enhancing the actual user experience.

Claims

1. A fall prediction wearable device based on transversus abdominis muscle electrical signals, characterized by: The device consists of an upper shell (1), a bottom shell (2), and an electrode plate (4) arranged from top to bottom. The bottom shell (2) has a cavity, in which a microprocessor (5) and a motion pose sensor (7) are arranged. The microprocessor (5) is electrically connected to the electrode plate (4) and the motion pose sensor (7). The lower surface of the electrode plate (4) is attached to the user's transverse abdominis muscle to collect the user's transverse abdominis muscle electromyography signal. The motion pose sensor (7) is used to collect the user's physical movement characteristics.

2. The fall-prediction wearable device of claim 1, wherein: An AD acquisition module (6) is also provided inside the cavity. The microprocessor (5) is electrically connected to the electrode (4) through the AD acquisition module (6). The microprocessor (5) and the AD acquisition module (6) are connected by a four-wire connection using the SPI protocol. The AD acquisition module (6) and the electrode (4) are connected by positive and negative electrode lines. The AD acquisition module (6) is used to convert the electromyographic signal of the user's transverse abdominis muscle acquired by the electrode (4) into a digital signal and then transmit it to the microprocessor (5).

3. The fall-prediction wearable device of claim 1, wherein: The motion posture sensor (7) has an inertial measurement unit, which consists of a three-axis accelerometer and a gyroscope. The three-axis accelerometer is used to obtain the static tilt angle and acceleration value of the user's posture at the device wearing point, and the gyroscope is used to obtain the instantaneous attitude angle of the user's posture at the device wearing point. The user's physical motion characteristics are composed of the data obtained by the inertial measurement unit. The microprocessor (5) and the motion posture sensor (7) are connected by a four-wire connection using the SPI protocol.

4. The fall-prediction wearable device of claim 1, wherein: The cavity is also equipped with a wireless communication module (8). The microprocessor (5) and the wireless communication module (8) are connected by a four-wire UART serial port protocol. The wireless communication module (8) is used for wireless communication between its own device and other fall prediction wearable devices worn by the same user, as well as for wireless communication between its own device and the user's personal terminal device.

5. The fall-prediction wearable device of claim 1, wherein: The cavity also contains a power management module (9), and the microprocessor (5) is connected to the power management module (9) by positive and negative power lines. The power management module (9) has a built-in power management chip and battery, and is used to power all the electrical components in the fall prediction wearable device.

6. The fall-prediction wearable device of claim 5, wherein: A standard interface (10) is provided on the bottom shell (2) at the location corresponding to the power management module (9); the standard interface (10) is connected to the microprocessor (5) and the power management module (9) respectively, and the standard interface (10) is used to provide battery charging input and external data input.

7. The fall-prediction wearable device of claim 1, wherein: The upper shell (1) and the bottom shell (2) are both made of flexible TPU material; a highly breathable lightweight textile fabric (3) is also provided between the bottom shell (2) and the electrode sheet (4); the highly breathable lightweight textile fabric (3) has a preset thickness value, which is used to provide breathable isolation between the bottom shell (2) and the electrode sheet (4).

8. The fall-prediction wearable device of claim 1, wherein: The upper shell (1) is also provided with a function button (11). The microprocessor (5) is electrically connected to the function button (11). The function button (11) is used to feed back the operation instructions from the user to the microprocessor (5).

9. The fall-prediction wearable device of claim 1, wherein: A buzzer is also provided inside the cavity, and the microprocessor (5) is electrically connected to the buzzer; a speaker hole is provided on the upper shell (1) at the location of the buzzer.

10. The fall-prediction wearable device of claim 1, wherein: Multiple fall prediction wearable devices are worn on the transverse abdominis muscle of the same user, and the various fall prediction wearable devices communicate with each other wirelessly.

Images