Background
The functional electric stimulation is widely applied to the gait function rehabilitation field, and the low-frequency pulse current is sent out by a plurality of groups of electrodes positioned at the corresponding parts of the lower limbs of the patient to stimulate muscles, induce the muscle movement or simulate the autonomous movement so as to achieve the purpose of rehabilitation physiotherapy. The trigger in posture correction is equivalent to a switch for controlling low-frequency current, and is used for controlling the output and stop of the stimulation current, and then the low-frequency current reaches the electrode by using a single path or multiple paths.
At present, the walker recorded in the market or literature uses functional electric stimulation to perform rehabilitation treatment on the patients with the difficulty in walking of lower limbs, but the control mode of low-frequency stimulation current and the stimulation intensity easily cause the patients to fall down and increase the falling times of the patients in the early stage of the treatment of the patients, in particular to the period of inadaptation to the treatment mode or the current stimulation intensity. The control mode and the stimulation intensity of the frequency stimulation current of the traditional walking aid are mainly divided into two types, namely fixed type and comparative type. Under the condition of fixed type, the control mode of the low-frequency stimulation current is fixed and cannot be adjusted, the walking aid cannot be changed according to the gait condition of the patient, more patients cannot adapt to the walking aid during use, the patient is easy to fall down, and the correction success rate is low; the comparison mode is to collect the gait information and the standard gait information of the patient, and determine the control mode and the output intensity of the low-frequency stimulation current according to the difference value of the gait information and the standard gait information of the patient. The walking aid adopting gait comparison type is still in existence, the problem that the patient falls easily is caused, normal person gait information is adopted by standard data at the beginning of establishment, the situation of the patient is quite different, the difference value of the gait parameters is quite different, although the machine can set corresponding correction strategies aiming at the difference, the adaptability problem of the patient in the correction process is ignored, namely, the difference between the gait parameters of part of the patients and the standard gait parameters is quite large at the beginning of correction, the difference value compared by the machine is difficult to adapt to the patient, the patient falls easily, and the problem that the concrete situation of the patient cannot be adapted to due to the complex difference of tolerance, health degree and the like of part of the patients still exists due to the simple difference between the standard gait information and the gait information of the patient. Therefore, the pulse electric stimulation intensity and the control mode of the two modes still cannot be suitable for all users, cannot be adjusted according to the real-time gait characteristics of the users, and risk that the patients fall easily exists.
Disclosure of Invention
Aiming at the situation, the invention firstly provides a multichannel foot drop dynamic electric stimulation walking aid based on the Internet of things, and aims to solve the problems that the traditional walking aid cannot adjust the low-frequency current stimulation intensity and the control mode according to the specific situation of a patient, so that the patient falls easily.
The technical scheme is that the system comprises a server end, a WEB end, a host end, a trigger end, an electrode and a sensor end; the sensor end is configured to be capable of measuring original parameter information of the lower limb posture of a patient walking, and sending and transmitting the original parameter information to the signal receiving end of the host end; the multiple groups of electrodes are configured into a structure which can be attached to and electrically conductive with muscle groups/groups required by the walking of the lower limbs of the patient; the trigger end is configured to receive an instruction sent by the host end and control one or more groups of electrodes to send or stop electric stimulation according to the instruction; the host end is configured to receive the manual control instruction and send an electrical stimulation control time sequence and low-frequency current stimulation intensity corresponding to the manual control instruction to the trigger end; the server side obtains a control algorithm capable of outputting a time sequence of each electrode output node and outputting electric stimulation intensity by using a rehabilitation patient treatment correction big data RNN (RNN cyclic neural network) module and a GRU (grid-controlled cyclic unit) network module; the host end comprises a signal receiving module, a signal sending module and a signal processing module, wherein the signal receiving module is used for receiving signals sent by the sensor end and signals from the service end, and the signal processing module is configured to process original information describing the walking posture of the lower limb of the patient from the sensor end to obtain standard posture information describing the walking of the lower limb of the patient and conforming to a set rule; the standard posture information is sent to a server through a signal sending module, and the server obtains a suggested treatment scheme capable of outputting a time sequence of each electrode output including a node and outputting electric stimulation intensity through an input control algorithm for the standard posture information; the WEB terminal is configured to share data of the server terminal, the WEB terminal is configured to send control information comprising an electrical stimulation control time sequence and low-frequency current stimulation intensity to the host terminal, the control information is received through a signal receiving module of the host terminal, and the control information controls the trigger terminal to send out control instructions corresponding to one or more groups of electrodes to send or stop electrical stimulation.
In one embodiment, the sensor end includes an array of pressure sensors located at the plantar aspect, the array of pressure sensors configured to measure signals of pressure distribution and pressure variation at the plantar aspect of the patient; the device also comprises an acceleration sensor for measuring the acceleration information of the ankle joint; the device also comprises an angular velocity sensor for measuring the angular deflection of the lacquer joints when the patient takes a step; and each sensor converts the physical quantity change of the lower limb of the patient into an electric signal, namely the original parameter information.
In an embodiment, the host side is further configured as a device for recording usage information of each sensor, and each parameter information of the usage information of each sensor and posture information of the lower limb of the patient is simultaneously sent to the server side through the signal sending module and stored in the storage module located at the host side.
In an embodiment, the WEB side is configured to display and manage data of the lower limb posture information of the patient at the server side and use record information of the patient instrument.
Aiming at the situation, the invention also provides a multichannel foot drop dynamic electric stimulation control method based on the Internet of things, which aims to solve the problems that the traditional walking aid device cannot adjust the low-frequency current stimulation intensity and the control mode according to the specific situation of a patient, and the patient falls easily.
The sensor end sends the original parameter information to a signal receiving end of the host end; the signal processing module processes the obtained original parameter information of each sensor to obtain gait characteristic parameters describing the posture information of the lower limb of the patient, namely the standard posture information; inputting the gait characteristic parameters of each sensor into an RNN (RNN-RN) circulating neural network module algorithm according to a set time sequence, and obtaining an output node time sequence corresponding to each group of electrodes by the circulating neural network module; the gait characteristic parameters of the sensors are input into a GRU gating circulation unit network module according to a set time sequence to establish dynamic signals of functional electrical stimulation intensity; the functional information formed by matching the sequence of the output node with the stimulation intensity of the low-frequency current of each group of electrodes is actively received or received by a signal receiving module at a host end, and multichannel dynamic control of functional electric stimulation is generated by a control module positioned in a trigger end; the WEB terminal determines a corrective treatment strategy according to a reference scheme and treatment experience, and the corrective treatment strategy can be actively or passively transmitted to the host terminal through a signal receiving module positioned on the host terminal; the trigger end is configured to receive an instruction sent by the host end and control one or more groups of electrodes to send or stop electric stimulation according to the instruction; the host end is configured to receive the manual control instruction and send the electrical stimulation control mode and intensity corresponding to the manual control instruction to the trigger end.
The invention does not agree with the walking aid device recorded in the market or literature, the scheme combines the treatment data of the same type of patients in the whole treatment stage through the neural network algorithm to give a reference treatment scheme, the corresponding reference treatment scheme is shared and managed by the WEB end at the same time, medical staff positioned at the WEB end can intervene and adjust the treatment scheme at any time by combining the specific condition and treatment experience of the patients, and the medical staff can directly control and adjust the treatment scheme of the patients in the whole treatment scheme according to the specific condition of the patients, and can monitor and review the whole correction process of the patients at any time.
Detailed Description
For further explanation of the aspects of the present patent, the aspects of the present invention will be further described in detail by the following examples or embodiments with reference to the accompanying drawings.
In some embodiments, server side 1, WEB side 6, host side 2, trigger side 3, electrodes 4, and sensor side 5 are included. The server 1 adopts a cloud server, the WEB 6 is in communication connection with the server 1 through a TCP/IP communication protocol or other network communication modes, and the WEB 6 is configured to display, review, manage, share files, pictures, documents, audio, video and other data located at the server 1, in particular to data related to a patient treatment scheme. The cloud server is also used for an operation control algorithm, the operation control algorithm comprises an RNN (RNN) circulating neural network module and a GRU (grid-controlled unit) circulating unit network module, the control algorithm is used for generating corresponding recommended treatment schemes by matching patient data with big data of similar patients, the data of the similar patients are patient data with good treatment results and suffering from the same diseases, the patient data comprise output node time sequences and electric stimulation intensities of the electrodes 4 in each treatment stage in the whole treatment period, the output node sequences are output sequences and output time periods of the electrodes 4, the data of the similar patients are mainly used for training the control algorithm so that the recommended treatment schemes outputted by the control algorithm are close to or equal to ideal treatment schemes, the recommended treatment schemes are displayed on the WEB end 6, and medical staff positioned on the WEB end 6 can regulate and control the treatment schemes at any time.
In the above embodiment, the host end 2 is provided with a signal transmitting and receiving module, where the signal transmitting and receiving module may include, but is not limited to, a WIFI module, and a sensor located at the sole, ankle, knee joint, etc. of the patient in the sensor end 5 transmits a signal for sensing and obtaining walking posture information of the lower limb of the patient to the host end 2 through bluetooth or wired or wireless communication, and forms a digital signal after noise reduction, sampling and normalization processing, where the digital signal forms gait characteristic parameters for describing gait information of the patient and transmits the gait characteristic parameters to the cloud server through WIFI or other communication modules, and the digital signal is subjected to operation processing by a control algorithm of the cloud server to obtain a suggested treatment scheme.
In some embodiments of the present invention, the host 2 further includes a storage module, where the storage module is configured to store the gait feature parameters of the patient and the usage information of each sensor, such as the usage time, the usage number, etc., and the gait feature parameters and the practical information are simultaneously transmitted to the cloud server through the WIFI or other communication modules, and the cloud server establishes a personal database for the WEB 6 to review.
In some embodiments, the system further includes a client 7, where the client 7 may access the cloud server through a TCP/IP communication protocol, query and manage patient treatment plan information and usage record information of the cloud server, and the client 7 may set and manage a patient treatment plan and generate a new treatment plan to be transmitted to the host 2 through a WIFI module of the host 2.
In some embodiments, the host end 2 and the trigger end 3 may be integrated in the same electronic device, where the host end 2 is configured as a device with a display module, and the display module displays a man-machine interaction interface for a user to operate or review treatment plan information, use record information, and the like, where the treatment plan information and the use record information are stored in a storage module located at the host end 2; the trigger end 3 comprises a control module, and the control module is used for controlling the output parameters of the electric stimulation pulse, including the frequency corresponding to the time sequence of the output node and the amplitude corresponding to the electric stimulation intensity.
In some embodiments, the noise reduction, sampling and normalization of the electrical signals of the sensors of the sensor terminal 5 are completed at the sensor terminal 5, and the sensor terminal 5 includes a bluetooth communication module and a microprocessor module, where the microprocessor module is used to convert the electrical signals of the sensors into digital signals and transmit the digital signals to the host terminal 2 through the bluetooth communication module.
In a specific use process, the walking aid device mainly comprises a platform part and a physical device, wherein the platform part consists of a WEB end 6, a user end 7 (mobile phone APP) and a cloud server; the entity device part consists of a host, a trigger positioned on the sole of the foot and one or more groups of stimulation electrodes 4; the host computer is connected with the stimulating electrode 4 through a lead, the host computer is connected with the pressure sensor positioned on the sole of the foot through Bluetooth, and the host computer is connected with the platform through Wifi of 2.4 GHZ.
Acceleration data acquired by an acceleration sensor, angular velocity deflection acquired by an angular velocity sensor, and pressure variation acquired by a pressure sensor are used for obtaining a gait characteristic curve graph and a three-dimensional model of a user. And then comparing and analyzing the model through a control algorithm, primarily obtaining a treatment reference scheme of a user for personnel to reference, and simultaneously recording all parameters and use conditions of a patient to form a statistical graph which is convenient to compare and improve. The WEB terminal 6 specially manages all data of the cloud through a network, downloads and displays treatment scheme information of a patient from the cloud, and can also read real-time use states, instrument parameters and use records, and the situation is improved by comparing before and after use so as to facilitate the patient to adjust a rehabilitation training scheme timely. The user terminal 7 operation interface downloads and displays user real-time information, use state, walking step number, walking time, instrument real-time output parameters, gait curve graph and the like from the cloud through Wifi, and the user terminal 7 can manually change the operation parameters and control commands of the instrument.
The pressure sensor in the plantar trigger is used for collecting pressure distribution and variation of the sole of a user in real time, the acceleration sensor in the host is used for collecting ankle joint acceleration data in real time when the user takes a step, and the angular velocity sensor is used for collecting knee joint angular velocity deflection in real time when the user takes a step.
The control algorithm comprises an RNN (RNN-cycle neural network) module and a GRU (gate-controlled loop unit) network module. The MCU in the host computer converts the received plantar pressure signal, ankle joint acceleration signal and knee joint angular velocity signal into gait characteristic parameters, the sensor data are input according to a given sequence through a single time-sequence circulation unit or all circulation units of the RNN circulation neural network module, and then the GRU gating circulation unit network module establishes a dynamic signal of the functional electrical stimulation intensity.
In some embodiments, the stimulating electrode 4 is a magnetic button electrode 4 sheet, the pressure sensor is a multipoint film pressure sensor, the acceleration sensor is a piezoresistive acceleration sensor, the angular velocity sensor is a three-coordinate gyroscope angular velocity sensor, the WEB terminal 6 supports various systems and versions for PC software, the user terminal 7 is a smart phone or a tablet computer, and the system can support android and IOS plug systems.
The invention does not agree with the walking aid device recorded in the market or literature, the scheme combines the treatment data of the same type of patients in the whole treatment stage through a neural network algorithm to give a reference treatment scheme, the corresponding reference treatment scheme is shared and managed by the WEB end 6 at the same time, medical staff positioned at the WEB end 6 can interfere and adjust the treatment scheme at any time by combining the specific condition and treatment experience of the patients, and the medical staff can directly control and adjust the treatment scheme of the patients in the whole treatment scheme according to the specific condition of the patients in a remote mode, and can monitor and review the whole correction process of the patients at any time.