CN105078449B - Senile dementia monitor system based on health service robot - Google Patents

Abstract

The invention discloses an Alzheimer's monitoring system based on a health service robot, which includes a health service robot, an intelligent terminal and a cloud server. The health service robot includes a robot body, a main control unit, a human-computer interaction unit and a medical detection unit; The interaction unit is connected to the main control unit, which includes a tablet computer, which is placed on the chest of the robot body; the medical detection unit is connected to the main control unit, which includes an EEG detection device independent of the robot body, and the EEG detection device passes through The Bluetooth signal is connected to the smart terminal and the tablet computer; the smart terminal and the tablet computer are connected to the cloud server through the mobile Internet, and the data interaction between the smart terminal and the tablet computer is realized through wireless signals. The invention can realize the automatic auxiliary diagnosis and treatment of senile dementia, improves the accuracy of diagnosis, is beneficial to the prevention and early detection of senile dementia, alleviates the aggravation of the disease, and achieves the purpose of curing.

Description

Senile dementia monitoring system based on health service robot

technical field

The invention relates to a health service system, in particular to an Alzheimer's monitoring system based on a health service robot, belonging to the fields of disease monitoring, auxiliary treatment and nursing.

Background technique

Alzheimer's disease (also known as "Alzheimer's disease") is a neurodegenerative disease that can lead to cognitive dysfunction in memory, executive power, visual space, language communication, abstract thinking, learning and calculation. Alzheimer's disease often occurs in old age and early old age, and the risk of the disease doubles with age. The prevalence rate is generally 4%-8% after the age of 60, and increases to 10% after the age of 65. will exceed 30%. In recent years, the number of Alzheimer's patients in China has shown a significant upward trend. According to statistics, the total number of patients with Alzheimer's disease in my country is about 6 million, ranking first in the world. About 1/4 of the global patients are in China. According to the "Anti-Alzheimer's Disease Market Research Report", as the problem of my country's aging population becomes more and more prominent, it is predicted that by 2020, there will be 10.2 million people with Alzheimer's disease in China. At the same time, due to the insidious onset of Alzheimer's disease, which is easily confused with physiological aging, it is difficult to detect Alzheimer's disease early, and it is easy to be ignored by patients and their families, thus losing the best opportunity for treatment.

At present, Alzheimer's disease is mainly evaluated by cognitive-spiritual scales after the appearance of dementia symptoms, combined with imaging and other examinations to make a clinical diagnosis, but when the clinical symptoms are obvious, the patients with Alzheimer's disease diagnosed are basically in the middle and late stages , and senile dementia is pathologically irreversible. Under the existing circumstances, on the one hand, there is no effective treatment method for senile dementia at home and abroad. In the specific stage of dementia, the disease can be alleviated or stabilized to a limited extent, and the curative effect cannot be achieved. However, the use of music feedback therapy and cognitive function training to treat Alzheimer's patients is more conducive to the physical and mental health of patients. Using biofeedback therapy, music therapy can improve the elderly's The quality of life of patients with dementia, these improvements are manifested in improved sleep, improved memory, emotional stability, enhanced expressive ability, and it is non-invasive and has no side effects. In addition, drug treatment is only a link in the prevention and treatment of Alzheimer's disease, and it can only be implemented after the patient has symptoms of dementia. Therefore, it is difficult for this treatment to achieve satisfactory curative effect. Clinical information system and auxiliary diagnosis system, auxiliary treatment system for dementia patients. Due to the relatively insufficient number of neurologists in China, the early diagnosis, daily care and adjuvant treatment of patients are greatly affected, which objectively leads to difficulties in the diagnosis of Alzheimer's disease and delays in diagnosis time, as well as the lack of daily care and adjuvant treatment. , that is, lack of timeliness and real-timeness. Therefore the senile dementia monitoring system based on the health service robot of the present invention will solve the above problems.

Existing dementia care requires a combination of palliative care and effective nursing care, and in many countries, especially China, family caregivers often see this not just as a duty, but as a way of expressing love and being loyal to their relationship , so to a large extent, Alzheimer's patients rely on the family system for care. Caregivers of dementia patients are under various pressures. They not only need to learn knowledge about diseases and medications, but also master nursing skills and psychological adjustment. Patients have little understanding of disease-related health knowledge, especially lack of communication skills, lack of safety nursing knowledge, lack of methods of exercising self-care for patients, and lack of knowledge about medication methods and adverse reactions. A large number of caregivers have not yet realized the importance of daily monitoring and adjuvant treatment for patients, they will overestimate the patient's physical and psychological functions, and ignore the fact that Alzheimer's patients cannot recognize the direction due to brain pathological factors, and cannot understand simple Incomprehensible explanations or incomprehensible simple instructions lead to depression in patients and even caregivers themselves.

Contents of the invention

The purpose of the present invention is to solve the defects of the above-mentioned prior art, and provide a kind of senile dementia monitoring system based on the health service robot. The automatic auxiliary diagnosis and treatment can improve the accuracy of diagnosis, which is beneficial to the prevention and early detection of Alzheimer's disease, alleviate the aggravation of the disease, and achieve the purpose of cure; it can solve the difficulty and diagnosis time of Alzheimer's disease in a timely and real-time manner It can also provide more scientific and reasonable guidance for the care of patients with dementia, so as to reduce the physical pain and psychological burden of patients with dementia and improve the quality of life of patients.

The purpose of the present invention can be achieved by taking the following technical solutions:

The senile dementia monitoring system based on the health service robot includes a health service robot, an intelligent terminal and a cloud server, and the health service robot includes a robot body, a main control unit, a human-computer interaction unit and a medical detection unit; the human-computer interaction unit It is connected with the main control unit, which includes a tablet computer, which is placed on the chest of the robot body; the medical detection unit is connected with the main control unit, which includes an EEG detection device independent of the robot body, and the EEG detection device is connected to the main control unit. The device is connected to the smart terminal and the tablet computer through the Bluetooth signal; the smart terminal and the tablet computer are connected to the cloud server through the mobile Internet, and data interaction is realized between the smart terminal and the tablet computer through wireless signals; wherein:

The EEG detection device is used to obtain the EEG information of Alzheimer's patients or healthy people in real time, and send the EEG information to tablet computers and smart terminals;

The tablet computer is used to receive EEG information of dementia patients or healthy people, collect voice information of children of dementia patients, and complete cognitive-psychological evaluation, sleep quality assessment and sleep quality assessment of dementia patients or healthy people. Cognitive function training, and upload EEG information, children's voice information, cognitive-psychological assessment, sleep quality assessment and cognitive function training information to the cloud server;

The intelligent terminal is used to receive EEG information of Alzheimer's patients or healthy people and input clinical information of Alzheimer's patients or healthy people, and upload the EEG information and clinical information to the cloud server;

The cloud server is used to receive the information uploaded by the tablet computer and the smart terminal, and perform data processing to complete the auxiliary diagnosis, generate corresponding nursing guidance suggestions, and feed back the auxiliary diagnosis results and nursing guidance suggestions to the smart terminal.

Preferably, the health service robot also includes a motion control unit, a binocular vision capture unit, an environment perception sensor unit and a power supply unit, and the motion control unit, the binocular vision capture unit and the environment perception sensor unit are respectively connected with the main control unit The power supply unit is used to supply power to the main control unit, motion control unit, binocular vision capture unit, human-computer interaction unit, environment perception sensor unit and medical detection unit.

Preferably, the EEG detection device includes a hat, an EEG sensor, an integrated analog front end, a mixed-signal microcontroller, a Bluetooth module, an input module, an indicator light module, and a power supply module; the EEG sensor is placed inside the hat, and The forehead of Alzheimer's patients or healthy people is in contact with the integrated analog front-end; the integrated analog front-end is connected with the mixed-signal microcontroller through SPI; the Bluetooth module is connected with the mixed-signal microcontroller through UART, and the Bluetooth module Used to connect with external equipment; the power supply module is used to supply power for the EEG sensor, the integrated analog front end, the mixed-signal microcontroller, the Bluetooth module and the indicator light module; the input module and the indicator light module are respectively connected with the mixed-signal microcontroller The input module is a switch of the EEG detection device, and the indicator light module is used to display the connection status between the EEG detection device and the main control unit, and the EEG detection function status of the EEG detection device.

Preferably, the bluetooth module includes a main control module, a radio frequency core module, a general peripheral device interface module and a sensor interface module; When transmitting externally, the signal is passed into the radio frequency core module, and the main control module includes a main controller, JTAG, ROM, flash memory and SRAM connected by wires; the radio frequency core module is used to receive the main Control the incoming signal of the module, and transmit the signal outward by the antenna. The radio frequency core module includes a co-controller connected by wires, a digital phase-locked loop, a DSP modem, SRAM, ROM and an amplifier, and the amplifier is connected to the antenna; The general peripheral device interface module includes I ² C, UART and SPI connected by wires; the sensor interface module includes a sensor controller, ADC and comparator connected by wires; the main control module communicates with the radio frequency core module, The general peripheral device interface module is connected with the sensor interface module.

Preferably, the main control unit includes a central processing unit, a general-purpose peripheral device interface module, a memory module, and a communication interface module; The data information of the capture unit, the human-computer interaction unit, the environmental perception sensor unit and the medical detection unit, the data information is stored in the memory module after being processed, and the central processing unit controls the motion control unit and the binocular vision capture unit through the communication interface module , human-computer interaction unit, environmental perception sensor unit and medical detection unit work.

Preferably, the motion control unit includes a motor drive module, an optocoupler isolation module, a motor unit, and a speed measuring encoder; the motor drive module is isolated from the main control unit by an optocoupler isolation module, and drives the motor unit to rotate; The speed measuring encoder is connected with the motor unit for real-time feedback of the position information and speed information of the motor unit to realize the closed-loop control of the position and speed of the motor unit; the motor unit is used to control the head rotation, waist rotation, mechanical Arm movement and chassis movement.

Preferably, the binocular vision capture unit selects the Kinect somatosensory sensor of Microsoft Corporation to realize the navigation and positioning functions of the robot, as well as the planning of the optimal path.

Preferably, the environment perception sensor unit includes a photoelectric switch, a gyroscope sensor, a touch sensor, an infrared sensor and an ultrasonic sensor; the photoelectric switch, the touch sensor, the infrared sensor, and the ultrasonic sensor work together to identify and avoidance; the gyro sensor interprets the attitude of the robot body; the environment perception sensor unit uses multi-sensor information fusion technology to process the sensed data, and performs feedback control through the main control unit.

Preferably, the power supply unit includes a charging base, a battery charging interface, a battery, and a voltage conversion module; the battery charging interface, the battery, and the voltage conversion module are integrated into the robot body, and the charging base is fixed indoors; , the voltage conversion module converts the voltage provided by the battery into the voltage required by the main control unit, motion control unit, binocular vision capture unit, human-computer interaction unit and environment perception sensor unit. When the threshold is reached, the robot automatically returns to the charging base for charging through the environment perception sensor unit.

Preferably, the tablet computer realizes the simulated emotional communication between the Alzheimer's patient and the children by collecting the voice information of the children of the Alzheimer's patient, specifically:

a. The tablet computer collects the voice information of the children according to the daily telephone conversations between the dementia patients and their children;

b. The tablet computer uploads the child's voice information to the cloud server for storage and processing, and establishes the child's voice information database;

c. According to the children's voice information database, the tablet computer uses voice recognition technology to simulate the speaking timbre of the children of Alzheimer's patients, and interacts with the Alzheimer's patients by voice.

Preferably, the tablet computer completes the sleep quality assessment of senile dementia patients or healthy people, specifically:

The method of combining energy features and least squares support vector machine is used to complete the automatic sleep staging, and then the sleep disorders are evaluated according to the automatic sleep staging received by the sleep quality assessment software. Quality is assessed.

Preferably, the tablet computer completes the cognitive function training of Alzheimer's patients or healthy people, specifically:

Establish personal cognitive function training files of Alzheimer's patients or healthy people, formulate training plans, and carry out unit model management. The training content includes: memory training and intelligence training, training in the form of graphic cognition and digital arithmetic cognition A course of treatment is divided into two stages, namely training and strengthening training.

Preferably, the cloud server processes the received EEG information as follows:

Use the ICA method to remove artifacts caused by irregular eye movements;

Mining the EEG information processed by the ICA method: using the correlation dimension method to analyze the nonlinear EEG signal to describe the complexity of the nervous system; Complexity; use EEG coherence analysis method to analyze EEG synchronization.

Preferably, the cloud server processes the received child voice information as follows:

a. Using linear predictive cepstral coefficients, Mel frequency scale cepstral coefficients and their dynamic features to form a mixed feature parameter method to extract speech features;

b. Adopt multi-feature mixed improved algorithm based on feature and Gaussian mixture model to model, and combine time-domain features and frequency-domain features, short-term stationarity and local variation rules through multi-feature combination.

Preferably, the cloud server performs deep learning and data mining based on big data information on the received EEG information, children's voice information, clinical information, cognitive-psychological evaluation information, sleep quality evaluation information, and cognitive function training information. ,as follows:

a. Use data divide and conquer and parallel processing strategy to carry out basic processing of big data information;

b. Use tensor decomposition for feature selection of big data information: use Tucker decomposition method for data decomposition, and use FSOM algorithm for feature extraction;

c. Use semi-supervised learning algorithms to classify big data information;

d. Use the FCM clustering algorithm to cluster big data information, and use the MapReduce model for large-scale parallel processing of data;

e. Use the Apriori algorithm to conduct association analysis on big data information

Preferably, the tablet computer also has a music feedback therapy function, and the music feedback therapy function is used to assist in the treatment of senile dementia patients, specifically:

According to the different conditions and different psychological characteristics of the patients, firstly, different therapeutic music is automatically selected, and secondly, a personal feedback program adapted to the different characteristics of each patient is established, and different biofeedback indicators such as EEG, respiration, and skin temperature are selected according to different conditions. The curative effect can be judged by dynamically observing the changes of physiological parameters during the training process of patients.

Compared with the prior art, the present invention has the following beneficial effects:

1. The senile dementia monitoring system of the present invention can receive the EEG information of senile dementia patients or healthy people through the smart terminal and the tablet computer on the health service robot, and can input clinical information of senile dementia patients or healthy people through the smart terminal. Information, the voice information of the children of dementia patients can be collected through the tablet computer, and the cognitive-psychological evaluation, sleep quality assessment and cognitive function training of the dementia patients or healthy people can be completed, and then the cloud server receives the smart terminal and tablet computer The uploaded information, using the deep learning and data mining technology based on big data information, can realize the automatic auxiliary diagnosis of Alzheimer's disease, avoid the one-sided and inconsistent diagnosis results caused by the over-reliance on the doctor's personal level and experience in the existing diagnostic methods, and improve the The accuracy of diagnosis, combined with cognitive function training to achieve the purpose of adjuvant therapy, alleviate the disease, and improve the quality of life of Alzheimer's patients. In addition, the Alzheimer's monitoring system based on health service robots has the advantages of good human-computer interaction, easy to move, easy to use, and long-term use.

2. The senile dementia monitoring system of the present invention can realize real-time detection of the EEG signals of senile dementia patients or healthy people by making senile dementia patients or healthy people wear EEG detection devices. Monitor the current condition of Alzheimer's disease and predict its development trend; for healthy people, it can effectively prevent Alzheimer's disease, detect Alzheimer's disease as early as possible, and take active intervention measures to slow down the physical damage and mental burden caused by Alzheimer's disease .

3. In the senile dementia monitoring system of the present invention, the tablet computer on the health service robot completes voice communication by simulating the voice of children, which can effectively alleviate the loneliness of dementia patients; through improving the sleep quality of dementia patients or healthy people Evaluation can help them keep abreast of their sleep health status, monitor changes in human body status, take appropriate treatment plans in time, reduce the risk of high-risk diseases for patients, and achieve the purpose of improving human sleep quality and human health.

4. In the senile dementia monitoring system of the present invention, the tablet computer on the health service robot can carry out music feedback therapy according to the psychological personality characteristics of senile dementia patients, and combined with cognitive function training to carry out auxiliary treatment for patients, which is beneficial to senile dementia patients. Improve the physical and mental health of patients with dementia, alleviate the disease, and improve the quality of life of patients with Alzheimer's disease. This improvement is manifested in improved sleep, improved memory, emotional stability, and enhanced expressive ability. It is non-invasive and has no side effects.

5. The senile dementia monitoring system of the present invention can automatically generate nursing guidance suggestions on the cloud server, and feed back to the smart terminal together with the remote medical auxiliary diagnosis results, and the smart terminal is synchronized to the tablet computer on the health service robot to give caregivers Provide scientific and reasonable nursing guidance and suggestions, so as to better control the condition of Alzheimer's patients, and also provide psychological adjustment suggestions for caregivers, so as to prevent caregivers from developing depression.

Description of drawings

Fig. 1 is the overall structural diagram of the senile dementia monitoring system of the present invention.

Fig. 2 is a structural block diagram of the health service robot of the present invention.

Fig. 3 is a functional principle diagram of the motion control unit of the present invention.

FIG. 4 is a structural block diagram of the EEG detection device of the present invention.

FIG. 5 is a structural block diagram of the Bluetooth module in the EEG detection device of the present invention.

Fig. 6 is a flow chart of the cloud server of the Alzheimer's disease monitoring system of the present invention.

Detailed ways

Example 1:

The present invention will be further described in detail below in conjunction with the embodiments and the accompanying drawings, but the embodiments of the present invention are not limited thereto.

As shown in Figure 1, the Alzheimer's disease monitoring system of this embodiment is applied in a certain family, including a health service robot, an intelligent terminal and a cloud server; wherein:

As shown in Figure 2, the health service robot includes a robot body, a main control unit, a motion control unit, a binocular vision capture unit, a human-computer interaction unit, an environment perception sensor unit, a medical detection unit and a power supply unit; The control unit, the motion control unit, the binocular vision capture unit, the human-computer interaction unit and the environmental perception sensor unit are arranged on the robot body; the main control unit communicates with the motion control unit, binocular vision and The capture unit, human-computer interaction unit, environmental perception sensor unit and medical detection unit are connected; among them, the motion control unit, binocular vision capture unit, human-computer interaction unit, environment perception sensor unit and medical detection unit are the top functional units.

The underlying operating system (software processing platform) of the robot body adopts an open source robot operating system (Robot Operating System, ROS), which includes hardware abstract description, underlying driver program management, execution of shared functions, message transfer between programs, and program distribution. Package management, distributed process framework, and system association supporting code base; the open source robot operating system is installed under the ubuntu (Ubuntu) system of the Linux kernel, and communicates with the main control unit through the serial port, thereby controlling the motion control unit, dual The working methods of the visual capture unit, human-computer interaction unit, environment perception sensor unit and medical detection unit.

The main control unit includes a central processing unit (CPU), a general-purpose peripheral device interface module, a memory module, and a communication interface module; Capture the data information of the unit, the human-computer interaction unit, the environmental perception sensor unit and the medical detection unit, and integrate the data. Stored in the memory module, the central processing unit sends and receives instructions through the communication interface module, and then controls the working mode of the motion control unit, binocular vision capture unit, human-computer interaction unit, environmental perception sensor unit and medical detection unit; The communication interface module includes I ² C (Inter-Integrated Circuit), CAN (Controller Area Network, controller area network) bus and UART (Universal Asynchronous Receiver/Transmitter, Universal Asynchronous Receiver Transmitter) and SPI (Serial Peripheral Interface, serial peripheral Interface) serial port communication module to meet the communication interface requirements between different functional units; the central processing unit can communicate with the operating system to complete the functional requirements of motion control, navigation and positioning, human-computer interaction, and data communication.

As shown in Figure 3, the motion control unit includes a motor drive module, an optocoupler isolation module, a motor unit and a speed measuring encoder; the motor drive module is used to receive the PWM (Pulse Width Modulation, pulse width modulation) sent by the main control unit. ) control signal to drive the motor unit to rotate, and the motor drive module is isolated from the main control unit by an optocoupler isolation module to protect the main control unit from the influence of motor voltage fluctuations; the speed measuring encoder is connected to the motor unit, It is used to feed back the position information and speed information of the motor group in real time to realize the closed-loop control of the position and speed of the motor group; the motor group can be composed of servo motors, DC motors, stepping motors, and high-torque steering gears, and is used to control the robot body head rotation, waist rotation, arm motion, and chassis motion.

The binocular vision capture unit selects the Kinect somatosensory sensor of Microsoft Corporation, and the Kinect somatosensory sensor is used to establish a 3D stereoscopic environment through three cameras, and through image recognition and processing, realizes the navigation and positioning functions of the robot, and the planning of the optimal path , so as to improve the decision-making ability of the main control unit.

The environment perception sensor unit includes a photoelectric switch, a gyroscope sensor, a touch sensor, an infrared sensor and an ultrasonic sensor; the photoelectric switch, the touch sensor, the infrared sensor, and the ultrasonic sensor work together to identify and avoid obstacles; The gyroscope sensor interprets the attitude of the robot body; the environment perception sensor unit uses multi-sensor information fusion technology to process the sensed data, and performs feedback control through the main control unit.

The power supply unit is used to supply power to the main control unit, motion control unit, binocular vision capture unit, human-computer interaction unit and environment perception sensor unit, which includes a charging base, a battery charging interface, a battery, and a voltage conversion module; The battery charging interface, battery and voltage conversion module are integrated into the robot body, and the charging base is fixed indoors; when the robot is working, the voltage conversion module converts the voltage provided by the battery into the main control unit, motion control unit, binocular The voltage required by the visual capture unit, the human-computer interaction unit, and the environment perception sensor unit. When the battery power is lower than the set threshold, that is, when the battery power is too low, the robot will automatically return to the charging base for charging through the environment perception sensor unit .

The human-computer interaction unit includes a voice interaction module and a tablet computer, and the voice interaction module includes a voice recognition unit, a voice synthesis unit, and a voice prompt unit, and the voice recognition unit is used to recognize voice instructions from the user; the voice synthesis The unit is used to process the recognized voice data, synthesize machine codes (which can be recognized by the top-level functional units), and send them to the main control unit for decision-making; the voice prompt unit can use a voice prompter for receiving the main control unit. The incoming control instructions give voice prompts to the user to realize the interactive function between the user and the robot; the tablet computer is placed on the chest of the robot body and can be displayed by touch.

The medical detection unit includes an EEG detection device independent of the robot body. The EEG information of Alzheimer's patients or healthy people in the family is acquired in real time by the EEG detection device. The EEG detection device communicates with the smart terminal through Bluetooth signals. , the tablet computer is connected; the tablet computer and the intelligent terminal are connected to the cloud server through the mobile Internet, and the data interaction is realized between the intelligent terminal and the tablet computer through wireless signals (the data channel can select cellular network/wireless local area network/Bluetooth).

As shown in Figure 4, the EEG detection device is a wearable device with low power consumption and high precision and adopts single-lead detection, which includes a hat (not shown in the figure), an EEG sensor, an integrated analog front end, Mixed-signal microcontroller, bluetooth module, input module (i.e. button), indicator light module and power supply module; the EEG sensor is placed inside the hat, in contact with the forehead of Alzheimer's patients or healthy people, and integrated with the analog front-end connected; the integrated analog front end is connected with the mixed-signal microcontroller through SPI; the bluetooth module is connected with the mixed-signal microcontroller through UART, and the bluetooth module is used to connect with external equipment; the power supply module is used for the EEG The sensor, the integrated analog front end, the mixed-signal microcontroller, the bluetooth module and the indicator light module are powered; The indicator light module is used to display the connection status between the EEG detection device and the main control unit, and the EEG detection function status of the EEG detection device.

As shown in Figure 5, the bluetooth module adopts low-power bluetooth standard V4.0 equipment, which can not only ensure high-speed transmission, but also solve the problem of excessive power consumption. It includes a main control module, a radio frequency core module, and general peripheral equipment Interface module and sensor interface module; the main control module is used to receive and store the signal from the mixed-signal microcontroller, and when the signal needs to be transmitted externally, transmit the signal to the radio frequency core module, the main control module includes wires Connected main controller, JTAG (Joint Test Action Group, joint test working group), ROM (Read-Only Memory, read-only memory), flash memory and SRAM (Static Random Access Memory, static random access memory); the radio frequency The core module is used to receive the incoming signal from the main control module when the signal needs to be transmitted externally, and transmit the signal externally through the antenna. The radio frequency core module includes a co-controller connected by wires, a digital phase-locked loop, a DSP modem, SRAM, ROM and amplifier, the amplifier is connected with the antenna; the general peripheral equipment interface module includes I2C, UART and low power consumption SPI connected by wire; the sensor interface module includes sensor controller, ADC (Analog) connected by wire to Digital Converter, analog-to-digital converter) and a low-power comparator; the main control module is connected to the radio frequency core module, the general peripheral device interface module and the sensor interface module through wires respectively.

The tablet computer can receive the EEG information of Alzheimer's patients or healthy people, collect the voice information of children of Alzheimer's patients, and complete the cognitive-psychological evaluation and sleep quality of Alzheimer's patients or healthy people on the user interface. Evaluation and cognitive function training, and upload EEG information, children's voice information, cognitive-psychological evaluation, sleep quality evaluation and cognitive function training information to the cloud server.

The tablet computer realizes the simulated emotional communication between the Alzheimer's patient and the children by collecting the voice information of the children of the Alzheimer's patient, specifically:

1) The tablet computer collects the voice information of the children according to the daily telephone conversations between the dementia patients and their children;

2) The tablet computer uploads the child's voice information to the cloud server for storage and processing, and establishes the child's voice information database;

3) Based on the children's voice information database, the tablet uses voice recognition technology to simulate the voice of the children of Alzheimer's patients. When the children fail to communicate emotionally with the patients, they can interact with the patients with Alzheimer's, such as simulating Alzheimer's disease. The patient's childlike voice reads stories or jokes aloud, alleviating the loneliness of Alzheimer's patients.

Described tablet computer completes the sleep quality evaluation of senile dementia patient or healthy person, specifically:

Automatic sleep staging is completed by combining energy features and least squares support vector machine (LS-SVM), and then sleep disorders are evaluated according to the automatic sleep staging received by the sleep quality assessment software, and the multimedia sleep quality is also used An assessment scale was used to assess sleep quality.

Described tablet computer completes the cognitive function training of senile dementia patient or healthy person, specifically:

Establish personal cognitive function training files of Alzheimer's patients or healthy people, formulate training plans, and carry out unit mode management. The training content includes: memory training (short-term, long-term) and intelligence training (acceptance, response ability, coping ability) , training in the form of graphic cognition, number arithmetic cognition, etc., each course of treatment is divided into two stages, namely training and intensive training.

Described tablet computer also has music feedback therapy function, and this music feedback therapy function is used for auxiliary treatment senile dementia patient, is specifically:

According to the different conditions and psychological characteristics of the patients, firstly, different therapeutic music is automatically selected, and secondly, a personal feedback program adapted to the different characteristics of each patient is established, and different biofeedback indicators such as EEG, respiration, and skin temperature are selected according to different conditions. The curative effect can be judged by dynamically observing the changes of physiological parameters during the training process of patients.

The intelligent terminal can receive the EEG information of Alzheimer's patients or healthy people, and input the clinical information of Alzheimer's patients or healthy people on the user interface, and upload the EEG information and clinical information to the cloud server; Clinical information includes the height, weight, gender, age, medical history, family history basic information and body temperature, blood oxygen saturation, blood pressure physiological parameters detection and psychological personality characteristics data of Alzheimer's patients or healthy people; in addition, due to smart terminals and tablets The data is interactive between the computers, so the data collected and processed by the tablet can also be obtained.

The cloud server workflow is shown in Figure 6. The cloud server receives the information uploaded by the tablet computer and the smart terminal, namely EEG information, children's voice information, clinical information, cognitive-psychological evaluation information, sleep quality evaluation information, cognitive Functional training information, and carry out corresponding processing; Described EEG information processing, EEG information comprises sleep EEG and non-sleep EEG, and non-sleep EEG utilizes ICA algorithm to process combined with EEG information, establishes a diagnostic model, diagnoses The severity of the disease in patients with Alzheimer's disease, while the sleep EEG adopts the method of combining energy characteristics and least squares support vector machine (LS-SVM) to complete the automatic sleep staging, and the sleep quality is assessed by the tablet computer combined with the sleep quality assessment scale. Evaluate; Described children's speech information processing, adopt the mixed feature parameter method that linear predictive cepstral coefficient, Mel frequency standard cepstral coefficient and their dynamic features form jointly to extract speech feature, and adopt the multiple based on feature and Gaussian mixture model Feature mixing improves algorithm modeling, and through an innovative multi-feature combination method, combines time-domain features and frequency-domain features, short-term stability and local variation rules, improves detection accuracy, and effectively handles common problems in tablet applications. The middle and low signal-to-noise ratio of the noise, and then carry out timbre processing, which is used for the simulation of the tablet computer and emotional communication with the children; the clinical information processing, clinical information includes physiological and pathological data, psychological personality characteristics, and uses data mining and deep learning to classify the disease Feature extraction, and the establishment of electronic medical records, and the establishment of personal feedback programs based on psychological personality characteristics, and music feedback therapy for tablet computers.

Described cloud server carries out following processing to the received EEG information (non-sleep EEG):

Use ICA (Independent Component Analysis, Independent Component Analysis) method to remove artifacts caused by irregular eye movements;

Mining the EEG information processed by the ICA method through a deeper analysis method to obtain more information, including:

1) Use the correlation dimension method to analyze nonlinear EEG signals to describe the complexity of the nervous system. In patients with cognitive impairment, the function and structure of the brain have changed, the connections between neurons have decreased, and the activity of the cerebral cortex has decreased. These representations are reflected by the relevant dimensions;

2) Using the Lempel-Ziv complexity (LZC for short) algorithm to determine the complexity of different brain regions when the brain is in different functional states, the better the brain development or function, the higher its complexity, and there is cognitive Patients with disorders exhibit lower complexity compared with normal individuals;

3) Using the EEG coherence analysis method to analyze the EEG synchronization can reflect the consistency of the two signals in a certain frequency range, and can indirectly reflect the degree of connection between the corresponding cerebral cortex. The greater the coherence coefficient between the two leads of the point, the stronger the cortical connection at the point where the lead is located. Compared with the normal elderly, the interhemispheric coherence of Alzheimer's patients has a more obvious decline.

Due to the huge amount of data received, the cloud server performs big data-based information on the received EEG information, children's voice information, clinical information, cognitive-psychological evaluation information, sleep quality evaluation information, and cognitive function training information. Deep learning and data mining, including the following steps:

1) Use data divide and conquer and parallel processing strategy to carry out basic processing of big data information;

2) Use tensor decomposition for feature selection of big data information: use MET (Memory-Efficient Tucker Decomposition) memory to use a more efficient Tucker decomposition method for data decomposition, and use FSOM (Fast Self-organizing Map, fast self-organizing map ) algorithm for feature extraction;

3) Use semi-supervised learning algorithm to classify big data information;

4) Use FCM (Fuzzy c-means, fuzzy c-means) clustering algorithm to cluster big data information, and use the MapReduce model for large-scale parallel processing of data;

5) Use the Apriori algorithm to conduct association analysis on big data information.

The cloud server automatically completes the auxiliary diagnosis and generates a corresponding combined with the results of remote medical auxiliary diagnosis, and then feed back the auxiliary diagnosis results and nursing guidance suggestions to the smart terminal. Guidance suggests syncing to tablet.

The intelligent terminal in the foregoing embodiments may be a smart phone, a PDA handheld terminal, and the like.

In summary, the senile dementia monitoring system of the present invention can realize the automatic auxiliary diagnosis and treatment of senile dementia, improve the accuracy of diagnosis, facilitate the prevention and early detection of senile dementia, alleviate the aggravation of the disease, and achieve the goal of cure. Purpose: It can solve the difficulty of senile dementia diagnosis and the delay of diagnosis time in a timely and real-time manner, as well as the lack of daily care and adjuvant treatment; it can also provide more scientific and reasonable guidance for the care of senile dementia patients, so as to alleviate the senile dementia. The physical pain and psychological burden of dementia patients, and improve the quality of life of patients.

The above is only a preferred embodiment of the patent of the present invention, but the scope of protection of the patent of the present invention is not limited thereto. Equivalent replacements or changes to the technical solutions and their inventive concepts all fall within the scope of protection of the invention patent.

Claims (7)

1. the senile dementia monitor system based on health service robot, it is characterised in that：Including health service robot, intelligence Can terminal and Cloud Server, the health service robot include robot body, main control unit, man-machine interaction unit, Motion control unit, binocular vision capture unit, environment sensing sensor unit and power supply power supply unit medical treatment detection unit；Institute State man-machine interaction unit, motion control unit, binocular vision capture unit and environment sensing sensor unit and main control unit It is connected, the power supply power supply unit is used for as main control unit, motion control unit, binocular vision capture unit, human-computer interaction Unit, environment sensing sensor unit and medical detection unit power supply, man-machine interaction unit includes tablet computer, the tablet computer It is placed in the front of robot body；The medical treatment detection unit is connected with main control unit comprising machine-independent human body EEG checking device, the EEG checking device is connected by Bluetooth signal with intelligent terminal, tablet computer；The intelligence is eventually End and tablet computer expert cross mobile Internet and are connected with Cloud Server, pass through wireless communication between the intelligent terminal and tablet computer Number realize data interaction；Wherein：

The underlying operating system of the robot body is using robot operating system of increasing income comprising hardware abstraction description, bottom Message transmission, program distribution package management, distributed process block between layer driver management, the execution of common functions, program Frame and support code library it is system combined；Robot operating system of increasing income is mounted under the ubuntu systems of linux kernel, is led to It crosses serial ports to be communicated with main control unit, and then controlled motion control unit, binocular vision capture unit, human-computer interaction list The working method of member, environment sensing sensor unit and medical detection unit；

The EEG checking device, the brain electric information for obtaining patients of senile dementia or Healthy People in real time, and by brain telecommunications Breath is sent to tablet computer and intelligent terminal；

The tablet computer, for receiving the brain electric information of patients of senile dementia or Healthy People, collecting patients of senile dementia Children's voice messaging, and complete cognition-Psychological Evaluation of patients of senile dementia or Healthy People, sleep quality assessment and recognize Know functional training, and by brain electric information, children's voice messaging, cognition-Psychological Evaluation, sleep quality assessment and training of cognitive function Information uploads to Cloud Server；Tablet computer realizes senile dementia by children's voice messaging of collection patients of senile dementia The simulation emotion communication of disease patient and children, specially：

Tablet computer is conversed according to patients of senile dementia and the call routine of children, collects the voice messaging of children；

The voice messaging of children is uploaded to Cloud Server and stored and processed by tablet computer, establishes children's voice messaging data Library；The children's voice messaging received is handled as follows in Cloud Server：It is fallen using linear prediction residue error, Mei Er frequency markings The composite character parametric technique that spectral coefficient and their behavioral characteristics constitute jointly extracts phonetic feature；Using feature based and The multiple features mixing innovatory algorithm of gauss hybrid models models, and by combination of multiple features mode, and temporal signatures and frequency domain are special Sign is combined, short-term stationarity and local changing rule are combined；

Tablet computer simulates patients of senile dementia children's according to children's voice messaging database, using speech recognition technology It speaks tone color, voice interface is carried out with patients of senile dementia；

The intelligent terminal, the brain electric information for receiving patients of senile dementia or Healthy People and input patients of senile dementia Or the clinical information of Healthy People, and brain electric information and clinical information are uploaded into Cloud Server；

The Cloud Server, the information uploaded for receiving tablet computer and intelligent terminal, and data processing is carried out, to complete At auxiliary diagnosis, and corresponding introduction on discharge suggestion is generated, and auxiliary diagnosis result and introduction on discharge suggestion are fed back into intelligence Terminal.

2. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute It includes cap, eeg sensor, integrated simulation front end, mixed signal microcontroller, bluetooth module, input to state EEG checking device Module, indicating lamp module and power module；The eeg sensor is placed on the inside of cap, with patients of senile dementia or health The forehead of people contacts, and is connected with integrated simulation front end；The integrated simulation front end passes through SPI and mixed signal microcontroller phase Even；The bluetooth module is connected by UART with mixed signal microcontroller, and the bluetooth module with external equipment for connecting；Institute State power module for be eeg sensor, integrated simulation front end, mixed signal microcontroller, bluetooth module and indicating lamp module Power supply；The input module and indicating lamp module are connected with mixed signal microcontroller respectively, and the input module is brain electric-examination The switch of device is surveyed, the indicating lamp module is used to show the connection status and brain of EEG checking device and main control unit The brain electro-detection functional status of electric detection means.

3. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute The sleep quality assessment that tablet computer completes patients of senile dementia or Healthy People is stated, specially：

Sleep mode automatically is completed by stages using the method that energy feature and least square method supporting vector machine are combined, then according to sleep The sleep mode automatically that quality evaluation software receives by stages assesses sleep disturbance, while also utilizing the sleep matter of multimedization Amount assessment scale assesses sleep quality.

4. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute The training of cognitive function that tablet computer completes patients of senile dementia or Healthy People is stated, specially：

It establishes the personal view functional training archives of patients of senile dementia or Healthy People and formulates drill program, carry out unit Schema management, training content have：Memory training and intelligent training are trained in the form of figure cognition, digital arithmetic cognition, Each course for the treatment of is divided into two stages, i.e. training and intensive training.

5. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute Cloud Server is stated the brain electric information received is handled as follows：

Artefact caused by removing irregular eye movement using ICA methods；

To by ICA methods, treated that brain electric information excavates：Non-linear EEG signals are carried out with Correlation Dimension method Analysis, portrays nervous system complexity；It is different when brain is in different function state with Lempel-Ziv product complexity theories The complexity of brain area；The synchronization analysis of brain electricity is carried out with brain electricity coherent analysis method.

6. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute Cloud Server is stated to the brain electric information, children's voice messaging, clinical information, cognition-Psychological Evaluation information, the sleep quality that receive Assessment information, training of cognitive function information carry out the deep learning based on big data information and data mining, as follows：

A, basic handling is carried out to big data information using data partition and parallel processing strategy；

B, the feature selecting of big data information is carried out using tensor resolution：Data decomposition is carried out using Tucker decomposition methods, with And carry out feature extraction using FSOM algorithms；

C, classified to big data information using semi-supervised learning algorithm；

D, big data information is clustered using FCM clustering algorithms, and the extensive of data is carried out with MapReduce model Parallel processing；

E, analysis is associated to big data information using Apriori algorithm.

7. the senile dementia monitor system according to claim 1 based on health service robot, it is characterised in that：Institute Stating tablet computer also has the function of that music feeds back therapy, which feeds back therapy function and suffer from for auxiliary treatment senile dementia Person, specially：

According to the different state of an illness of patient and different psychological characteristics of personality, different treatment music is automatically selected first, next is built The vertical personal feedback process for adapting to each patient's different characteristics selects the difference biology of brain electricity, breathing, Pi Wen according to the different state of an illness Index is fed back, by the variation of physiological parameter in dynamic observation patient's training process to judge curative effect.