A user memory status early warning device and method

Abstract

The present application proposes a user memory status early warning device, which comprises at least a data analysis module, and is characterized in that the data analysis module analyzes the correlation between the user's specific memory status and external conditions, and pushes a memory training plan based on the correlation to trigger the specific memory status.

Description

[0001] The present invention is a divisional application of application number 201810217913.7, application date March 16, 2018, application type is invention, and application name is a divisional application of a memory training device for the elderly. Technical Field

[0002] The present invention relates to the field of artificial intelligence, and in particular to a user memory status early warning device and method. Background Art

[0003] Memory training is common in children and adolescents, when intelligence is developing rapidly, and is often used to promote intelligence development to achieve a better level of intelligence development. The memory state of the elderly is in a declining stage or a stable stage. Therefore, in daily life, the correlation between the memory state and pathology of the elderly is blurred, and the importance of memory training for the elderly is often ignored. For example, severe memory decline in the elderly can induce serious complications such as Alzheimer's disease, confusion, mental depression, and communication difficulties, further causing the elderly to lose the ability to take care of themselves.

[0004] Due to individual differences among the elderly and different external conditions, the same memory training program cannot be applied to every user to the greatest extent. At the same time, compared with childhood or adolescence, the memory state of the elderly is easily affected by many factors such as illness, weather conditions, emotional state, sleep state, etc., and the memory of the elderly often has an explosive exponential decline after a downward trend, which in turn causes a series of serious complications such as Alzheimer's disease, confusion, slurred speech, etc. At this time, it is too late to carry out memory training for the elderly. Therefore, how to personalize memory training to suit different users and to warn the elderly of the trend of memory deterioration in advance and accurately and to carry out memory training in time to improve their memory state is an urgent problem to be solved.

[0005] The patent document with the publication number CN101478912A in the prior art discloses a method for analyzing an individual suffering from a dementia-type disorder, comprising: receiving a plurality of electroencephalogram data related to the individual; receiving a plurality of cardiovascular risk factor data related to the individual; receiving a plurality of cognitive data related to the individual; and determining whether the individual has an indication value of risk of suffering from a dementia-type disorder based at least in part on the electroencephalogram data, the cardiovascular risk factor data, and a portion of the cognitive data. The invention can provide a system and method for analyzing and evaluating dementia and dementia-type disorders by integrating the application of electroencephalogram (EEG), neuropsychological tests, and cardiovascular risk factors. The embodiments of the invention can utilize nonlinear analysis of EEG data that is not a linear analysis, thereby statistically combining the results of the nonlinear EEG data analysis with neuropsychological tests or cognitive tests and the determination of cardiovascular risk factors. Compared with the linear EEG determination used in conventional systems and methods, the invention can provide more reliable predictive information.

[0006] The above technical solution is to use EEG test data, psychological tests or cognitive tests, and cardiovascular risk factors in combination, and use statistical principles to predict dementia risk.

[0007] A patent document with publication number CN107533583A in the prior art discloses a human user mental damage monitoring system, which comprises: a measurement system (23) for repeatedly measuring an index of unconscious adjustment achieved by a human user, the measurement system comprising a processor (6), the processor being able to access a memory (7), the memory storing a time-related user-specific data set, the time-related user-specific data set comprising both a time-related user-specific physiological data set and a time-related user-specific behavioral data set, the time-related user-specific physiological data set comprising data representing a user's physiological parameters, the time-related user-specific behavioral data set comprising data representing a user's behavioral parameters; the processor (6), comprising a calculator (24), for repeatedly analyzing the time-related a user-specific data set so as to repeatedly determine the value of the unconscious adjustment index achieved by the human user through the time-related user-specific data set as a comprehensive index of the user-specific physiological data and the user-specific behavioral data; a computing system (25) for calculating a pattern index and pre-set structured experimental data, the pattern index indicating the possibility of the user displaying a mental damage pattern, indicating the position of the user within a mental damage grid, the mental damage grid including the mental damage pattern from the prediction model (16), based at least on the value of the unconscious adjustment index measured for the user, the pre-set structured experimental data stored in a memory (17) and indicating the relationship between the unconscious adjustment index value achieved by the experimental human user and the mental damage history data of the experimental human user.

[0008] The above technical solution focuses on the quantification of unconscious adjustments made by users as the main data available for assessing mental impairment. Doctors can take this data into account when diagnosing or treating mental health. The invention focuses on the quantification of unconscious adjustments made by users in their daily lives.

[0009] A patent document in the prior art with publication number CN104902806A discloses a method for providing an assessment related to a neurological disorder or a neuropsychiatric disorder of a subject, comprising: selecting an analysis category indicating one or more aspects of cognitive or sensory functions associated with a neurological disorder or a neuropsychiatric disorder; presenting a stimulation sequence to the subject, the stimulation sequence being based on the selected analysis category; collecting physiological signals of the subject before, during and after the presentation of the stimulation sequence to generate physiological data; and processing the physiological data to generate an information set containing one or more quantitative values ​​associated with the selected analysis category.

[0010] The above technical solution discloses a system, device and method for collecting and using physiological information (e.g., brain signals) to characterize the pathology and / or susceptibility of a subject's nervous system or neuropsychiatric disorder and / or evaluate the treatment of such a disorder. In other words, the above technical solution analyzes information about nervous system-related disorders by providing stimulation to the subject and detecting physiological data before, during and after the stimulation.

[0011] Compared with the prediction or detection devices for nervous system diseases described in the above-mentioned technical solutions, the present invention proposes a memory training device, which targets technical problems and achieves technical effects that are completely different from those of the above-mentioned technical solutions.

[0012] The above-mentioned multiple technical solutions only involve the technical solution of determining the type of the patient's disease or the degree of the disease according to the test data. They do not involve the technical feature proposed in this application of analyzing the correlation between the user's specific memory state and the external conditions, and pushing a memory training program based on the correlation to trigger the specific memory state. Summary of the invention

[0013] The present invention is derived from the discovery that no one person can accurately describe the cognitive or sensory experience brought about by a certain memory state. Although there are methods such as brain wave scanning and characteristic signal detection, and a large number of artificial intelligence algorithms such as CNN (convolutional neural network), RNN (recurrent neural network), and DNN (deep neural network) are used to construct memory models, the specific application effects of each can only be said to be unsatisfactory.

[0014] From a psychologist's perspective, memory status is manageable, or at least guideable. This management and guidance requires scientific methods and devices to assist in achieving the desired results. Therefore, the task of the present invention is to provide a device and method for the user to calibrate and record the memory status and its associated feelings, and thereby train the memory in a targeted manner. Through the present invention, on the one hand, the training of the memory analysis algorithm can be completed through the teaching process, and on the other hand, the user can continuously adjust the difference between the actual memory status and the theoretical memory status, so that the present invention can gradually learn the user's memory training characteristics, and then effectively manage the user's memory status.

[0015] A memory training device for the elderly, comprising at least a data interaction module for automatically collecting and / or inputting actual memory state information by a user, a data acquisition module for collecting body signals by indirectly or directly contacting the user's body, and a storage module for storing data information, which are respectively connected to a data analysis module in a wired and / or wireless manner. The data analysis module determines the theoretical memory state information corresponding to the body signal by using a memory analysis algorithm. The data analysis module completes the analysis and correction of the theoretical memory state change trend based on the user's input feelings, real-time memory state information and / or external conditions. The theoretical memory state change trend that exceeds the critical value at a foreseeable time point is classified and stored in the storage module in a manner that the user's input feelings, the actual memory state information, the body signal and the external conditions are all associated with each other. The data analysis module forms a boundary trigger condition for the theoretical memory state change trend exceeding the critical value based on the above-mentioned data information associated with each other, analyzes the correlation between the user's actual memory state information and the boundary trigger condition, and makes an early warning based on the correlation, pushes a memory training plan and / or a suggestion for improving the external conditions.

[0016] Due to the individual differences of the elderly and the differences in external conditions, the same memory training program cannot be applied to every user to the greatest extent. In order to realize the complete intelligence of memory training, it is necessary to train the memory analysis algorithm so that it can continuously narrow the gap between the theoretical memory state and the actual memory state. The above-mentioned technical means help the memory analysis algorithm to grasp the memory training characteristics of the current user, so that personalized memory training can be carried out. At the same time, it is also particularly important to determine the timing of carrying out memory training, especially for the elderly. For example, affected by many factors such as illness, weather conditions, emotional state, sleep state, etc., the elderly may have an exponential decline in memory state in an environment that is continuously unsuitable for external conditions, thereby causing a series of serious complications such as Alzheimer's disease, confusion, slurred speech, etc. At this time, it is too late to carry out memory training for the elderly. Therefore, the present invention includes the body signals, feelings and external conditions of the elderly into the monitoring test scope, and determines the boundary trigger conditions by analyzing the correlation between the elderly and the memory state change trend under different feelings, different body signals and different external conditions, so that the present invention can more accurately determine the timing of carrying out memory training.

[0017] According to a preferred embodiment, the memory training device further comprises a memory test module electrically connected to the data analysis module. The memory test module applies test information capable of obtaining the user's memory state information to the user in the form of text, voice, video and / or graphics. The data interaction module determines the changing trend of the memory state based on the feedback result of the test information. The data analysis module analyzes the teaching process of memory training based on the boundary trigger condition and / or the changing trend of the memory state. The parameters of the memory analysis algorithm are pre-configured according to at least two extreme memory state information determined in the teaching process. After determining the memory analysis algorithm, the actual memory state is accurately analyzed based on the boundary trigger condition and the changing trend of the memory state, so as to improve the analysis probability of the actual memory state of the individual, which is conducive to the formulation of a personalized memory analysis algorithm. The extreme memory state of the user leads to a serious decline in the user's memory, which in turn induces serious consequences such as a vague understanding of surrounding things, unclear personal consciousness, and loss of self-care ability. In the above case, the demand memory analysis algorithm has high accuracy to provide early warning information as soon as possible.

[0018] According to a preferred embodiment, the memory test module applies test information capable of obtaining the user's memory status information to the user in the form of text, voice, video and / or graphics. And the data interaction module automatically collects test feedback information corresponding to the test information. At the same time, the data acquisition module collects the user's body signals corresponding to the test feedback information, the user's input feelings and / or external conditions. The data analysis module pre-configures the parameters of the memory analysis algorithm based on the test feedback information, the external conditions, the user's input feelings and / or the body signals, and corrects the theoretical memory status information.

[0019] According to a preferred embodiment, the data analysis module generates a user memory profile related to the current user based on continuously adjusting the theoretical memory state information qualitatively and / or quantitatively, and analyzes the teaching process of memory training based on the boundary trigger condition, the change trend of the memory state and / or the user memory profile.

[0020] According to a preferred embodiment, the data analysis module records the external conditions that trigger the actual memory state of the user and stores them in the storage module in a form associated with the corresponding actual memory state, and stores the current memory state information input by the user and the external conditions in the storage module in a form associated with each other. After analyzing the correlation between the user's specific memory state and the external conditions, a memory training plan is pushed based on the correlation to trigger the specific memory state.

[0021] According to a preferred embodiment, the data analysis module analyzes and determines the corresponding theoretical memory state based on the collected body signals, and outputs stimulation information in the form of text, voice, video and / or graphics to change the external conditions and / or directly provide memory training to the user to improve the user's current memory state.

[0022] According to a preferred embodiment, the current memory state information input by the user and the external conditions related to the actual memory state information are stored in the storage module in a manner associated with each other. The user retrieves the actual memory state information in a manner associated with the external conditions through the data interaction module.

[0023] According to a preferred embodiment, the data analysis module corrects the theoretical memory state determined based on the analysis of collected body signals according to the actual memory state information, and the memory management profile composed of the corrected theoretical memory state is stored in the storage module in a manner that can be retrieved based on body signals.

[0024] According to a preferred embodiment, the data analysis module records the actual memory state and the theoretical memory state based on iterative analysis of at least one external condition, body signal and / or feeling at each moment. And the data interaction module displays the change trend of external condition changes, body signal changes, feeling changes and actual memory state in the form of a mapping-associated mark, reminds the user of the extreme change trend of the actual memory state in the form of color change, sound and / or vibration, and / or displays suggestions for changing the actual memory state.

[0025] The present invention also provides a memory training method for the elderly, which at least includes: automatically collecting and / or inputting actual memory state information by the user. Collecting body signals by indirectly or directly contacting the user's body. Determining the theoretical memory state information corresponding to the body signal using a memory analysis algorithm based on a preset database. Wherein, the analysis and correction of the theoretical memory state change trend are completed based on the user's input feelings, real-time memory state information and / or external conditions. The theoretical memory state change trend that exceeds the critical value at a foreseeable time point is classified and stored in a manner that is associated with the user's input feelings, the actual memory state information, the body signal and the external conditions. Wherein, a boundary trigger condition for the theoretical memory state change trend exceeding the critical value is formed based on the above-mentioned mutually associated data information. Analyze the correlation between the user's actual memory state information and the boundary trigger condition, and make early warnings based on the correlation, push memory training plans and / or suggestions for improving external conditions.

[0026] Beneficial technical effects of the present invention:

[0027] The present invention continuously narrows the gap between the theoretical memory state and the actual memory state by continuously training the memory analysis algorithm, and at the same time, enables the memory analysis algorithm to learn the memory training characteristics of the current user, so that personalized memory training can be carried out. Furthermore, the present invention includes the body signals, feelings and external conditions of the elderly into the monitoring test range, and determines the boundary trigger conditions by analyzing the correlation between the elderly's memory state change trend under different feelings, different body signals and different external conditions, so that the present invention can more accurately determine the timing of memory training, and avoid the exponential decline of the elderly's memory state in an environment that is continuously unsuitable for external conditions, thereby causing a series of serious complications such as Alzheimer's disease, confusion, slurred speech, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the connection relationship of each module of the present invention;

[0029] Figure 2 is a schematic diagram of the logical structure of the present invention; and

[0030] Figure 3 It is a display schematic diagram of the data interaction module of the present invention.

[0031] Reference numerals list

[0032] 10: Data interaction module; 20: Data acquisition module; 30: Data analysis module; 31: Database; 40: Storage module; 50: Memory test module. DETAILED DESCRIPTION

[0033] The following is a detailed description with reference to the accompanying drawings.

[0034] Example 1

[0035] like Figure 1 As shown, the data analysis module is connected to a data interaction module for automatically collecting and / or inputting actual memory status information by a user, a data acquisition module for collecting body signals by indirectly or directly contacting the user's body, a storage module for storing data information, and a memory test module in a wired and / or wireless manner. Preferably, the data interaction module can be a mobile terminal. The data acquisition module can be a detector. The data analysis module can be a cloud server. The memory test module can be a memory span tester, an intelligent terminal equipped with memory software, etc. The storage module can be a storage chip, a hard disk, a storage server, etc.

[0036] Preferably, the data analysis module includes at least three bidirectional interfaces, several data input interfaces and at least one data output interface. The data interaction module, the storage module and the database are connected to the data analysis module through different bidirectional interfaces for bidirectional communication. The data acquisition module and the memory test module are connected to the data analysis module through different data input interfaces. Preferably, the data output interface is used to connect and communicate with the upper or lower level equipment to transfer the data in the memory training device to the upper or lower level equipment. The data output interface is configured in a one-way communication mode in which data can only be output but not input to avoid the risk of data information in the memory device being stolen.

[0037] like Figure 1 , Figure 2 and Figure 3As shown, the present invention provides a device for memory training for the elderly. The memory training device of the present invention comprises at least a data interaction module 10, a data acquisition module 20 and a data analysis module 30. The data interaction module 10 is used to automatically collect and / or input actual memory status information by the user. Preferably, the data interaction module 10 comprises an image acquisition device and a video acquisition device. The data interaction module can be an intelligent terminal, such as a notebook, a mobile phone, a smart bracelet, a smart watch and other mobile devices, or a camera device. Preferably, the camera device comprises an ordinary camera device and a night camera device. Preferably, the automatic acquisition is to automatically collect the conversation records of the user in normal interpersonal conversation occasions, the user's description of a specific event, etc. Alternatively, the user inputs his or her memory status described in words into the data interaction module.

[0038] Preferably, the memory test module is connected to the data interaction module by wired and / or wireless means. The memory training module outputs stimulation information in the form of text, voice, video and / or graphics to change the external conditions and / or directly provide training stimulation to the user to improve the user's current memory state. The memory test module applies test information that can obtain the user's memory state information to the user in the form of text, voice, video and / or graphics, and automatically collects test feedback information corresponding to the test information. The data analysis module pre-configures the parameters of the memory analysis algorithm based on the test feedback information and / or body signals during the teaching process, and corrects the theoretical memory state information corresponding to the body signals. Preferably, the memory test module can be a VR device worn by the user. The VR device includes at least a sensing unit. The sensing unit can collect information related to the user's brain activity and generate an output signal during the memory test or memory training process. The output signal can be sent to the data interaction module wirelessly and / or wired.

[0039] The data acquisition module of the present invention is used to collect body signals by indirectly or directly contacting the user's body. The data acquisition module includes a variety of modules for collecting human physiological signals. The detection module of the data acquisition module includes at least one or more of a pulse sensor, a heartbeat sensor, a blood pressure sensor, a breathing rate sensor, a sound acquisition module, a hand vibration module and a footstep frequency sensor. Preferably, the detection module in the data acquisition module can also be equipped with a microwave signal sensor to detect changes in brain waves as required. The detection module in the data acquisition module can also be equipped with electrodes set on the head as required. Preferably, the body signals collected by the data acquisition module 20 are sent to the data analysis module 30 via the data interaction module 10.

[0040] The data analysis module 30 of the present invention is used to determine the theoretical memory state information corresponding to the body signal using a memory analysis algorithm based on a preset database. Preferably, the data analysis module 30 includes a database 31 storing a number of memory analysis algorithms. The data analysis module 30 analyzes and determines the corresponding theoretical memory state based on the body signal collected by the data acquisition module 20 and feeds back to the data interaction module 10. Preferably, the memory analysis algorithm includes machine learning algorithms such as Bayesian classification algorithm, neural network, support vector machine, decision tree, case-based reasoning learning, association rule learning, etc.

[0041] For example, the memory analysis algorithm of the data analysis module is described as follows.

[0042] In this embodiment, the minimum attribution degree fuzzy support vector machine or the average attribution degree fuzzy support vector machine is used to classify the memory state. The memory state can be divided into, for example, a healthy state, a sub-healthy memory state, and a severe memory decline state. Specifically, the classification steps of the memory analysis algorithm include:

[0043] A1: Define the fuzzy membership function

[0044] In order to avoid the generation of inseparable regions, the fuzzy membership function is introduced and defined as:

[0045]

[0046] A2: Define the attribution function

[0047] There are two ways to define the attribution function: minimum attribution and average attribution.

[0048] Minimum degree of belonging: m i =Min(m ij (x)), j=1, 2,...k, j≠i.

[0049] Average degree of belonging:

[0050] A3: Classification

[0051] Get k m i Then, put x into m i The largest category: arg{max(m i (x))}, i=1, 2,…k.

[0052] Preferably, the data analysis module 30 completes the teaching process of the data analysis module 30 based on the user's operation of the data interaction module 10. Preferably, the data analysis module includes at least one or more of a CPU, a processor, a microprocessor, a server, and a server group. First, the data analysis module 30 requires user cooperation to complete the teaching process. The teaching in the present invention refers to exemplary artificial intelligence programming.

[0053] For example, during the teaching process of the user using the device of the present invention, it is necessary to input emotional feelings, sleep feelings and current memory state information through the data interaction module. During the user using the device of the present invention, the data interaction module 10 actively collects the user's current external conditions. External conditions include weather conditions, noise index, user's current mental state, etc. Preferably, the user inputs the current emotional feelings and sleep feelings, and within the specified time range interval, the user again inputs the current emotional feelings and sleep feelings after moderate adjustment to enter the teaching process. During the teaching process, the data analysis module forms a memory analysis algorithm that matches the user based on the user's external conditions, sleep feelings and emotional feelings, and the memory state change trend. At the same time, the theoretical memory state is corrected.

[0054] For example: The teaching process includes:

[0055] S1: Start teaching mode;

[0056] S2: User selection: Select the teaching object according to the user name. If the teaching object is a user that has been recorded in the data storage module, the selection is made directly; if the teaching object is a user that has not been recorded, the user's memory status information is recorded through the data interaction module before selection. The user memory information includes the user name and user face photo, user sample memory and the external conditions and feelings associated with it;

[0057] S3: Sample selection: If the sample memory content and teaching matching content of this teaching have been stored under the user name, the stored sample memory content and teaching matching content are directly selected; if the sample memory content and teaching matching content of this teaching are not included in the user name, the teaching matching content of this teaching is read in through the data interaction module, and the sample memory content and teaching matching content of this teaching are stored in the user name, and then the stored sample memory content and teaching matching content are selected;

[0058] S4: Memory state change: The user's memory state is not exactly the same in different time periods. Collect the user's body signals, external conditions, and memory test feedback results after repeated memory training for at least one month and perform theoretical memory state analysis. Through the changes in memory state in different time periods, multiple memory state samples are formed;

[0059] S5: Effect evaluation: The data analysis module modifies the memory analysis algorithm by comparing the memory state analyzed by the theoretical memory state and the user memory state information entered during the teaching process, and records the external conditions and feelings associated with the memory state.

[0060] Preferably, the user's sample memory is the optimal memory state information associated with body signals, external conditions, emotional feelings and sleep feelings formed by the user performing multiple memory tests within a certain time range and saved in the form of video, audio, text and / or pictures.

[0061] Preferably, when the user's actual memory state, body signals, and external conditions are stored in the data analysis module in an associated form, the data analysis module 30 writes the associated relationship between the actual memory state, body signals, and external conditions, as well as the memory analysis algorithm, into the memory through a teaching process, thereby forming a personalized memory analysis algorithm for the user. The more times the user and the data analysis module 30 are taught through the data interaction module 10, the more accurate the memory analysis algorithm of the data analysis module 30 is. For example, the weight parameters of various information are adjusted according to the sample information input by the user for multiple times, so that the theoretical memory state of the memory analysis algorithm is as consistent or similar as possible with the actual memory state of the user.

[0062] The data analysis module 30 pre-configures the parameters of the memory analysis algorithm according to at least two extreme memory state information determined in the teaching process.

[0063] Preferably, during the teaching process, the data interaction module 10 applies stimulation information that can trigger the memory state to the user, and automatically collects and / or inputs the actual memory state information corresponding to the stimulation information by the user. At the same time, the data acquisition module 20 collects the user's body signals corresponding to the stimulation information. The data analysis module pre-configures the parameters of the memory analysis algorithm based on at least two of the actual memory state information and the body signals during the teaching process. Preferably, the data analysis module 30 corrects the theoretical memory state information corresponding to the body signal based on the actual memory state.

[0064] Preferably, the stimulus information includes information such as videos, pictures, text information and sounds that can cause the user to recall the response. The stimulus information causes the user to recall the response through the information such as specific events, specific scenes, memory training questions, etc. that it carries. For example, the video information records the user's continuous memory information of the occurrence and end of a certain event within a certain time range. After the data interaction module displays a certain segment, it determines whether the user can obtain detailed information about the entire event based on the segment by asking questions, or divides the user's memory state into levels based on the degree of completion of the entire event recall.

[0065] Preferably, the data analysis module 30 also forms a personalized memory state change curve based on the change trend of the actual memory state during the teaching process. For example, the memory state changes as follows: blurred consciousness → intermittent clear consciousness → normal consciousness but can only remember simple events → can remember some details of complex events → can remember all details of the entire event. Under normal circumstances, the user will not instantly switch from a blurred consciousness state to a normal memory state.

[0066] By continuously analyzing the theoretical memory state of the current user, the memory training device of the present invention can use the data analysis module 30 to determine the changing trend of the memory state. Preferably, the data analysis module or the data interaction module analyzes the changing trend of the actual memory state based on the body signal, the actual memory state information and / or the external condition information. For example, the user's current actual memory state is tested by the memory test module that communicates with the data analysis module in real time. The test can be performed in cycles and frequencies, for example, four sampling tests are performed in a month. During the test, the memory test module outputs test information to the test subject in the form of pictures, videos, audio and / or text and accepts the results of the test subject's feedback. By summarizing and analyzing the results of multiple tests within a certain time period, the user's memory state changing trend within the time period is obtained. Preferably, when the test subject has an extremely poor memory state in the first test, the data analysis module and / or the data interaction module tests the test subject's current mood, physical condition and sleep quality and adjusts the test subject's mood and sleep quality according to the test results, or trains the test subject's memory. Modulating the emotions of the test subjects can alleviate their loneliness through voice systems such as Siri, which can talk to the elderly.

[0067] Preferably, when the data interaction module 10 analyzes the actual memory state information within at least one continuous time interval and confirms that the changing trend of the actual memory state of the current user will exceed the critical value at a foreseeable time point, and / or when the data analysis module analyzes and confirms that the changing trend of the theoretical memory state of the current user will exceed the critical value at a foreseeable time point, the data analysis module sends an early warning prompt to the current user through the data interaction module. Specifically, the data interaction module 10 or the data analysis module 30 analyzes the changing trend of the actual memory state of the current user and confirms that the changing trend of the actual memory state will exceed the critical value at a foreseeable time point. At this time, the data analysis module 30 sends a variety of prompts such as vibration, sound, color change, etc. to the user through the data interaction module 10 to remind the user that there is an abnormality in the memory state and start memory training for the user.

[0068] Preferably, the data interaction module 10 stores or provides the user's feelings input in the form of text, voice, video and / or graphics in association with the corresponding automatically collected actual memory state to the data analysis module 30. Alternatively, the data interaction module 10 records the external conditions that trigger the user's actual memory state and stores or provides them to the data analysis module 30 in an associated form with the corresponding actual memory state. The data analysis module 30 analyzes the correlation between the user's specific memory state and the external relationship, and warns the triggering of the specific memory state based on the correlation. Preferably, the external conditions are, for example, weather conditions, noise index, the user's current mental state, etc.

[0069] For example, in rainy weather, the user is in an unfamiliar geographical location with a noisy living environment. The user's actual memory state is affected by the user's current emotional state and / or sleep state, causing the memory state to be lower than its normal level. At this time, the data acquisition module collects the user's body signals in real time. The data interaction module 10 or the data analysis module 30 associates and stores the external conditions, body signals and actual memory state at the same time, and records the memory state change trend within the previous limited time of the actual memory state and the memory state change trend within the next limited time. The data analysis module 30 analyzes the correlation between the user's extremely poor memory state and the external conditions, emotional state and sleep state, and warns of the occurrence of an extremely poor memory state based on the correlation.

[0070] Preferably, the data analysis module 30 stores the current memory state information input by the user and the external conditions related to the actual memory state information provided by the data interaction module 10 in a manner associated with each other, which facilitates information retrieval. The data interaction module 10 is configured to allow the user to retrieve the actual memory state information stored in the data analysis module 30 and / or the data interaction module 10 in a manner associated with the external conditions.

[0071] Preferably, the theoretical memory state change trend exceeding the critical value at a foreseeable time point is classified and stored in the database of the data analysis module in a manner that the user's input feeling, the actual memory state information, the body signal and the external conditions are all associated with each other, wherein the data analysis module forms a boundary trigger condition for the theoretical memory state change trend exceeding the critical value based on the above-mentioned associated data information in the database, and the data analysis module is configured to analyze the correlation between the user's actual memory state information and the boundary trigger condition, and make an early warning based on the correlation, improve the external conditions and / or the working mode of the memory training program. For example, the classification calculation results of memory can be divided into a healthy state, a sub-healthy memory state and a severe memory decline state. Among them, the sub-healthy memory state and the severe memory decline state can be divided into the scope of the memory state change trend exceeding the critical value. The data analysis module records the body signals, external conditions, user input feelings and actual memory state information in the sub-healthy memory state or the severe memory decline state. For example, based on multiple monitorings within a certain time period, during the first monitoring, the weather was sunny, the climate was warm, the sleep quality was poor, the mood was irritable, and the actual memory state was that the items were always forgotten when going out. During the second monitoring, the weather was sunny, the climate was warm, the sleep quality was good, the mood was irritable, and the actual memory state was clear thinking and quick response. Both were in the sub-healthy state of memory. It can be confirmed that irritability and poor sleep quality are one of the boundary triggering conditions that cause the trend of memory state change to exceed the critical value. The data analysis module analyzes and concludes that under the above conditions, the actual memory state is closely correlated with sleep quality and emotional state to a greater extent. Based on the above correlation, the data analysis module can give suggestions such as improving the living environment, adjusting the work and rest time, exercising appropriately, participating in community activities, etc. for the user's current situation. Based on the above method, the device of the present invention can learn the user's memory state adjustment characteristics, and then use a simple and effective method to guide the user's memory state in advance to avoid its trend of deteriorating state.

[0072] Memory state retrieval is an important means in the memory training process. The user can retrieve the actual memory state information stored in the data analysis module by retrieving the body signal that triggers a certain bad memory state. Preferably, the actual memory state information can be a certain inner feeling recorded by the user in the form of text, picture and / or video, or it can be sound, video or image information collected by the data interaction module 10. Through retrieval, the user can determine the external conditions, emotional state, sleep state and historical change records of the actual memory state that affect his or her memory state, thereby helping the user or caregiver to train his or her memory.

[0073] Preferably, the step in which the user inputs the actual memory status in the data interaction module 10 includes: the user selects the current external conditions, emotion type and level, sleep type and level, memory level in a point-and-click manner, and / or the user inputs the actual memory status change trend over a period of time in the form of text, voice, video or graphics.

[0074] Preferably, the step in which the data interaction module 10 automatically collects the user's actual memory status includes: the data interaction module 10 tests the user through the memory testing module and collects the actual memory status in the form of audio and / or video recording, and compares the collected actual memory status with the current external conditions, emotion type and level, sleep type and level, and memory level selected by the user in a point-by-point manner, thereby correcting the collected actual memory status.

[0075] In particular, the data interaction module 10 can provide actual memory state information as a search result in a manner related to external conditions. For example, by inputting and searching with the external condition "cloudy day-winter" or body signal, several related actual memory state change trends can be retrieved. The user can understand the influence of external conditions on himself and recall the situation at that time.

[0076] Although the data analysis module 30 can analyze the body signal and the actual memory state, they are all based on theoretical research. Each individual has a different personality. Some people like cloudy days and have better memory on cloudy days. Some people dislike cloudy days and have poorer memory on cloudy days. Therefore, a teaching process is required to correct the theoretical memory state so that the analysis of the data analysis module 30 can adapt to independent individuals. Preferably, the data analysis module 30 corrects the theoretical memory state determined by the body signal analysis collected by the data acquisition module according to the actual memory state information. The memory management configuration file composed of the corrected theoretical memory state is stored in the data interaction module in a manner that can be retrieved according to the body signal. Preferably, the corrected theoretical memory state can be retrieved by customizable keywords. Preferably, the retrieved theoretical memory state can deliver the retrieval results in a manner that is provided together with the approximate actual memory state. Providing the theoretical memory state together with the approximate actual memory state is conducive to users determining the desired retrieval information based on the retrieval results and improving the accuracy of information retrieval.

[0077] Preferably, the step of correcting the theoretical memory state according to the actual memory state includes: comparing the actual memory state with the theoretical memory state, and adjusting the theoretical memory state qualitatively and / or quantitatively, thereby generating a user memory profile related to the current user. Preferably, the user memory profile also takes into account the external conditions related to each actual memory state information. Such a configuration is conducive to eliminating the influence of unfavorable external conditions and guiding users to adapt to external conditions in an appropriate manner, and promoting their memory state to develop in a better direction under the condition of providing memory training or rehabilitation care. Specifically, when adjusting the theoretical memory state, personalized conditions related to the theoretical memory are set in the user memory profile. Personalized conditions include external conditions and body signals that match the user. When the user is in an extreme memory state and needs guidance, the user memory profile adjusts the nursing plan and memory training plan of the data interaction module based on the personalized conditions related to the theoretical memory state, or changes the user's external conditions to gradually adjust the user's memory state to change slowly.

[0078] Preferably, the display of the data interaction module 10 includes an external condition identifier, a body characteristic identifier, a memory state identifier and a feeling identifier. Among them, the changes of the external condition identifier, the body characteristic identifier, the memory state identifier and the feeling identifier are related to each other, and at least one external condition qualitatively and / or quantitatively changes the body characteristic. At least one body characteristic qualitatively and / or quantitatively triggers the change of the memory state identifier. At least one user feeling qualitatively and / or quantitatively triggers the change of the user's memory state. Preferably, when the user inputs the feeling, the data analysis module analyzes the actual memory state of the corresponding external conditions, body signals and feelings sent by the data interaction module and displays it as a memory state identifier. Preferably, the external condition identifier, the body characteristic identifier, the memory state identifier and the feeling identifier are displayed in real time in the data interaction module in a synchronously changing manner. The user can simultaneously view the external condition identifier, the body characteristic identifier, the memory state identifier and the feeling identifier in the data interaction module, as well as the changes of other identifiers caused by the change of one of the identifiers.

[0079] For example, the data interaction module 10 stores and displays external condition information, body signal information, actual memory state information and feeling information in a manner in which the mapping associations of external conditions, body signals, actual memory states and feelings gradually increase. Each external condition is associated with at least one body signal mapping. Each body signal is associated with at least one actual memory state mapping. Each actual memory state is associated with at least one feeling mapping.

[0080] Preferably, the data analysis module records the actual memory state and the theoretical memory state based on iterative analysis of at least one external condition, body signal and / or feeling at each moment. The mobile terminal 10 displays the change trend of external condition changes, body signal changes, feeling changes and actual memory state by displaying a marker with mapping association, reminds the user of the extreme change trend of the actual memory state by color change, sound and / or vibration, and / or displays suggestions for changing the actual memory state. Suggestions may be, for example: suggesting that the user do exercise, travel, visit relatives, go shopping to change external conditions or user feelings.

[0081] The data interaction module 10 displays the user's input feelings, the actual memory state associated therewith, the external conditions and the body signals in the form of a circular array composed of at least two circles, wherein each circle is divided into a plurality of spaces for recording information.

[0082] like Figure 3As shown, the data interaction module 10 stores and displays the user input feelings, the actual memory state associated therewith, the external conditions and the body signals in the form of a circular list composed of four circles. Among them. The circular list includes an outer circle mark and an inner circle mark. The inner circle mark includes a first inner circle mark 11 and a second inner circle mark 12 whose radius is greater than the first inner circle mark. The outer circle mark includes a first outer circle mark 13 and a second outer circle mark 14 whose radius is greater than the first outer circle mark.

[0083] Each space of the first inner circle mark 11 is used to store external conditions associated with the actual memory state. Each space of the second inner circle mark 12 is used to store physical characteristics associated with the actual memory state. Each space of the first outer circle mark 13 is used to store the actual memory state. Each space of the second outer circle mark 14 is used to store the user input feeling. Changes in each type of information may cause changes in other information.

[0084] For example, the noise index in the external conditions increases, and each space of the external condition identification is recorded and changed in color clockwise, and each space of the physical characteristic identification is recorded in clockwise and changes in color accordingly. After the noise index increases and exceeds the individual threshold, each space of the feeling identification is recorded in clockwise and changes in color accordingly. For example, the feeling input by the user is irritable level 2. The memory state identification changes based on the changes in external conditions, physical characteristics, feelings and their levels, resulting in a change in the color of the memory state identification. When the actual memory state of the user appears in an extreme state of no improvement or continuous decline within a period of time, the data interaction module 10 warns the user of the deterioration trend of the user's memory state in the form of sound, vibration or flashing, reminding the user to improve the external conditions, actively train the user's memory or guide the caregiver to provide appropriate rehabilitation care for the user. Preferably, the data analysis module 30 can send suggestions to the user through the data interaction module 10 to guide the change of external conditions, such as sports suggestions, music song suggestions, elderly activity suggestions, etc.

[0085] Example 2

[0086] This embodiment is a further improvement on Embodiment 1, and the repeated contents will not be repeated here.

[0087] The present embodiment provides a memory training method for the elderly, which at least includes the steps of: automatically collecting and / or inputting actual memory state information by the user. Collecting body signals by indirectly or directly contacting the user's body. Determining the theoretical memory state information corresponding to the body signal using a memory analysis algorithm. Wherein, the analysis and correction of the theoretical memory state change trend are completed based on the user's input feelings, real-time memory state information and / or external conditions. The theoretical memory state change trend that exceeds the critical value at a foreseeable time point is classified and stored in a manner that is associated with the user's input feelings, the actual memory state information, the body signal and the external conditions. Wherein, a boundary trigger condition for the theoretical memory state change trend exceeding the critical value is formed based on the above-mentioned mutually associated data information. Analyze the correlation between the user's actual memory state information and the boundary trigger condition, and make early warnings, push memory training plans and / or suggestions for improving external conditions based on the correlation.

[0088] Preferably, test information capable of obtaining the user's memory status information is applied to the user in the form of text, voice, video and / or graphics, and the change trend of the memory status is determined based on the feedback result of the test information. The teaching process of memory training is analyzed based on the boundary trigger condition and / or the change trend of the memory status, wherein the parameters of the memory analysis algorithm are pre-configured according to at least two extreme memory status information determined in the teaching process.

[0089] Preferably, test information capable of obtaining the user's memory status information is applied to the user in the form of text, voice, video and / or graphics, and test feedback information corresponding to the test information is automatically collected. At the same time, the user's body signals corresponding to the test feedback information, the user's input feelings and / or external conditions are collected. The parameters of the memory analysis algorithm are preconfigured based on the test feedback information, the external conditions, the user's input feelings and / or the body signals, and the theoretical memory status information is corrected.

[0090] Preferably, a user memory profile associated with the current user is generated based on continuously qualitatively and / or quantitatively adjusting the theoretical memory state information. The teaching process of memory training is analyzed based on the boundary trigger condition, the change trend of the memory state and / or the user memory profile.

[0091] Preferably, the external conditions that trigger the actual memory state of the user are recorded and stored in a form associated with the corresponding actual memory state, and the current memory state information input by the user and the external conditions are stored in a form associated with each other. After analyzing the correlation between the user's specific memory state and the external conditions, a memory training plan is pushed to trigger the specific memory state based on the correlation.

[0092] Preferably, the collected body signals are analyzed to determine the corresponding theoretical memory state. Stimulation information is output in the form of text, voice, video and / or graphics to change the external conditions and / or directly provide memory training to the user to improve the user's current memory state.

[0093] Preferably, the current memory state information input by the user and the external conditions related to the actual memory state information are stored in a manner associated with each other. The user retrieves the actual memory state information in a manner associated with the external conditions.

[0094] Preferably, the theoretical memory state determined based on the analysis of collected body signals is corrected according to the actual memory state information, and the memory management profile consisting of the corrected theoretical memory state is stored in a manner that can be retrieved according to the body signals.

[0095] Preferably, the actual memory state and the theoretical memory state are recorded based on iterative analysis of at least one external condition, body signal and / or feeling at each moment. The changing trends of the external condition changes, body signal changes, feeling changes and the actual memory state are displayed in the form of a marker with mapping association, and the user is reminded of the extreme changing trend of the actual memory state by color change, sound and / or vibration, and / or a suggestion for changing the actual memory state is displayed.

[0096] It should be noted that the above specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention description and its drawings are illustrative and do not constitute a limitation of the claims. The protection scope of the present invention is defined by the claims and their equivalents.

Claims

1. A user memory status early warning device, comprising at least a data analysis module, characterized in that: The data analysis module analyzes the correlation between the user's specific memory state and external conditions, and pushes a memory training plan to trigger the specific memory state based on the correlation. The user memory state warning system also includes a data interaction module. The data analysis module stores the current memory state information input by the user and the external conditions related to the actual memory state information provided by the data interaction module in a mutually associated manner. The data interaction module is configured to allow the user to retrieve the actual memory state information stored in the data analysis module and / or the data interaction module in a manner related to the external conditions. The data analysis module associates and stores the external conditions, body signals and actual memory state at the same time, and performs The memory state change trend within the previous limited time of the actual memory state and the memory state change trend within the next limited time are recorded, the data analysis module analyzes the correlation between the user's extremely poor memory state and the external conditions, emotional state and sleep state, and based on the correlation, warns of the initiation of the extremely poor memory state, the theoretical memory state change trend that exceeds the critical value at a foreseeable time point is classified and stored in the database of the data analysis module in a manner that is associated with the user's input feelings, the actual memory state information, body signals and external conditions, and the data analysis module forms a boundary trigger condition for the theoretical memory state change trend exceeding the critical value based on the above-mentioned associated data information in the database.

2. The user memory status early warning device according to claim 1, characterized in that: The data analysis module is configured to analyze the correlation between the user's actual memory status information and the boundary trigger condition, and to make early warnings based on the correlation, improve external conditions and / or memory training programs.

3. The user memory status early warning device according to claim 2, characterized in that: The user can retrieve the actual memory state information stored in the data analysis module by retrieving the body signal that triggers a certain bad memory state.

4. A user memory status early warning method, characterized in that: At least: Analyze the correlation between the user's specific memory state and external conditions, and push a memory training program to trigger the specific memory state based on the correlation. The current memory state information input by the user and the external conditions related to the actual memory state information are stored in a mutually associated manner, The user retrieves the actual memory state information in a manner related to external conditions, The external conditions, body signals and actual memory status at the same time are stored in association, and the memory status change trend within the previous limited time and the memory status change trend within the next limited time of the actual memory status are recorded. Analyze the correlation between the user's extremely poor memory state and the external conditions, emotional state and sleep state, and issue an early warning of the occurrence of the extremely poor memory state based on the correlation. The theoretical memory state change trend that exceeds the critical value at a foreseeable time point is classified and stored in a manner that is associated with the user's input feelings, the actual memory state information, body signals and external conditions. Based on the above-mentioned interrelated data information, a boundary trigger condition is formed when the trend of change in the theoretical memory state exceeds the critical value.

Images