A memory support system based on transcranial direct current stimulation
By combining transcranial DC stimulation and n-back working memory paradigm training, the electrical stimulation parameters are adjusted by using EEG signal acquisition and analysis, the problem of short-lasting working memory enhancement effect in the existing technology is solved, and the Memory effect continues and improves.
Patent Information
- Application Number
- CN202011609685.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-12-30
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2040-12-30
AI Technical Summary
The existing transcranial DC stimulation has the problem of not lasting effect in improving working memory performance, and cannot continue to improve the effect after being disconnected from the stimulus.
Combined with transcranial DC stimulation and n-back working memory paradigm training, through electrical stimulator, acquisition device and training device, the stimulation position, duration and current density are adjusted according to EEG signal acquisition and analysis, and combined with Baddeley's multi-component model setting training Pattern, build brain functional networks, adjust electrical stimulation parameters to improve the durability of memory effects.
It significantly improves the effect of working memory and extends it until the end of electrical stimulation, improving the persistence and effectiveness of memory.
Smart Images

Figure CN112604163B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of pattern recognition, and relates to an auxiliary memory system based on transcranial direct current stimulation, and in particular to an auxiliary memory system combining transcranial direct current stimulation with n-back working memory paradigm training. Background Art
[0002] Transcranial direct current stimulation (tDCS) is a non-invasive technique for regulating the activity of the cerebral cortex. It regulates the activity of neurons by applying a weak and stable direct current on the surface of the scalp. Transcranial direct current stimulation has two electrodes, namely the stimulating electrode and the reference electrode. According to the polarity of the stimulating electrode, it can be divided into: anodal stimulation, cathodal stimulation and sham stimulation. The current intensity is generally set to 1-2 mA, and the duration varies from 5 to 20 minutes. Unlike other non-invasive stimulation methods that use external suprathreshold stimulation to induce neuronal cell discharge, tDCS stimulation activates or inhibits the excitability of the cortex by regulating the subthreshold activation level of neuronal cells. When anodal tDCS stimulation is applied to the motor cortex area, it can depolarize the resting membrane potential of neurons, lower the activation threshold and enhance the excitability of the cortex, while cathodal tDCS stimulation hyperpolarizes the resting membrane potential of neurons, thereby increasing their activation threshold to inhibit the excitability of the cortex. tDCS stimulation can regulate the activity of N-methyl-D-aspartate receptors and calcium homeostasis, change the physiological environment of synapses and induce long-term stimulation aftereffects.
[0003] Working memory refers to a cognitive system that can temporarily store task-related information and process it. Working memory is closely related to many complex cognitive skills, such as reading comprehension, verbal comprehension, vocabulary learning, writing, logical reasoning, and complex learning. Among the many working memory models, the most influential is Baddeley's multi-component model, which includes the speech loop, visuospatial template, central executive system, and context buffer. The speech loop consists of two parts: speech storage and pronunciation control. The speech storage can keep speech information for about 2 seconds, and its items are all represented by speech structures; the pronunciation control can reactivate the speech that is about to disappear through silent reading. The visuospatial template is mainly responsible for the encoding of visual and spatial information, including visual elements related to color and shape and spatial elements related to position. Studies have shown that verbal text materials can activate the left dorsolateral prefrontal cortex more actively, while spatial information materials can more effectively activate the right dorsolateral prefrontal cortex. It can be seen that using different load materials to carry different forms of information can deepen working memory.
[0004] Initially, in 2005, Fregni found that anodal tDCS applied to the left dorsolateral prefrontal cortex can improve working memory performance. In addition, some studies have shown that the effect of anodal stimulation is related to the difficulty of working memory tasks, and the positive effects of working memory training using anodal tDCS can be transferred to similar unfamiliar training tasks. However, some studies have shown that the effect of anodal stimulation on working memory only stays during the stimulation process and cannot be extended after the stimulation is completed. These research results show that the use of transcranial direct current stimulation has a positive effect on the enhancement of working memory, but without a reasonable stimulation position and intensity, and with the training method, the memory effect is not good after the stimulation is removed. Summary of the invention
[0005] In view of the shortcomings of the prior art, the present invention proposes an auxiliary memory system based on transcranial direct current stimulation, including an electrical stimulator, a collection device and a training device. Different memory load materials are selected for different working memory sub-components, and the corresponding brain areas are activated to the maximum extent in combination with tDCS. The EEG information during the stimulation process is counted and analyzed, and the stimulation duration and current density of tDCS stimulation are changed to improve the effect of working memory, and the improvement effect is extended to after the electrical stimulation is removed.
[0006] An auxiliary memory system based on transcranial direct current stimulation comprises an electrical stimulator, a collection device and a training device.
[0007] The electrical stimulator is a transcranial direct current stimulator, which places the anode electrode and the cathode electrode in different brain regions of the user, and adjusts the position, duration and intensity of the stimulation according to the information collected by the acquisition device. The acquisition device collects the EEG signals of the user in different states through an EEG acquisition cap, constructs a brain function network, and provides a reference for the parameter adjustment of the electrical stimulator and the training device. The training device includes a computer screen and a speaker, and three different training modes are set according to Baddeley's multi-component model.
[0008] The three training modes are color-position training, semantic-sound training and audio-visual combined training. In the color-position training mode, the computer screen is divided into four blocks: upper, lower, left and right, and four colors of magenta, yellow, cyan and black are randomly presented in these four positions. In the semantic-sound training mode, a single word is randomly displayed on the screen, and the computer randomly emits high-frequency or low-frequency sounds. In the audio-visual combined training mode, any position of the upper, lower, left and right of the computer screen will randomly display a single word and a color, and the computer randomly emits high-frequency or low-frequency sounds.
[0009] Preferably, the frequency of the high-frequency sound is 2000 Hz, and the frequency of the low-frequency sound is 500 Hz.
[0010] The method of using the transcranial direct current stimulation-based memory assistance system comprises the following steps:
[0011] Step 1. Before electrical stimulation, use an EEG acquisition cap to collect the user's EEG signals in three training modes; use a bandpass filter with a bandwidth of 0.15-30 Hz to filter the collected EEG signals to filter out invalid frequency bands, and use independent component analysis to remove blinking and eye movement artifacts from the remaining valid frequency bands, and then use the last 200ms of data as the baseline value.
[0012] Step 2: After electrical stimulation, use an EEG acquisition cap to collect the user's EEG signals in the three training modes; subtract the potential of the EEG signal from the baseline value obtained in step 1 to obtain a new potential value. Superimpose the signals of all channels obtained in each mode and average them to obtain the relevant potential. Determine the distribution area of the relevant potential in different training modes.
[0013] Step 3: Construct brain function network and extract network features. The specific process is as follows:
[0014] Step 3.1, adopt the cross-correlation analysis method to calculate the cross-correlation coefficients between the EEG signals of each channel to obtain the connection coefficient matrix:
[0015]
[0016] Where x and y represent EEG time series, i and j represent the sampling nodes of EEG signals, and n represents the total number of sampling nodes. xy The value of is between 0 and 1, where a value of 0 indicates that the signals are completely uncorrelated and a value of 1 indicates that the signals are completely correlated.
[0017] Step 3.2: Select a threshold to convert the connection coefficient matrix into a binary matrix, obtain the topological structure of the brain functional network based on the binary matrix, and calculate the node degree, clustering coefficient and global efficiency of the sampled nodes.
[0018] The node degree of node i is:
[0019]
[0020] where h ij is an element in the binary matrix. When there is a connection between nodes i and j, h ij =1, otherwise h ij =0.
[0021] The clustering coefficient of node i is:
[0022]
[0023] where d ijis the number of connecting edges passed by the shortest communication path between nodes i and j, and V is the set of all nodes in the brain functional network.
[0024] The node degree values of all sampled nodes are calculated, and the area with larger node degree values is taken as the node concentration area.
[0025] Step 4: The overlapping area of the distribution area of the relevant potential determined in step 2 and the node concentration area determined in step 3 is used as the core brain area for training. When the core brain area is located in the left dorsolateral prefrontal cortex, the anode electrode of the electrical stimulator is placed at F3. When the core brain area is the right dorsolateral prefrontal cortex, the anode electrode of the electrical stimulator is placed at F4. In both cases, the cathode electrode of the electrical stimulator is placed in the supraorbital area on the opposite side of the core brain area. The positions of F3 and F4 refer to the international 10-20 electroencephalogram positioning system.
[0026] Step 5, place the cathode and anode electrodes at the positions determined in step 4, and train the user under the condition of electrical stimulation; then, perform another round of training on the user without electrical stimulation, and collect the EEG signals at this time.
[0027] Step 6: Based on the EEG signals collected in step 5 without electrical stimulation, a brain function network is constructed and features are extracted. The features are compared with the features of the previous training to obtain the changes in brain connectivity characteristics caused by the aftereffects of transcranial direct current stimulation. The stimulation duration and current density of the electrical stimulator during training are adjusted accordingly.
[0028] The present invention has the following beneficial effects:
[0029] 1. Combining transcranial direct current stimulation with a training method based on Baddeley's multi-component model setting can improve the effect of working memory during the stimulation process and extend the effect after the stimulation ends.
[0030] 2. Collect the user's EEG signals at different stages and construct a brain function network. Extract network features and compare them. Adjust the duration and current density of electrical stimulation according to the changes in features to improve the stimulation effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] Figure 1 A block diagram showing the principle of how to use the system;
[0032] Figure 2 There are three training modes in the embodiment; DETAILED DESCRIPTION
[0033] The present invention will be further explained below with reference to the accompanying drawings;
[0034] An auxiliary memory system based on transcranial direct current stimulation comprises an electrical stimulator, a collection device and a training device.
[0035] The electrical stimulator is a transcranial direct current stimulator, which places the anode electrode and the cathode electrode in different brain regions of the user, and adjusts the position, duration and intensity of the stimulation according to the information collected by the acquisition device. The acquisition device collects the EEG signals of the user in different states through an EEG acquisition cap, constructs a brain function network, and provides a reference for the parameter adjustment of the electrical stimulator and the training device. The training device includes a computer screen and a speaker, and three different training modes are set according to Baddeley's multi-component model.
[0036] like Figure 1 As shown, the method of using the system includes collecting EEG data before electrical stimulation, calculating event-related potentials, conducting electrical stimulation for a period of time, and analyzing the activated brain areas under different training modes to determine the electrode placement position of the electrical stimulator. Then, different working memory training is performed under transcranial direct current stimulation. The EEG signal features after each training are collected, and the transcranial direct current stimulation parameters in the next training are adjusted according to the feature changes.
[0037] Eight healthy adults were selected as users, including four males and four females. The subjects were aged 22 to 25 years old (average 23.8 years old), all right-handed, with normal vision, no history of neurological diseases, and similar educational levels. Before the experiment, the subjects did not take drugs or food that affect the central nervous system.
[0038] like Figure 2 As shown, the user is trained in working memory three times a week for a period of three weeks, the training of the first week is color-position training, the training of the second week is semantic-sound training, and the training of the third week is audio-visual combination training. The training material of color-position is: the computer screen is divided into four blocks of upper, lower, left and right, and magenta, yellow, cyan and black four colors are randomly presented in these four positions. The training material of semantic-sound is: a single word is randomly displayed on the screen, and the computer randomly emits high-frequency or low-frequency sound. The training of audio-visual combination is: a single word and a color will be randomly displayed at any position of the upper, lower, left and right of the computer screen, and the computer randomly emits high-frequency or low-frequency sound. The three trainings per week are set to three different difficulties of 0-back, 1-back and 2-back respectively.
[0039] The 0-back difficulty is to let the user judge whether the current material is the same as the first material; the 1-back difficulty is to let the user judge whether the current material is the same as the previous material adjacent to it; the 2-back difficulty is to let the user judge whether the current material is the same as the previous material separated by one position. The display time of each material is 3000ms, and the time interval between the appearance of two materials is 500ms.
[0040] In each working memory training, the user's EEG signals are first collected using a 64-lead EEG acquisition device, the relevant potentials are calculated and the brain function network is constructed. After the electrode position is determined, the user is trained under electrical stimulation. After the training under electrical stimulation is completed, the user is trained again without electrical stimulation to strengthen the aftereffects of the stimulation. The EEG signals after training are collected and compared with those before training. If the brain connectivity performance after stimulation is not significantly improved, consider increasing the density of the circuit and the duration of stimulation.
Claims
1. A memory assistance system based on transcranial direct current stimulation, characterized in that: It includes an electrical stimulator, a collection device, and a training device; The electrical stimulator is a transcranial direct current stimulator, which places anode electrodes and cathode electrodes in different brain regions of the user, and adjusts the location, duration and intensity of stimulation according to the information collected by the collection device; The acquisition device collects EEG signals of the user in different states through an EEG acquisition cap, constructs a brain function network, and provides a reference for adjusting parameters of the electrical stimulator and the training device; The training device included a computer screen and audio, and three different training modes were set up according to Baddeley's multi-component model; The three training modes are color-position training, semantic-sound training and audio-visual combined training; in the color-position training mode, the computer screen is divided into four blocks: upper, lower, left and right, and four colors, magenta, yellow, cyan and black, are randomly presented at these four positions; in the semantic-sound training mode, a single word is randomly displayed on the screen, and the computer randomly emits high-frequency or low-frequency sounds; in the audio-visual combined training mode, a single word and a color will be randomly displayed at any position of the upper, lower, left or right of the computer screen, and the computer randomly emits high-frequency or low-frequency sounds; The method of using the system is: Step 1: Before electrical stimulation, use an EEG acquisition cap to collect the user's EEG signals in three training modes; After preprocessing, the last 200ms of data were used as baseline values; Step 2: After electrical stimulation, use an EEG acquisition cap to collect the user's EEG signals in three training modes; Subtract the potential of the EEG signal from the baseline value obtained in step 1 to obtain a new potential value; superimpose and average the signals of all channels obtained in each mode to obtain the relevant potential; Determine the distribution area of relevant potentials under different training modes; Step 3: Take the cross-correlation analysis method to calculate the cross-correlation coefficients between the EEG signals of two channels to obtain the connection coefficient matrix, and then convert the connection coefficient matrix into a binary matrix to obtain the topological structure of the brain function network; then calculate the node degree, clustering coefficient and global efficiency of the sampling nodes, and take the area with larger node degree value as the node concentration area. The specific steps are as follows: Step 3.1, adopt the cross-correlation analysis method to calculate the cross-correlation coefficients between the EEG signals of each channel to obtain the connection coefficient matrix: Where x and y represent EEG time series, i and j represent sampling nodes of EEG signals, and n represents the total number of sampling nodes; R xy The value of is between 0 and 1, where 0 indicates that the signals are completely uncorrelated, and 1 indicates that the signals are completely correlated; Step 3.2, select a threshold to convert the connection coefficient matrix into a binary matrix, obtain the topological structure of the brain functional network based on the binary matrix, and calculate the node degree, clustering coefficient and global efficiency of the sampled nodes; The node degree of node i is: where h ij is an element in the binary matrix. When there is a connection between nodes i and j, h ij =1, otherwise h ij =0; The clustering coefficient of node i is: where d ij is the number of connecting edges passed by the shortest communication path between nodes i and j, and V is the set of all nodes in the brain functional network; Step 4: The overlapping area of the distribution area of the relevant potential determined in step 2 and the node concentration area determined in step 3 is used as the core brain area for training; when the core brain area is located in the left dorsolateral prefrontal cortex, the anode electrode of the electric stimulator is placed at F3; when the core brain area is the right dorsolateral prefrontal cortex, the anode electrode of the electric stimulator is placed at F4; in both cases, the cathode electrode of the electric stimulator is placed in the supraorbital area on the opposite side of the core brain area; the positions of F3 and F4 refer to the international 10-20 electroencephalogram positioning system; Step 5, placing the cathode and anode electrodes at the positions determined in step 4, and training the user under the condition of electrical stimulation; then, performing another round of training on the user without electrical stimulation, and collecting the EEG signals at this time; Step 6: Based on the EEG signals collected in step 5 without electrical stimulation, a brain function network is constructed and features are extracted. The features are compared with the features of the previous training to obtain the changes in brain connectivity characteristics caused by the aftereffects of transcranial direct current stimulation. The stimulation duration and current density of the electrical stimulator during training are adjusted accordingly.
2. A memory aid system based on transcranial direct current stimulation as claimed in claim 1, characterized in that: The frequency of high frequency sound is 2000Hz and the frequency of low frequency sound is 500Hz.
3. A memory assistance system based on transcranial direct current stimulation as claimed in claim 1, characterized in that: The method of preprocessing the extracted EEG signals in step 1 is: using a bandpass filter with a bandwidth of 0.15-30 Hz to filter the collected EEG signals, filter out invalid frequency bands, and use independent component analysis method to remove blinking and eye movement artifacts from the remaining valid frequency bands.
Citation Information
Patent Citations
Multi-lead correlation analysis electroencephalo-graph (EEG) feature extraction method
CN103258215A
Transcranial direct current stimulation device and data processing method
CN110585590A
