Personalized threshold determination method and application of detection indicators in SSVEP asynchronous system

Abstract

The present invention discloses a method for determining personalized thresholds of detection metrics in an SSVEP asynchronous system and its application. The method for determining personalized thresholds includes the following steps: obtaining the maximum correlation coefficient C of the spontaneous electroencephalogram signal of the user; obtaining real-time electroencephalogram data; calculating the maximum correlation coefficient ρ of the evoked electroencephalogram signal of the user max and the second largest correlation coefficient ρ 2nd ; calculating the real-time detection metric ε, where ε = (ρ 2nd - ρ max )(ρ max - C); determining whether the real-time detection metric ε is less than a preset threshold; in the case where the real-time detection metric ε is less than the preset threshold, determining whether the threshold is stable; and in the case where the threshold is unstable, calculating the metric threshold at the shortest data time to update the threshold. According to the method for determining personalized thresholds of the present invention, the user can use the asynchronous system with the highest efficiency, and at the same time, when the SSVEP recognition accuracy rate is the same, the false touch rate of the spontaneous signal is reduced.

Description

Technical Field

[0001] The present invention relates to a SSVEP asynchronous system, and in particular to a method for determining a personalized threshold in the SSVEP asynchronous system. Background Art

[0002] SSVEP (steady-state visual evoked potential) refers to the fact that when a fixed-frequency visual stimulus is received, the visual cortex of the human brain will produce a continuous response related to the stimulus frequency (the fundamental frequency and multiples of the stimulus frequency). Brain-computer interfaces based on SSVEP are widely used due to their high information transmission rate. Among them, asynchronous brain-computer interface systems are more user-friendly in actual use, so the research on asynchronous systems has gradually increased.

[0003] Due to the non-stationarity of EEG and the differences between users, asynchronous systems generally adopt the method of dynamic data length, that is, increasing the data length to increase the signal-to-noise ratio when the signal-to-noise ratio is poor. The dynamic change of data length is mainly achieved by constructing detection indicators and setting thresholds. When the detection indicator reaches the threshold, stop adding data for output, otherwise continue to add data. Regardless of the indicator used, the setting of the threshold is crucial. The usual method is to regard the threshold as a hyperparameter and calculate it using offline data.

[0004] The detection indicators in existing asynchronous systems are basically divided into two types: (1) Maximum correlation coefficient ρ max (2) The difference between the largest correlation coefficient and the second largest correlation coefficient ρ max -ρ 2nd Or the ratio ρ max / ρ 2nd ,The indicator threshold is mainly determined using offline data through grid search and other methods, and usually all users use the same threshold.

[0005] Both indicators in the existing methods do not consider spontaneous signals, and the detection logic of the two indicators is different and only considers some situations. In addition, the thresholds in the existing methods are only determined by offline data, and usually all users use the same threshold, without considering the differences caused by changes in the status of users, devices, etc. in actual use. Therefore, when using the same threshold, the asynchronous system may not operate at the maximum ITR (information transfer rate) of the actual user.

[0006] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0007] The object of the present invention is to provide a method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system, which allows a user to use the asynchronous system with maximum efficiency.

[0008] To achieve the above object, an embodiment of the present invention provides a method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system, which comprises the following steps:

[0009] Obtaining the maximum correlation coefficient C of the user's spontaneous EEG signal;

[0010] Obtain real-time EEG data;

[0011] Calculate the maximum correlation coefficient ρ of the user's evoked EEG signal max and the second largest correlation coefficient ρ 2nd ;

[0012] Calculate the real-time detection index ε, where ε=(ρ 2nd -ρ max )(ρ max -C);

[0013] Determine whether the real-time detection index ε is less than a preset threshold;

[0014] When the real-time detection index ε is less than the preset threshold, it is determined whether the threshold is stable; and when the threshold is unstable, the index threshold for the shortest data is calculated to update the threshold.

[0015] In one or more implementations, when the threshold is stable, the preset threshold is used as the final threshold.

[0016] In one or more embodiments, when the threshold is unstable, the threshold is calculated when the data is the shortest to update the threshold, including: gradually reducing the data until the shortest data can be correctly detected, calculating the index threshold when the data is the shortest, using this index threshold as the target to update the threshold through the gradient descent method, using the latest threshold as the threshold for the next detection, and repeatedly updating the index threshold until the index threshold is stable.

[0017] In one or more embodiments, the detection index when the data is the shortest is calculated. Then the threshold T is updated using the gradient descent method.

[0018]

[0019]

[0020] The constant α>0 is the learning rate.

[0021] In one or more embodiments, when the spontaneous signal of the user cannot be obtained, C is set to zero.

[0022] In one or more embodiments, the method for determining whether the threshold is stable includes: determining by setting a fixed threshold, when Stop updating when the value is less than the fixed threshold.

[0023] In one or more embodiments, the method for determining whether the threshold is stable includes: setting the number of updates, for example, to 20 times.

[0024] Another aspect of the present invention provides an electronic device, comprising: a processor; and a memory, wherein the memory stores instructions, and when the instructions are executed by the at least one processor, the at least one processor executes the personalized threshold determination method for detection indicators in the SSVEP asynchronous system as described above.

[0025] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for determining personalized thresholds of detection indicators in the SSVEP asynchronous system as described above.

[0026] Compared with the prior art, the personalized threshold determination method according to the embodiment of the present invention reduces the false touch rate of spontaneous signals while maintaining the same SSVEP recognition accuracy. In addition, personalized parameters are set for each user by updating parameters online, so that the system operates at the user's maximum information transmission rate (ITR). BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 is a flow chart of a method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system according to an embodiment of the present invention;

[0028] Figure 2 is another flow chart of a method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings, but it should be understood that the protection scope of the present invention is not limited by the specific implementation modes.

[0030] Unless explicitly stated otherwise, throughout the specification and claims, the term “comprise” or variations such as “include” or “comprising”, etc., will be understood to include the stated elements or components but not to exclude other elements or components.

[0031] When the human body is subjected to a visual stimulus of a fixed frequency flickering or changing pattern, the potential activity of the cerebral cortex will be modulated, thereby producing a continuous response related to the stimulus frequency (at the fundamental frequency and multiples of the stimulus frequency). This response has a periodic rhythm similar to that of visual stimulation, which is the steady-state visual evoked potential (SSVEP). The SSVEP signal is manifested in the EEG signal as a spectrum peak at the stimulus frequency and harmonics in the power spectrum. By analyzing and detecting the frequency corresponding to the spectrum peak, the stimulus source of the user's visual gaze can be detected, thereby identifying the user's intention.

[0032] An asynchronous system is a brain-computer interface system that is designed from the user's perspective. Unlike a synchronous system, an asynchronous system does not have a unified trigger clock. The initiative is in the hands of the user, who can issue control instructions at the required time. That is, only when the user wants to perform a task operation will the system control the external device according to the instruction. Compared with a synchronous system, an asynchronous system has obvious advantages in terms of ease of use. Users can issue instructions at any time, allowing users to perform standby operations. When the user does not issue an instruction, the system will wait silently until a new user operation is detected, providing a basis for practical operations based on SSVEP.

[0033] like Figure 1 to Figure 2 As shown, a method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system according to a preferred embodiment of the present invention includes the following steps S1-S6.

[0034] In step S1 , the maximum correlation coefficient C of the user's spontaneous EEG signal is obtained.

[0035] In an embodiment of the present invention, the maximum correlation coefficient C of the spontaneous signal is determined by collecting the spontaneous signal before the user officially uses it, and C can be set to 0 when there is no spontaneous signal. By collecting the spontaneous signal of the user and determining the maximum correlation coefficient C of the spontaneous signal, it can be ensured that the interference factor of the user's spontaneous signal is considered when inducing the EEG signal. In addition, since the maximum correlation coefficient C of the spontaneous signal of different users is different, the personalized threshold determination method of the detection index of the present invention can perform personalized threshold determination for different users, so that the detection result is more accurate.

[0036] In steps S2 and S3, real-time EEG data is obtained and the maximum correlation coefficient ρ of the user's induced EEG signal is calculated. max and the second largest correlation coefficient ρ 2nd .

[0037] In the embodiment of the present invention, the optimal detection index threshold of the user's induced EEG signal is calculated by real-time EEG data, while the threshold in the prior art is only determined by offline data, and usually all users use the same threshold, without considering the differences caused by the state changes of users, devices, etc. in actual use. Therefore, when the same threshold is used, the asynchronous system may not operate at the maximum ITR (information transfer rate) of the actual user. The specific correlation coefficient calculation method can be calculated by the canonical correlation analysis (CCA) method and other methods such as FBCCA, etc., which belongs to the prior art in this field and will not be repeated here.

[0038] In step S4, the real-time detection index ε is calculated, where ε = (ρ 2nd -ρ max )(ρ max -C).

[0039] In the embodiment of the present invention, the inventors have found through a large number of experiments and theoretical studies that the new detection index, which takes into account the operational relationship between the maximum correlation coefficient C of the user's spontaneous EEG signal and the maximum correlation coefficient and the second largest correlation coefficient of the induced EEG signal, can more accurately detect the user's control intention. The detection index ε is usually effective when it is negative, that is, it requires that the maximum correlation coefficient of the induced signal is greater than the maximum correlation coefficient of the spontaneous signal.

[0040] In steps S5 and S6, it is determined whether the real-time detection index ε is less than a preset threshold; and when the real-time detection index ε is less than the threshold, it is determined whether the threshold is stable;

[0041] In the embodiment of the present invention, the idea of ​​determining the personalized threshold of the user is: presetting a smaller threshold to ensure that the detection accuracy is close to 100%. In the case of a stable threshold, the preset threshold is used as the final threshold.

[0042] In step S7, when the threshold is unstable, the index threshold for the shortest data is calculated to update the threshold.

[0043] When the user's detection index reaches the preset threshold, the detection result is equivalent to the actual label, and the data is gradually reduced until the shortest data is available for correct detection. The index threshold at the shortest data is calculated, and this index threshold is used as the target to update the threshold through the gradient descent method. The latest threshold is used as the threshold for the next detection, and the index threshold is repeatedly updated until the index threshold is stable.

[0044] In one or more embodiments, the detection index when the data is the shortest is calculated. Then the threshold T is updated using the gradient descent method.

[0045]

[0046]

[0047] The constant α>0 is the learning rate.

[0048] In one or more embodiments, the method for determining whether the threshold is stable includes: determining by setting a fixed threshold, when Stop updating when the value is less than the fixed threshold.

[0049] In one or more embodiments, the method for determining whether the threshold is stable includes: setting the number of updates, for example, to 20 times.

[0050] Another aspect of the present invention provides an electronic device, comprising: a processor; and a memory, wherein the memory stores instructions, and when the instructions are executed by the at least one processor, the at least one processor executes the personalized threshold determination method for detection indicators in the SSVEP asynchronous system as described above.

[0051] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for determining personalized thresholds of detection indicators in the SSVEP asynchronous system as described above.

[0052] The personalized threshold determination method of the detection index in the SSVEP asynchronous system of the present invention takes into account the maximum correlation coefficient of the user's spontaneous EEG signal, so that the detection result is more accurate. In addition, the threshold is gradually updated through the user's data when using online, and finally the optimal threshold of the user is reached.

[0053] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0054] The foregoing description of specific exemplary embodiments of the present invention is for the purpose of illustration and demonstration. These descriptions are not intended to limit the present invention to the precise form disclosed, and it is clear that many changes and variations can be made based on the above teachings. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the present invention and its practical application, so that those skilled in the art can realize and utilize various different exemplary embodiments of the present invention and various different selections and changes. The scope of the present invention is intended to be limited by the claims and their equivalents.

Claims

1. A method for determining personalized thresholds of detection indicators in an SSVEP asynchronous system, characterized in that: The following steps are involved: Obtaining the maximum correlation coefficient C of the user's spontaneous EEG signal; Obtain real-time EEG data; Calculate the maximum correlation coefficient ρ of the user's evoked EEG signal max and the second largest correlation coefficient ρ 2nd ; Calculate the real-time detection index ε, where ε=(ρ 2nd -ρ max )(ρ max -C); Determine whether the real-time detection index ε is less than a preset threshold; When the real-time detection index ε is less than the preset threshold, determining whether the threshold is stable; as well as When the threshold is unstable, the indicator threshold with the shortest data is calculated to update the threshold.

2. The personalized threshold determination method according to claim 1, characterized in that: When the threshold is stable, the preset threshold is used as the final threshold.

3. The personalized threshold determination method according to claim 1, characterized in that: When the threshold is unstable, the threshold is calculated when the data is the shortest to update the threshold, including: gradually reducing the data until the shortest data can be correctly detected, calculating the index threshold when the data is the shortest, using this index threshold as the target to update the threshold through the gradient descent method, using the latest threshold as the threshold for the next detection, and repeatedly updating the index threshold until the index threshold is stable.

4. The personalized threshold determination method according to claim 3, characterized in that: Calculate the detection index when the data is the shortest Then the threshold T is updated using the gradient descent method. The constant α>0 is the learning rate, T n Indicates the threshold value for the nth shortest data.

5. The personalized threshold determination method according to claim 1, characterized in that: When the spontaneous signal from the user cannot be obtained, C is set to zero.

6. The personalized threshold determination method according to claim 4, characterized in that: The method for judging whether the threshold is stable includes: judging by setting a fixed threshold. Stop updating when the value is less than the fixed threshold.

7. The personalized threshold determination method according to claim 4, characterized in that: The method of judging whether the threshold is stable includes: setting the number of updates.

8. An electronic device, characterized in that: include: processor; as well as A memory storing instructions, which, when executed by the processor, causes the processor to execute the personalized threshold determination method for detection indicators in the SSVEP asynchronous system as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the method for determining a personalized threshold value of a detection indicator in an SSVEP asynchronous system according to any one of claims 1 to 7.

Images