A method for measuring critical flicker fusion frequency based on electroencephalogram

Abstract

The present invention relates to the field of CFF measurement technology, and in particular to a method for measuring critical flicker fusion frequency based on electroencephalogram. The method includes calibration; visual stimulation; obtaining electroencephalogram signals for filtering and denoising to obtain target EEG signals; analyzing target EEG signal segments to adjust stimulation parameters; analyzing timestamps according to CFF feedback data to obtain critical flicker fusion frequency values. The present invention precisely positions the six electrodes on the EEG head-mounted device in the visual center area of ​​the subject's head, and starts an LED light source to visually stimulate the subject. The EEG acquisition device records the subject's electroencephalogram signals under visual stimulation in real time, and the collected electroencephalogram signals are filtered and denoised by a signal processing unit to accurately extract features related to visual stimulation for analysis, thereby dynamically adjusting the visual stimulation parameters, improving the effectiveness of visual stimulation, and increasing the accuracy of the detection results.

Description

Technical Field

[0001] The invention relates to the technical field of CFF measurement, and in particular to a method for measuring critical flicker fusion frequency based on electroencephalogram. Background Art

[0002] Critical Flicker Fusion Frequency (CFF) refers to the lowest frequency at which the human visual system perceives flickering light as stable light. When the flicker frequency of light exceeds the critical flicker frequency, the eye will perceive it as stable light. CFF reflects the temporal resolution limit of the visual system and the level of optic nerve function, and is highly sensitive to retinal and optic nerve functional lesions. When patients have fundus diseases such as macular degeneration, optic neuropathy, glaucoma, and diabetic retinopathy, the CFF measurement value will decrease, that is, damage to the visual conduction pathway will lead to a decrease in CFF. The effect of refractive medium turbidity on the CFF value is limited, because refractive medium turbidity mainly affects the spatial resolution of the visual system rather than the temporal resolution.

[0003] The cones in the retina are mainly L-cone, M-cone and S-cone, which have different sensitivities to different wavelengths of light (red, green and blue), and have the highest density in the macular area (where vision is most acute). CFF detection of different colors of light can provide a variety of visual function information. However, since the CFF examination is a subjective examination, its results depend on the patient's subjective perception of flickering light, so there are some inherent defects and limitations. The patient's cognitive level, reaction ability, fatigue level and concentration level may significantly affect the test results. For example, when the cognitive level is low, the attention is not focused, the visual task is long or the body is tired, the patient may not be able to accurately judge the frequency of flickering, resulting in reduced reliability of the measurement results.

[0004] EEG (electroencephalogram) is a method of recording brain activity using electrophysiological indicators. When the brain is active, the postsynaptic potentials of a large number of neurons are formed after being summed up. It records the changes in electric waves during brain activity and is a general reflection of the electrophysiological activity of brain nerve cells in the cerebral cortex or scalp surface. In ophthalmic examination technology, visual pathway function can be evaluated by recording the electrophysiological response of the brain to visual stimulation. Traditional EEG examinations use 12 electrodes, while in visual stimulation examinations, the main focus is on the 6 electrodes of the visual center. Although EEG provides objective electrophysiological data in receiving EEG responses, it also has certain limitations. When measuring visual stimulation, it will be disturbed in the case of turbidity of the refractive medium, affecting the accuracy of the results. Therefore, if CFF is combined with EEG, that is, by recording the EEG signal of the visual center during visual stimulation to form a stimulation-reception loop, the influence of subjective factors on CFF results can be reduced, and the objectivity and accuracy of CFF measurement can be improved, which is of great significance for visual function evaluation.

[0005] The patent document with the Chinese patent authorization announcement number CN109938755B discloses a method for measuring the critical value of flicker fusion frequency based on digital set recognition, wherein N Japanese-shaped digital frames are arranged in a background light area flashing within a set frequency, and a name is assigned to each stroke of the Japanese-shaped digital frame; ten groups of different frequency combinations are set to be applied to the N Japanese-shaped digital frames respectively; a stroke of the Japanese-shaped digital frame is assigned to each frequency in each group of different frequency combinations; for each Japanese-shaped digital frame in the N Japanese-shaped digital frames, one group is randomly selected from the ten groups of different frequency combinations; the value of each frequency in each group of the N groups of frequency combinations is determined; the subject recognizes the flashing numbers on the N Japanese-shaped digital frames in the background light area within the set frequency range; and the range of the subject's flicker fusion frequency critical value or the final flicker fusion frequency critical value is determined according to the result of the subject's recognition; it can be seen that the existing method for measuring the critical value of flicker fusion frequency lacks a measurement method that combines CFF with EEG, fails to analyze the EEG signal, and fails to adjust the stimulation frequency according to the analysis result, so that the measurement result is affected by subjectivity and is inaccurate. Summary of the invention

[0006] To this end, the present invention provides a method for measuring critical flicker fusion frequency based on electroencephalogram, so as to overcome the problem in the prior art that the measurement process of the flicker fusion frequency critical value is easily affected by the subjective factors of the subject, lacks objective analysis of EEG signals, and adjusts the stimulation frequency parameters according to the analysis results, resulting in low measurement accuracy.

[0007] To achieve the above object, the present invention provides a method for measuring critical flicker fusion frequency based on electroencephalogram, comprising:

[0008] Calibrate the LED light source and the EEG headset, adjust the EEG headset to the subject's head, turn on the signal processing unit, and connect the LED light source and the EEG headset to the signal processing unit via Bluetooth or WiFi;

[0009] An initial flickering frequency is set in the LED light source, and the LED light source is started to provide visual stimulation to the subject, wherein the visual stimulation process is to emit a light beam at the initial flickering frequency or the current flickering frequency after being adjusted upward, and a trigger instruction is generated at each time when the visual stimulation occurs to send a trigger signal to the EEG acquisition device;

[0010] The initial flickering frequency is adjusted upward multiple times with a standard adjustment frequency to update the current flickering frequency, and the visual stimulation occurs at the time corresponding to each time the light beam is emitted;

[0011] Acquire the EEG signal of the subject under visual stimulation in real time through the EEG acquisition device, filter and denoise the EEG signal, and obtain the target EEG signal;

[0012] Segmenting the target EEG signal into a plurality of target EEG signal segments, averaging any of the target EEG signal segments to generate an ERP waveform, marking each relevant component in the ERP waveform, and extracting key features in the ERP waveform, wherein the relevant components include a P1 wave, an N1 wave, and a P2 wave, and the key features include a peak value, an amplitude, and a delay;

[0013] Analyze each relevant component in any target EEG signal segment to determine whether to adjust the standard adjustment frequency;

[0014] When the signal peak in the target EEG signal segment is within the peak threshold range, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in a data feature library to obtain a real-time constant brightness feature signal and a real-time flicker feature signal, so as to construct a perceived flicker data set and a perceived constant brightness data set;

[0015] Analyzing the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the perceived constant brightness data set according to the CFF feedback data to obtain a critical flicker fusion frequency value;

[0016] Wherein, the CFF feedback data includes the CFF feedback time;

[0017] The start time of the real-time flicker characteristic signal is obtained to obtain a constant brightness timestamp.

[0018] Furthermore, filtering and denoising the EEG signal includes:

[0019] Intercepting the EEG signal according to the set standard frequency band range, acquiring the EEG signal falling within the standard frequency band range, and obtaining the original EEG signal;

[0020] The electrooculographic artifacts and electromyographic artifacts of the original EEG signal are identified and deleted to obtain a target EEG signal.

[0021] Further, identifying the electrooculographic artifacts and electromyographic artifacts of the original EEG signal and deleting them includes:

[0022] Acquiring the real-time frequency of the original EEG signal for spectrum analysis to divide the original EEG signal into different frequency bands;

[0023] When the real-time frequency does not fall within the first standard frequency band or the second standard frequency band, the artifact occurrence area in the original EEG signal is identified, and the artifact occurrence area is focused and analyzed to identify the electrooculographic artifacts and electromyographic artifacts for deletion.

[0024] Furthermore, obtaining the real-time frequency of the original EEG signal for spectrum analysis includes: determining the real-time frequency according to the first standard frequency band and the second standard frequency band,

[0025] If the real-time frequency is in the first standard frequency band, the original EEG signal is decomposed into a plurality of independent components or principal components and analyzed, components related to electrooculographic artifacts are identified, and deleted from the original signal;

[0026] If the real-time frequency is in the second standard frequency band, the signal is processed using the threshold method or wavelet transform method. When the signal amplitude exceeds the set threshold or shows abnormality in a specific wavelet coefficient, it is judged as an electromyographic artifact and deleted.

[0027] If the real-time frequency is neither in the first standard frequency band nor in the second standard frequency band, the original EEG signal is segmented according to the preset acquisition time into several original EEG signal segments, and a focused analysis is performed on the artifact occurrence area in any of the original EEG signal segments to identify the electrooculographic artifacts and electromyographic artifacts and delete them.

[0028] Further, obtaining the P1 wave in any target EEG signal segment for analysis includes:

[0029] Obtain the P1 wave in any target EEG signal segment, draw a curve of the P1 wave changing over time, and obtain a waveform change curve;

[0030] Determine the peak value of the waveform change curve,

[0031] When the peak value of the real-time signal is greater than the maximum value of the peak threshold range, the amplitude of the N1 wave or the P2 wave is obtained for determination;

[0032] When the peak value of the real-time signal is within the peak threshold range, the standard adjustment frequency is not adjusted, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in a data feature library;

[0033] When the signal peak is less than the minimum value of the peak threshold range, emotion recognition is performed on the target EEG signal segment according to the established emotion recognition model to obtain a recognition result, and it is determined whether to increase the standard adjustment frequency to the second corrected adjustment frequency according to the recognition result.

[0034] Further, the recognition results include anxious emotions, normal emotions, and tired emotions;

[0035] If anxiety is identified, the standard adjustment frequency is reduced to the third modified adjustment frequency;

[0036] If a normal emotion is identified, no adjustment is made to the standard adjustment frequency;

[0037] If fatigue is detected, the standard adjustment frequency is increased to the second corrected adjustment frequency.

[0038] Further, obtaining the amplitude of the N1 wave or the P2 wave for determination includes:

[0039] When the amplitude of the N1 wave or the P2 wave is less than the corresponding amplitude threshold, the standard adjustment frequency is increased to the first correction adjustment frequency;

[0040] When the amplitude of the N1 wave or the P2 wave is greater than the corresponding amplitude threshold, the P1 wave in the next adjacent target EEG signal segment is obtained, an adjacent continuous waveform curve is drawn, and the signal peak in the adjacent continuous waveform curve is obtained to continue the determination.

[0041] Furthermore, matching the real-time characteristic signal in a data characteristic library includes:

[0042] Use the decision tree to match the real-time feature signal with each standard signal in the data feature library one by one to obtain the matching result;

[0043] According to the matching results, the real-time characteristic signal is labeled and classified into a real-time constant light characteristic signal or a real-time flashing characteristic signal;

[0044] The real-time constant brightness feature signal marked in the matching result is stored in a perception constant brightness data set, and the real-time flicker feature signal marked in the matching result is stored in a perception flicker data set;

[0045] The perceived constant brightness data set includes a real-time constant brightness characteristic signal and a corresponding label, and the perceived flicker data set includes a real-time flicker characteristic signal and a corresponding label.

[0046] Further, analyzing the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the perceived constant brightness data set according to the CFF feedback data includes:

[0047] The actual time interval between the CFF feedback moment and the constant light timestamp is obtained. When the actual time interval is within the standard time interval range, the key feedback is determined to be normal, and the critical flicker fusion frequency value is obtained according to the key feedback.

[0048] When the actual time interval is less than the minimum value of the standard time interval range, a key error prompt is given, the current flashing frequency corresponding to the constant light timestamp is obtained, and the target perception result is obtained;

[0049] When the actual time interval is greater than the maximum value of the standard time interval range, the current flashing frequency corresponding to the constant light timestamp is obtained to obtain the target perception result.

[0050] Furthermore, the process of segmenting the target EEG signal is as follows:

[0051] Based on the stimulation time, the target EEG signal is extracted according to the preset stimulation duration to obtain the target EEG signal segment.

[0052] Compared with the prior art, the beneficial effects of the present invention are that by ensuring that the LED light source and the portable EEG headset are calibrated to provide accurate stimulation and recording, by adjusting the EEG headset to the subject's head to ensure that the six electrodes are accurately positioned in the visual center area, by starting the LED light source and showing the subject a flashing light to provide visual stimulation to the subject, the EEG acquisition device records the subject's EEG signal under visual stimulation in real time, and because the EEG data will be affected by external factors such as the subject's eye movement and muscle movement during the acquisition process, the collected EEG signal is filtered and denoised by the signal processing unit to accurately extract features related to the visual stimulation as real-time feature signals, and by analyzing the real-time feature signals, the real-time features are extracted from the EEG signal by using ERP analysis, and then the visual stimulation parameters are dynamically adjusted to improve the effectiveness of the visual stimulation and increase the stability of the detection results.

[0053] Furthermore, by filtering and denoising the EEG signal to extract features related to visual stimulation, the accuracy and reliability of signal acquisition are improved, while the feasibility of subsequent waveform recognition is increased.

[0054] Furthermore, by obtaining the actual time interval between the CFF feedback moment and the constant light timestamp and comparing it with the standard time interval, the state of the key feedback can be effectively judged. When the actual time interval is greater than the maximum value of the standard time interval, it means that the subject did not press the key in time. At this time, the perception result is obtained according to the EEG signal to avoid the influence of subjective factors on the measurement, thereby improving the detection accuracy. When the actual time interval is less than the minimum value of the standard time interval, it means that the subject pressed the key too quickly. At this time, a key error will be prompted to avoid distortion of the test data caused by abnormal reactions, ensuring that the reaction time of the subject during the test is within a reasonable range. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of a method for measuring critical flicker fusion frequency based on electroencephalogram in an embodiment of the present invention;

[0056] Figure 2 It is a schematic diagram of the structure of a CFF detection system according to an embodiment of the present invention;

[0057] Figure 3 A logic decision diagram for determining whether to adjust the standard adjustment frequency according to an embodiment of the present invention;

[0058] Figure 4 A schematic diagram of logic determination for adjusting the standard adjustment frequency according to an embodiment of the present invention;

[0059] In the figure: 1-LED light source, 2-EEG headset, 3-electrode, 4-signal processing unit, 5-feedback module. DETAILED DESCRIPTION

[0060] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0061] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0062] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0063] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0064] See also Figure 1 As shown, it is a flow chart of a method for measuring critical flicker fusion frequency based on electroencephalogram in an embodiment of the present invention. The present invention provides a method for measuring critical flicker fusion frequency based on electroencephalogram, comprising:

[0065] Step S1, calibrating the LED light source and the EEG headset, adjusting the EEG headset to the subject's head, turning on the signal processing unit, and connecting the LED light source and the EEG headset to the signal processing unit via Bluetooth or WiFi;

[0066] Step S2, setting an initial flickering frequency in the LED light source, starting the LED light source, to provide visual stimulation to the subject, wherein the visual stimulation process is to emit a light beam at the initial flickering frequency or the current flickering frequency after being adjusted upward, and generating a trigger instruction at each time when the visual stimulation occurs to send a trigger signal to the EEG acquisition device;

[0067] The initial flickering frequency is adjusted upward multiple times with a standard adjustment frequency to update the current flickering frequency, and the visual stimulation occurs at the time corresponding to each time the light beam is emitted;

[0068] Step S3, acquiring the EEG signal of the subject under visual stimulation in real time through an EEG acquisition device, filtering and denoising the EEG signal to obtain a target EEG signal;

[0069] Step S4, segmenting the target EEG signal into a plurality of target EEG signal segments, averaging any of the target EEG signal segments to generate an ERP waveform, marking each relevant component in the ERP waveform, and extracting key features in the ERP waveform, wherein the relevant components include P1 wave, N1 wave and P2 wave, and the key features include signal peak value, signal amplitude and signal delay;

[0070] The process of segmenting the target EEG signal is as follows:

[0071] Taking the stimulation time as the benchmark, extract the target EEG signal with the preset stimulation duration to obtain the target EEG signal segment;

[0072] Step S5, analyzing each relevant component in any target EEG signal segment to determine whether to adjust the standard adjustment frequency;

[0073] When the signal peak in the target EEG signal segment is within the peak threshold range, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in the data feature library to obtain a real-time constant brightness feature signal and a real-time flicker feature signal, so as to construct a perceived flicker data set and a perceived constant brightness data set;

[0074] Step S6, analyzing the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the perceived constant brightness data set according to the CFF feedback data to obtain a critical flicker fusion frequency value;

[0075] Wherein, the CFF feedback data includes the CFF feedback time;

[0076] The start time of the real-time flicker characteristic signal is obtained to obtain a constant brightness timestamp.

[0077] In this embodiment, before starting the measurement, the machine automatically calculates the required cursor size, light intensity, and measurement distance, and adjusts them. Finally, the doctor manually checks whether it is directly in front of the cornea. The subject holds an operating lever with a button on it. The subject uses the operating lever to provide feedback on whether he perceives flickering / constant light. Each time the machine adjusts the frequency, it will emit a "beep" sound. After the patient hears the "beep" sound, if he perceives flickering, he will short press the operating lever button. If he perceives constant light, he will long press the operating lever button. During the measurement, the signal processing unit associates the CFF feedback data and analyzes the collected EEG signals to avoid the influence of the subject's subjective factors on the test results, so as to accurately identify the perceived results and the non-perceived results and obtain the critical flicker fusion frequency value. In the measurement, the flicker frequency of the visual stimulus is controlled from 0 Hz to 2 Hz. The frequency is adjusted upward from 0.5Hz to 1.5Hz, and during the upward adjustment, feedback results are obtained. When feedback of not perceiving flicker is obtained, the frequency is adjusted downward by 1.5Hz until perceiving feedback results are obtained again, and then the frequency is adjusted upward by 1Hz until feedback of not perceiving flicker is obtained. Then the current flicker frequency is obtained and used as the target perception result. The current flicker frequency is continuously verified, that is, the frequency is adjusted downward by 1Hz once to obtain the first verification perception result, and the frequency is adjusted upward by 1Hz once to obtain the second verification perception result. The actual measured frequency averages corresponding to the target perception result, the first verification perception result, and the second verification perception result are taken to obtain the critical flicker fusion frequency value.

[0078] The preset stimulation duration set in this embodiment represents a period of time after the stimulation occurs, and the important stage of visual information processing is involved within the preset stimulation duration. Generally, the preset stimulation duration is set in the range of 200 milliseconds to 600 milliseconds to effectively capture the brain's response pattern to visual stimulation; the initial flickering frequency is set to 0 Hz.

[0079] ERP waveform (event-related potential) refers to the potential change in the electroencephalogram (EEG) signal caused by a specific event (light stimulation); the ERP waveform is obtained by averaging the EEG signals recorded in multiple trials, and can reflect the brain's temporal and potential response to specific stimuli.

[0080] By ensuring that the LED light source and the portable EEG headset are calibrated to provide accurate stimulation and recording, by adjusting the EEG headset to the subject's head, ensuring that the six electrodes are accurately positioned in the visual center area, by starting the LED light source, showing the subject a flashing light to provide visual stimulation to the subject, since the accuracy of the measurement results may be affected by the subject's subjective factors, therefore, during the test process, the subject's EEG signal is collected and analyzed to obtain a more objective and more accurate test result, and when the EEG acquisition device records the subject's EEG signal under visual stimulation in real time, since the EEG data will be affected by external factors such as the subject's eye movement and muscle movement during the acquisition process, therefore, the collected EEG signal is filtered and denoised by the signal processing unit to accurately extract features related to the visual stimulation as real-time feature signals, and by analyzing the real-time feature signals, the real-time features are extracted from the EEG signal by using ERP analysis, and then the visual stimulation parameters are dynamically adjusted to improve the effectiveness of the visual stimulation and increase the stability of the detection results.

[0081] See also Figure 2 As shown, it is a schematic diagram of the structure of the CFF detection system according to an embodiment of the present invention;

[0082] The present invention also provides a CFF detection system applied to the method for measuring critical flicker fusion frequency based on electroencephalogram, comprising an LED light source 1, an EEG acquisition device, an EEG head-mounted device 2, electrodes 3, a signal processing unit 4 and a feedback module 5, wherein:

[0083] LED light source 1, used for visually stimulating the subject by emitting a light beam;

[0084] An EEG acquisition device, which is connected to the LED light source, and is used to acquire the EEG signal of the subject under visual stimulation in real time, and to filter and denoise the EEG signal with a signal processing unit to obtain a target EEG signal; the EEG acquisition device can also acquire the EEG signal with a preset acquisition time and transmit it to the signal processing unit each time the trigger signal is received;

[0085] The EEG acquisition device includes an EEG headset 2 and six electrodes 3 disposed on the EEG headset and located in the visual center area, wherein the EEG headset is used to fix and connect the electrodes, and the electrodes are used to capture brain electrical signals;

[0086] Feedback module 5, used to generate CFF feedback data according to the triggering result of the key by the subject;

[0087] The signal processing unit 4 is connected to the EEG acquisition device and the feedback module, and is used to process and analyze the EEG signal, adjust the standard adjustment frequency according to the analysis result, and obtain a real-time characteristic signal;

[0088] The critical flicker fusion frequency value calculation unit is used to analyze the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the perceived constant brightness data set according to the CFF feedback data to obtain the critical flicker fusion frequency value.

[0089] Specifically, in step S3, filtering and denoising the EEG signal includes:

[0090] Step S31, intercepting the EEG signal according to the set standard frequency band range, acquiring the EEG signal falling within the standard frequency band range, and obtaining the original EEG signal;

[0091] Step S32, identifying the electrooculographic artifacts and electromyographic artifacts of the original EEG signal and deleting them to obtain a target EEG signal.

[0092] In this embodiment, a 4th-order Butterworth bandpass filter is used to obtain and determine that the EEG signal falls within the standard frequency band range to remove low-frequency drift and high-frequency noise in the EEG signal. The set standard frequency band range represents the effective frequency band range of the EEG signal, which is set in the range of 0.1Hz-50Hz to retain the useful information in the EEG signal. The set standard frequency band range can be selected and adjusted according to the accuracy requirements and the actual filter specifications. The filter can choose to use Butterworth filter, Chebyshev filter, etc., which is not specifically limited here.

[0093] By filtering and denoising the EEG signal to extract features related to visual stimulation, the accuracy and reliability of signal acquisition are improved, while the feasibility of subsequent waveform recognition is increased.

[0094] Specifically, in step S32, identifying the electrooculographic artifacts and electromyographic artifacts of the original EEG signal and deleting them includes:

[0095] Step S321, obtaining the real-time frequency of the original EEG signal to perform spectrum analysis to divide the original EEG signal into different frequency bands;

[0096] Step S322, when the real-time frequency does not fall within the first standard frequency band or the second standard frequency band, identifying the artifact occurrence area in the original EEG signal, and performing focused analysis on the artifact occurrence area to identify electrooculographic artifacts and electromyographic artifacts for deletion.

[0097] In this embodiment, electrooculographic artifacts refer to EEG signals caused by the subject's eye movements, such as blinking, eye movement, and nystagmus; electromyographic artifacts refer to EEG signals caused by the subject's muscle activities, such as head movements, facial expressions, and masseter muscle activities.

[0098] By using the threshold method to denoise the EEG signal, the impact of noise on subsequent analysis can be effectively reduced. Since the electrooculographic artifacts are more obvious in the low-frequency band (0.5-10Hz) and the electromyographic artifacts are more obvious in the high-frequency band (>30Hz), in the frequency band where the artifacts are less obvious, the artifacts are mixed with the effective EEG signals, which can easily lead to misjudgment. If the noise in a specific frequency range is simply removed, some useful EEG information will also be filtered out. Therefore, in this case, by segmenting the signal and focusing on the area where the artifacts occur, it can not only ensure the effective removal of the artifacts, but also maximize the retention of effective EEG signals.

[0099] Specifically, in step S321, obtaining the real-time frequency of the original EEG signal for spectrum analysis includes:

[0100] The real-time frequency is determined according to the first standard frequency band and the second standard frequency band.

[0101] If the real-time frequency is in the first standard frequency band, the EEG signal is decomposed into multiple independent components or principal components and analyzed by independent component analysis (ICA) or principal component analysis (PCA) to identify the components related to the electrooculographic artifacts and delete them from the original signal;

[0102] If the real-time frequency is in the second standard frequency band, the signal is processed using the threshold method or wavelet transform method. When the signal amplitude exceeds the set threshold or shows abnormality in a specific wavelet coefficient, it is judged as an electromyographic artifact and deleted.

[0103] Specifically, in step S322, if the real-time frequency is in a non-first standard frequency band and a non-second standard frequency band, the original EEG signal is segmented according to the preset acquisition time into several original EEG signal segments, and a focused analysis is performed on the artifact occurrence area in any of the original EEG signal segments to identify electrooculographic artifacts and electromyographic artifacts for deletion.

[0104] In this embodiment, the first standard frequency band is 0.5Hz-10Hz, which indicates a low frequency band where electrooculographic artifacts are more obvious, and the second standard frequency band is 30Hz-50Hz, which indicates a high frequency band where electromyographic artifacts are more obvious. The focused analysis is to locally amplify the area where the artifact occurs, so as to accurately identify sudden changes in signal amplitude, increase or decrease in specific frequency components, and abnormal components in signal morphology for deletion. For example, for specific frequency components in the signal, the time domain signal is converted to the frequency domain through fast Fourier transform, and the frequency components of the signal are then analyzed. The time domain waveform of the signal is analyzed to identify the signal part that does not meet expectations and delete it.

[0105] In this embodiment, components related to electrooculogram artifacts are identified, and components are judged as myoelectric artifacts and deleted, including: setting a threshold value of 70μV, setting the range of normal level wavelet coefficients to [-5,5], and when determining that the real-time frequency is in the second standard frequency band, if a certain signal amplitude exceeds 70μV, it is judged as an electromyographic artifact and marked as abnormal for deletion. When it is found that the wavelet coefficient of a certain layer (for example, c3=10) exceeds this range, it is also marked as abnormal for deletion, effectively detecting and removing myoelectric artifacts, thereby improving the quality of EEG signal data. Principal component analysis is to map data from high-dimensional space to low-dimensional space through linear transformation, remove redundant information, and retain the main components; independent component analysis is to decompose the signal into multiple independent components based on statistical independence, and use visualization tools (such as component diagrams) to check the waveform and spectral characteristics of each component. Usually, the components of electrooculogram artifacts will appear as high-amplitude spikes in the time domain, and may appear as enhancements of specific frequency bands in the frequency domain (for example, low-frequency range). Components related to electrooculogram artifacts can be identified in the following way: set an amplitude threshold, assuming it is 10μV, observe all component waveforms, if the peak amplitude of a component exceeds this threshold, it can be considered that this component is associated with electrooculogram artifacts, and remove these components from the original data through linear combination or inverse ICA / PCA methods, and use the remaining independent components or principal components to reconstruct the original signal, thereby obtaining an EEG signal with artifacts removed.

[0106] By real-time identification of the frequency band in which the EEG signal is located and selecting different artifact recognition methods, artifacts are removed online in real time in the low-frequency band (0.5-10Hz) where electrooculographic artifacts are more obvious, and in the high-frequency band (>30Hz) where electromyographic artifacts are more obvious, by utilizing the information between channels, while in the frequency band where artifacts are less obvious, the signal is segmented and focused on analyzing the area where the artifact occurs, so as to better remove noise while retaining the effective signal to the maximum extent.

[0107] See also Figure 3 and Figure 4 As shown, Figure 3 A logic decision diagram for determining whether to adjust the standard adjustment frequency according to an embodiment of the present invention. Figure 4 A schematic diagram of logic determination for adjusting the standard adjustment frequency according to an embodiment of the present invention;

[0108] Specifically, in step S5, analyzing the relevant components in any target EEG signal segment includes:

[0109] Step S51, obtaining the P1 wave in any target EEG signal segment, drawing a curve of the P1 wave changing over time, and obtaining a waveform change curve;

[0110] Step S52, determining the signal peak value of the waveform change curve,

[0111] When the signal peak value is greater than the maximum value of the peak threshold range, the amplitude of the N1 wave or the P2 wave is obtained for determination;

[0112] When the signal peak is within the peak threshold range, the standard adjustment frequency is not adjusted, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in the data feature library.

[0113] When the signal peak value is less than the minimum value of the peak threshold range, emotion recognition is performed on the target EEG signal segment according to the established emotion recognition model to obtain a recognition result, and whether to adjust the standard adjustment frequency is determined according to the recognition result;

[0114] Among them, the recognition results include anxiety, normal emotions, and tired emotions.

[0115] If anxiety is identified, the standard adjustment frequency is reduced to the third modified adjustment frequency;

[0116] If a normal emotion is identified, no adjustment is made to the standard adjustment frequency;

[0117] If fatigue is detected, the standard adjustment frequency is increased to the second corrected adjustment frequency;

[0118] Among them, Fb2'=Fb×[1+(Nmix-Ns) / Nmix], Fb3'=Fb×[1-(Nmix-Ns) / Nmix], Fb2' is the second corrected adjustment frequency, Fb3' is the third corrected adjustment frequency, Fb is the standard adjustment frequency, Nmix is ​​the minimum value of the peak threshold range, and Ns is the signal peak.

[0119] In this embodiment, the peak threshold range represents the normal range of the subject's visual processing sensitivity to stimulation, the minimum value of the peak threshold range is 0.5 μV, and the maximum value of the peak threshold range is 5 μV; the standard adjustment frequency is the stimulation frequency of the incident light source set during the measurement process, and the standard adjustment frequency is set to 2.5 Hz.

[0120] By analyzing the P1 wave in the target EEG signal segment and plotting its curve over time, the peak change of the signal can be accurately identified. When the signal peak is determined to exceed the preset peak threshold range, it means that the stimulus has appeared and the amplitude is large. At this time, the subject's visual processing of the stimulus has reached a certain sensitivity. Then, by obtaining the amplitude of the N1 wave or P2 wave, further judgment is made to increase the standard adjustment frequency and approach the perception result more quickly. If the amplitude of the N1 wave or P2 wave is determined to be less than the threshold at this time, it means that although the subject has recognized the stimulus, the brain's further processing or cognitive evaluation of this stimulus has not reached the expected level or intensity. This is because the subject's current emotional state affects the characteristics of the EEG signal. In this case, the stimulus is adaptively increased. Frequency is used to improve the subject's attention level and approach the perception result more quickly, thereby improving the accuracy and reliability of the measurement. When the signal peak is determined to be within the peak threshold range, it means that the brain's cognitive evaluation of the stimulus has reached the expected level. The critical flicker fusion frequency value is determined by distinguishing the signal category, that is, whether the signal is a constant brightness feature or a flicker feature. When the signal peak is determined to be less than the minimum value of the peak threshold range, it may be that the subject's current emotional state has affected the EEG signal characteristics. The emotion recognition model is used to analyze the EEG signal fragments to more accurately identify the subject's emotional state. The stimulation frequency is adjusted according to the emotion recognition results to optimize the visual stimulation parameters to better adapt to the subject's emotional state and improve the sensitivity of visual perception.

[0121] Specifically, the established emotion recognition model is a convolutional neural network classification model. A large amount of labeled EEG data under different emotional states is collected in advance for model training to perform emotion recognition on the target EEG signal fragments, that is, to classify the feature vectors and identify the subject's current emotional state. If anxiety is identified, the stimulation frequency is reduced, that is, the standard adjustment frequency is reduced to the third corrected adjustment frequency. If fatigue is identified, the stimulation frequency is increased to improve attention. At this time, the standard adjustment frequency is increased to the second corrected adjustment frequency. If normal emotions are identified, staff are required to conduct on-site inspection and processing.

[0122] Specifically, in step S52, obtaining the amplitude of the N1 wave or the P2 wave for determination includes:

[0123] If the amplitude Ws of the N1 wave or the P2 wave is smaller than the corresponding amplitude threshold Wb, the standard adjustment frequency Fb is increased to the first correction adjustment frequency;

[0124] If the amplitude of the N1 wave or the P2 wave is greater than the corresponding amplitude threshold, obtain the P1 wave in the next adjacent target EEG signal segment, draw an adjacent continuous waveform curve, obtain the signal peak in the adjacent continuous waveform curve and continue the above determination;

[0125] Among them, Fb1'=Fb×[1+(Wb-Ws) / Wb], Fb1' is the first correction adjustment frequency, Fb is the standard adjustment frequency, Wb is the amplitude threshold, and Ws is the amplitude of the N1 wave or the P2 wave.

[0126] In this embodiment, the amplitude threshold of the N1 wave or the P2 wave is set between 0.5 μV and 10 μV, which is related to the age and gender of the subject, and is selected and adjusted according to the experimental design to ensure that the subject's cognitive evaluation level of the stimulus can be accurately judged.

[0127] Specifically, in step S52, matching the real-time characteristic signal in the data characteristic library includes:

[0128] Step S521, using a decision tree to match the real-time feature signal with each standard signal in the data feature library one by one to obtain a matching result;

[0129] Step S522, adding a label to the real-time characteristic signal according to the matching result and classifying it into a real-time constant light characteristic signal or a real-time flickering characteristic signal;

[0130] Step S523, storing the real-time constant brightness characteristic signal marked in the matching result into the perceived constant brightness data set, and storing the real-time flicker characteristic signal marked in the matching result into the perceived flicker data set;

[0131] The perceived constant brightness data set includes a real-time constant brightness characteristic signal and a corresponding label, and the perceived flicker data set includes a real-time flicker characteristic signal and a corresponding label.

[0132] The EEG feature signals extracted in real time are classified by adopting a decision tree. The cosine similarity between the real-time feature signal and the standard signal is calculated, and the standard signal whose cosine similarity is greater than the similarity threshold is selected as the matching signal. The real-time feature signal is classified according to the type of matching signal to determine the validity of the match and quickly and accurately judge whether the signal belongs to a constant brightness feature or a flickering feature.

[0133] Specifically, in step S6, analyzing the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the perceived constant brightness data set according to the CFF feedback data includes:

[0134] Step S61, obtaining the actual time interval between the CFF feedback moment and the constant light timestamp, and when the actual time interval is within the standard time interval range, determining that the key feedback is normal, and obtaining the critical flicker fusion frequency value according to the key feedback;

[0135] Step S62, when the actual time interval is less than the minimum value of the standard time interval range, it is determined that the subject presses the key too fast, a key error prompt is given, the current flashing frequency corresponding to the constant light timestamp is obtained, and the target perception result is obtained;

[0136] When the actual time interval is greater than the maximum value of the standard time interval range, it is determined that the subject did not press the key in time, and the current flashing frequency corresponding to the constant light timestamp is obtained to obtain the target perception result.

[0137] Specifically, in step S61, obtaining the critical flicker fusion frequency value according to the key feedback includes:

[0138] When the target perception result is obtained, the standard verification frequency is further adjusted downward once to obtain the first verification perception result, and the standard verification frequency is further adjusted upward once to obtain the second verification perception result; the critical flicker fusion frequency value is calculated according to the target perception result, the first verification perception result and the second verification perception result;

[0139] The calculation subunit takes the average value of the actual measurement frequencies corresponding to the target perception result, the first verification perception result and the second verification perception result to obtain the critical flicker fusion frequency value.

[0140] The standard time interval in this embodiment is used to measure the normal reaction time of the subject to stimulation. Generally, the standard time interval is set between 0.2s and 1s, preferably, it is set to 0.3s; the constant light timestamp indicates the moment when the stimulation signal starts; the standard verification frequency is set to 1Hz.

[0141] By obtaining the actual time interval between the CFF feedback moment and the constant light timestamp and comparing it with the standard time interval, the state of the key feedback can be effectively judged. When the actual time interval is greater than the maximum value of the standard time interval, it means that the subject did not press the key in time. At this time, the perception result is obtained based on the EEG signal to avoid the influence of subjective factors on the measurement, thereby improving the detection accuracy. When the actual time interval is less than the minimum value of the standard time interval, it means that the subject pressed the key too quickly. At this time, a key error will be prompted to avoid distortion of the test data caused by abnormal reactions, ensuring that the reaction time of the subject during the test is within a reasonable range.

[0142] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0143] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for measuring critical flicker fusion frequency based on electroencephalogram, characterized in that: include: Calibrate the LED light source and the EEG headset, adjust the EEG headset to the subject's head, turn on the signal processing unit, and connect the LED light source and the EEG headset to the signal processing unit; An initial flickering frequency is set in the LED light source, and the LED light source is started to provide visual stimulation to the subject, wherein the visual stimulation process is to emit a light beam at the initial flickering frequency or the current flickering frequency after being adjusted upward, and a trigger instruction is generated at each time when the visual stimulation occurs to send a trigger signal to the EEG acquisition device; The initial flickering frequency is adjusted upward several times with the standard adjustment frequency to update the current flickering frequency, and the visual stimulation occurs at the time corresponding to each time the light beam is emitted; Acquire the EEG signal of the subject under visual stimulation in real time through the EEG acquisition device, filter and denoise the EEG signal, and obtain the target EEG signal; Segmenting the target EEG signal into a plurality of target EEG signal segments, averaging any of the target EEG signal segments to generate an ERP waveform, marking each relevant component in the ERP waveform, and extracting key features in the ERP waveform, wherein the relevant components include a P1 wave, an N1 wave, and a P2 wave, and the key features include a signal peak value and a signal amplitude; Analyze each relevant component in any target EEG signal segment to determine whether to adjust the standard adjustment frequency; When the signal peak in the target EEG signal segment is within the peak threshold range, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in a data feature library to obtain a real-time constant brightness feature signal and a real-time flicker feature signal, so as to construct a perceived flicker data set and a perceived constant brightness data set; Analyze the constant brightness timestamp corresponding to the real-time constant brightness feature signal in the constant brightness perception data set according to the CFF feedback data, The actual time interval between the CFF feedback moment and the constant light timestamp is obtained. When the actual time interval is within the standard time interval range, the key feedback is determined to be normal, and the critical flicker fusion frequency value is obtained according to the key feedback. When the actual time interval is less than the minimum value of the standard time interval range, a key error prompt indication is given, the current flashing frequency corresponding to the constant light timestamp is obtained, and the target perception result is obtained; When the actual time interval is greater than the maximum value of the standard time interval range, the current flashing frequency corresponding to the constant light timestamp is obtained to obtain the target perception result; Wherein, the CFF feedback data includes the CFF feedback time; The start time of the real-time flicker characteristic signal is acquired to obtain the constant brightness timestamp.

2. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 1, characterized in that: Filtering and denoising the EEG signal includes: Intercepting the EEG signal according to the set standard frequency band range, acquiring the EEG signal falling within the standard frequency band range, and obtaining the original EEG signal; The electrooculographic artifacts and electromyographic artifacts of the original EEG signal are identified and deleted to obtain a target EEG signal.

3. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 2, characterized in that: Identifying the electrooculographic artifacts and electromyographic artifacts of the original EEG signal and deleting them includes: Acquiring the real-time frequency of the original EEG signal for spectrum analysis to divide the original EEG signal into different frequency bands; When the real-time frequency does not fall within the first standard frequency band or the second standard frequency band, the artifact occurrence area in the original EEG signal is identified, and the artifact occurrence area is focused and analyzed to identify the electrooculographic artifacts and electromyographic artifacts for deletion.

4. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 3, characterized in that: Obtaining the real-time frequency of the original EEG signal for spectrum analysis includes: The real-time frequency is determined according to the first standard frequency band and the second standard frequency band. If the real-time frequency is in the first standard frequency band, the original EEG signal is decomposed into a plurality of independent components or principal components and analyzed, components related to electrooculographic artifacts are identified, and deleted from the original signal; If the real-time frequency is in the second standard frequency band, the signal is processed using the threshold method or wavelet transform method. When the signal amplitude exceeds the set threshold or shows abnormality in a specific wavelet coefficient, it is judged as an electromyographic artifact and deleted. If the real-time frequency is neither in the first standard frequency band nor in the second standard frequency band, the original EEG signal is segmented according to the preset acquisition time into several original EEG signal segments, and a focused analysis is performed on the artifact occurrence area in any of the original EEG signal segments to identify the electrooculographic artifacts and electromyographic artifacts and delete them.

5. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 1, characterized in that: The analysis of each relevant component in any target EEG signal segment includes: Obtain the P1 wave in any target EEG signal segment, draw a curve of the P1 wave changing over time, and obtain a waveform change curve; Determine the peak value of the waveform change curve, When the peak value of the real-time signal is greater than the maximum value of the peak threshold range, the amplitude of the N1 wave or the P2 wave is obtained for determination; When the peak value of the real-time signal is within the peak threshold range, the standard adjustment frequency is not adjusted, a real-time feature signal is extracted from the target EEG signal, and the real-time feature signal is matched in a data feature library; When the real-time signal peak is less than the minimum value of the peak threshold range, emotion recognition is performed on the target EEG signal segment according to the established emotion recognition model to obtain a recognition result, and it is determined whether to increase the standard adjustment frequency to the second corrected adjustment frequency according to the recognition result.

6. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 5, characterized in that: The recognition results include anxious emotions, normal emotions, and tired emotions; If anxiety is identified, the standard adjustment frequency is reduced to the third modified adjustment frequency; If a normal emotion is identified, no adjustment is made to the standard adjustment frequency; If fatigue is detected, the standard adjustment frequency is increased to the second corrected adjustment frequency.

7. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 5, characterized in that: Obtaining the amplitude of the N1 wave or the P2 wave for determination includes: When the amplitude of the N1 wave or the P2 wave is less than the corresponding amplitude threshold, the standard adjustment frequency is increased to the first correction adjustment frequency; When the amplitude of the N1 wave or the P2 wave is greater than the corresponding amplitude threshold, the P1 wave in the next adjacent target EEG signal segment is obtained, an adjacent continuous waveform curve is drawn, and the signal peak value in the adjacent continuous waveform curve is obtained to continue the determination.

8. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 5, characterized in that: Matching the real-time characteristic signal in the data characteristic library includes: Use the decision tree to match the real-time feature signal with each standard signal in the data feature library one by one to obtain the matching result; According to the matching results, the real-time characteristic signal is labeled and classified into a real-time constant light characteristic signal or a real-time flashing characteristic signal; The real-time constant brightness feature signal marked in the matching result is stored in a perception constant brightness data set, and the real-time flicker feature signal marked in the matching result is stored in a perception flicker data set; The perceived constant brightness data set includes a real-time constant brightness characteristic signal and a corresponding label, and the perceived flicker data set includes a real-time flicker characteristic signal and a corresponding label.

9. The method for measuring critical flicker fusion frequency based on electroencephalogram according to claim 1, characterized in that: The process of segmenting the target EEG signal is: Based on the stimulation time, the target EEG signal is extracted according to the preset stimulation duration to obtain the target EEG signal segment.

Images