CN106236083B - Equipment for removing ocular artifacts in sleep state analysis - Google Patents

Abstract

The invention relates to a device for removing electro-oculogram artifacts in sleep state analysis, including: EEG electrodes, electro-oculogram electrodes, reference electrodes and analog-to-digital converters connected thereto, and the processing connected through the analog-to-digital converters and filter circuits The electroencephalogram electrode is used to detect the original electroencephalogram signal; the oculoelectric electrode is used to collect the oculoelectric signal; the analog-to-digital converter is used for analog-to-digital conversion, and the filter circuit is used for low-frequency filtering and then input to the processor; the processor is used for The EEG signal after each frame filtering is subjected to empirical mode decomposition, decomposed into several eigenmode functions, and the correlation coefficient between each eigenmode function and the oculoelectric signal at the same time is calculated; find and delete the correlation coefficient greater than The eigenmode function with the preset threshold and the eigenmode function with the largest correlation coefficient are used to reconstruct the EEG signal of each frame by using the remaining eigenmode function. The invention can reduce the influence of the electrooculograph artifact removal process on the waveform of the electroencephalogram signal, and retain most of the detail information of the original signal.

Description

A device for removing oculograph artifacts in sleep state analysis

technical field

The invention relates to the technical field of sleep aids, in particular to a device for removing oculoelectric artifacts in sleep state analysis.

Background technique

During sleep, the human body undergoes a process of self-relaxation and recovery, so good sleep is a basic condition for maintaining good health; however, due to work pressure, irregular life schedule and other reasons, some people have poor sleep quality , manifested as insomnia, waking up in the middle of the night, etc.

There are already some devices on the market to help people fall asleep and improve the quality of sleep. For example, in a specific sleep state, manual intervention such as sound and light signals can be used to avoid waking up the user in a deep sleep state. For devices that assist sleep, in order to really achieve the purpose of improving the user's sleep quality, it is very important to correctly identify the user's sleep state.

Polysomnography (PSG), also known as sleep EEG, is currently the "gold standard" for clinical sleep diagnosis and analysis. Polysomnography uses a variety of vital signs such as EEG, EMG (submandibular), Oculoelectricity, respiration, blood oxygen, etc. to analyze sleep. Among these signs and signals, the electroencephalogram (EOG) is at the core. An electroencephalogram is a waveform signal that uses sophisticated electronic instruments to record and amplify the electrical activity from the cerebral cortex on the scalp. Since the EEG signal is very weak (microvolt level), it is easily interfered by bioelectrical signals from other parts. When the amplitude of the EEG signal is low (that is, there is no strong eyeball/eyelid activity such as blinking), the interference of the EEG signal on the EEG signal is relatively weak. When the amplitude of the EEG signal is high, because the frequency of the EEG signal is lower than that of the normal EEG signal, the high-amplitude EEG signal is superimposed on the EEG signal to form a phenomenon similar to baseline drift.

In order to reduce the impact of electro-oculogram signals, there are currently many methods for removing electro-oculogram artifacts. Independent component analysis (Indepdent component analysis, ICA) is a commonly used method. It first assumes that the input signals are a linear combination of statistically independent non-Gaussian signals, and then uses a linear transformation to separate them from the signals. Its disadvantages are (1) the assumed conditions of the input signal cannot be fully satisfied in actual use; (2) for multiple separated signals, it is necessary to further judge which signals are "pure" EEG signals and which signals are The separated electro-oculogram signal. In addition, there is a method that assumes an influence factor (such as 0.2) of the electrooculogram signal on the EEG signal, and then removes the electrooculogram artifact by subtracting the electrooculogram signal multiplied by the influence factor from the EEG signal, such as the formula: EEG _pure ＝EEG _original -0.2*EOG, due to individual differences and differences in the location of the electro-oculogram electrodes, a fixed impact factor cannot be well adapted to different individuals.

In addition, in the sleep state analysis, the waveform of the EEG signal is a very important sleep state indicator. For example, the appearance of spindle waves and K complex waves indicates that the S2 phase of non-rapid eye movement sleep has entered. The waveform of the EEG signal processed by the traditional method often changes, which affects the subsequent analysis of the EEG signal.

Contents of the invention

Based on this, it is necessary to address the above problems and provide a device for removing EEG artifacts in sleep state analysis, reducing the influence of the EEG artifact removal process on the waveform of EEG signals, and ensuring the subsequent analysis of EEG signals.

A device for removing electro-oculogram artifacts in sleep state analysis, comprising: EEG electrodes, electro-oculogram electrodes, reference electrodes, analog-to-digital converters, filter circuits and processors;

The EEG electrodes, oculoelectric electrodes, and reference electrodes are respectively connected to the analog-to-digital converter, and are connected to the processor through the analog-to-digital converter and the filter circuit in turn;

The EEG electrodes are used to detect the EEG signals of the user during sleep; the electro-oculogram electrodes are used to collect the EEG signals of the user during sleep;

The analog-to-digital converter converts the electro-oculogram signal and the electroencephalogram signal into digital signals, and the filter circuit performs low-frequency filtering on the electro-oculogram signal and the electroencephalogram signal before inputting them to the processor;

The processor is used to perform empirical mode decomposition on the filtered EEG signal of each frame, decompose it into several eigenmode functions, and calculate the correlation between each eigenmode function and the electrooculogram signal at the same time Coefficient; find and delete the intrinsic mode function with the correlation coefficient greater than the preset threshold and the intrinsic mode function with the largest correlation coefficient, and use the remaining intrinsic mode function to reconstruct the EEG signal of each frame.

In the sleep state analysis mentioned above, the equipment for removing the electrooculogram artifacts uses the electrooculogram signals of the user collected by the electro-oculogram electrodes and the electroencephalogram signals collected by the electroencephalogram electrodes. After analog-to-digital conversion and filtering, the processor filters each frame. The final EEG signal is subjected to empirical mode decomposition, decomposed into several eigenmode functions, and the correlation coefficient between each eigenmode function and the oculoelectric signal at the same time is calculated; find and delete the correlation coefficient greater than the preset threshold The eigenmode function and the eigenmode function with the largest correlation coefficient are used to reconstruct the EEG signal of each frame by using the remaining eigenmode function. The device can reduce the influence of the EEG artifact removal process on the waveform of the EEG signal, retain most of the details of the signal, and ensure the subsequent analysis of the EEG signal.

Description of drawings

FIG. 1 is a schematic structural diagram of a device for removing electrooculogram artifacts in sleep state analysis of an embodiment;

Fig. 2 is the algorithm flow chart of removing the electrooculograph artifact by the processor;

Fig. 3 is a schematic diagram of the experimental data results of removing electroocular artifacts.

detailed description

Embodiments of the device for removing oculograph artifacts in sleep state analysis of the present invention will be described below with reference to the accompanying drawings.

Referring to Fig. 1, Fig. 1 is a schematic structural diagram of a device for removing electrooculogram artifacts in sleep state analysis of the present invention, including: EEG electrodes, electro-oculogram electrodes, reference electrodes, analog-to-digital converters, filter circuits and processors ;

The EEG electrodes, oculoelectric electrodes, and reference electrodes are respectively connected to the analog-to-digital converter, and are connected to the processor through the analog-to-digital converter and the filter circuit in turn;

The EEG electrodes are used to detect the EEG signals of the user during sleep; the electro-oculogram electrodes are used to collect the EEG signals of the user during sleep;

The analog-to-digital converter converts the electro-oculogram signal and the electroencephalogram signal into digital signals, and the filter circuit performs low-frequency filtering on the electro-oculogram signal and the electroencephalogram signal before inputting them to the processor;

The processor is used to perform empirical mode decomposition on the filtered EEG signal of each frame, decompose it into several eigenmode functions, and calculate the correlation between each eigenmode function and the electrooculogram signal at the same time Coefficient; find and delete the intrinsic mode function with the correlation coefficient greater than the preset threshold and the intrinsic mode function with the largest correlation coefficient, and use the remaining intrinsic mode function to reconstruct the EEG signal of each frame.

In the device for removing electrooculogram artifacts in the sleep state analysis of the above-mentioned embodiments, the electrooculogram signals of the user collected by the electro-oculogram electrodes and the electroencephalogram signals collected by the electroencephalogram electrodes are processed by the processor after analog-to-digital conversion and filtering. The EEG signal after each frame filtering is subjected to empirical mode decomposition, decomposed into several eigenmode functions, and the correlation coefficient between each eigenmode function and the oculoelectric signal at the same time is calculated; find and delete the correlation coefficient greater than The eigenmode function with the preset threshold and the eigenmode function with the largest correlation coefficient are used to reconstruct the EEG signal of each frame by using the remaining eigenmode function. The device can reduce the influence of the EEG artifact removal process on the waveform of the EEG signal, retain most of the details of the signal, and ensure the subsequent analysis of the EEG signal.

Subsequently, the EEG signal output by the device can be used to monitor and analyze the sleep state, etc. Of course, the subsequent processing can also be implemented on the processor.

In one embodiment, the EEG electrodes are set on the user's forehead; the reference electrodes are set on the user's earlobe; the eye-electric electrodes are set on the corners of the eyes; That is "M" in the figure, the oculoelectric electrode includes left and right electrodes, namely "ROC" and "LOC" in the figure, the reference electrode is set on the user's earlobe, namely "R" and "L" in the figure, the acceleration sensor It is "AT" in the figure. The filter circuit is mainly for low-pass filtering and filtering out power frequency interference. In order to adapt to the processing of EEG signals and electro-oculogram signals, the filter circuit outputs signals in the 0-256Hz frequency band to the processor after filtering.

The function of removing the electro-ocular artifact is mainly performed by the processor, and based on the functions realized by the processor, corresponding algorithm modules can be configured in the processor.

The algorithm flow of the processor to remove the oculoelectric artifacts includes (1) to (5), as follows:

(1) The processor controls the electro-oculogram electrodes and EEG electrodes to collect the electro-oculogram signals and EEG signals of the user according to the set frame length;

For example, in the sleep state analysis of the user such as assisted sleep, the processor can set the frame length to collect the electro-oculogram and electroencephalogram signals generated by the user during sleep through the oculoelectric electrodes and electroencephalogram electrodes worn by the user. When collecting signals, 30s can be used as a frame for collection, and subsequent analysis and processing of each frame of electro-ocular signals and electroencephalogram signals.

(2) Carry out empirical mode decomposition to the frame of EEG signal, decompose it into several eigenmode functions, and obtain the eigenmode function set;

Here, the processor performs empirical mode decomposition on the EEG signal, and decomposes it into the form of the sum of several intrinsic mode functions (Intrinsic Mode Function, IMF) and residual function (Redisual, Re).

The set of eigenmode functions includes the following formulas:

In the formula, EEG _original represents the EEG signal, imf _i represents the i-th eigenmode function, and Re represents the residual function.

(3) Calculate the correlation coefficient between each eigenmode function of the eigenmode function set and the electrooculogram signal at the same moment respectively;

Referring to Fig. 2, Fig. 2 is a flow chart of the algorithm for the processor to remove electrooculogram artifacts. After the EEG signal is subjected to empirical mode decomposition, the set of intrinsic mode functions is obtained, and the intrinsic mode functions 1-n(imf ₁ ~imf _n ) Correlation coefficient 1-n (corrcoef ₁ ˜corrcoef _n ) with the electrooculogram signal EOG.

(4) find out the eigenmode function whose correlation coefficient is greater than the preset threshold and the eigenmode function with the largest correlation coefficient, and delete it from the eigenmode function set;

As shown in Figure 2, by setting the threshold, after the correlation coefficient is calculated, the eigenmode function with the correlation coefficient greater than the threshold and the eigenmode function with the largest correlation coefficient are deleted, leaving m eigenmode functions.

As an embodiment, after calculating the correlation coefficient, the processor can also be used in the intrinsic mode function whose correlation coefficient is greater than the preset threshold; calculate the Euclidean distance between the intrinsic mode function and the electrooculogram signal; from the Euclidean distance The smallest eigenmode function is eliminated from the set of eigenmode functions.

Similarly, after calculating the correlation coefficient, the processor can also be used in the eigenmode function whose correlation coefficient is greater than the preset threshold; calculate the cosine distance between the eigenmode function and the electrooculogram signal; from the eigenmode function with the smallest cosine distance The modulus function is excluded from the set of intrinsic modulus functions.

Through the above-mentioned embodiment, the determination of Euclidean distance or cosine distance is combined with the determination of the correlation coefficient, so that more residual electroocular artifacts that cannot be removed in the determination of the correlation coefficient can be removed.

(5) Utilize the remaining eigenmode functions in the eigenmode function set to reconstruct the frame of EEG signal;

After removing the electro-oculogram artifacts, the processor uses the remaining m eigenmode functions to reconstruct the EEG signals after removing the electro-oculogram artifacts. As an embodiment, when the processor reconstructs the EEG signal, according to the order of the eigenmode functions, the processor selects a number of eigenmode functions at the front in the eigenmode function set to reconstruct the EEG signal.

In this embodiment, since the order of the intrinsic mode functions is from large to small according to the frequency, and the intrinsic mode functions with the highest similarity to the electrooculogram signal are generally arranged in the middle position, when reconstructing the EEG signal, only Using the first few eigenmode functions, delete the eigenmode functions with lower frequency including the eigenmode function with the highest similarity, and then reconstruct the EEG signal.

Wherein, the processor can use the following formula to reconstruct the EEG signal:

In the formula, EEG _pure represents the reconstructed EEG signal, corrcoef represents the correlation coefficient, imf _i represents the i-th eigenmode function, EOG represents the oculoelectric signal, corrcoef _max represents the maximum correlation coefficient, and thre represents the preset correlation coefficient threshold .

In one embodiment, the processor may divide the EEG signal frame into N time windows before performing EMD on the EEG signal, and perform eigenmode function decomposition on the EEG signal in each time window ; and after reconstructing the EEG signal, combining the EEG signals reconstructed in each time window to obtain the EEG signal frame.

In the above embodiments, by dividing the collected EEG signal frames into multiple time windows for parallel processing, the speed of signal processing can be accelerated and the efficiency of sleep state analysis can be improved.

For example, taking a frame of 30s as an example, a time window length of 5s or 10s can be used.

The device for removing electrooculogram artifacts in sleep state analysis of the present invention only removes artifacts similar to baseline drift caused by high-amplitude electrooculograms, and retains most of the detailed information of the signal; the subsequent use of the EEG signal to perform Better results can be obtained when sleep state analysis is performed.

Referring to FIG. 3 , FIG. 3 is a schematic diagram of experimental data results for removing electrooculus artifacts. Figure 3(a) is the collected EEG signal, Figure 3(b) is the collected EEG signal, and Figure 3(c) compares the EEG signal before and after removing the EEG artifact (in the figure, ① is the EEG signal ;②It is the EEG signal after removing the oculograph artifacts), the figure below is an enlarged part of the intercepted part of the above figure, and it can be found that between the data points in the above interval, there is a deep V-shaped While the fluctuation is eliminated by this scheme, more information is retained.

Compared with traditional methods (such as ICA, etc.), which regard the input multi-channel signals as multi-channel source signals after linear combination, and try to separate these signals from each other, the traditional method can obtain better results on periodic signals. Moreover, for the EEG signal, since the EEG signal and the EEG signal can be regarded as random signals, and the EEG signal is easily subject to external interference, it is difficult to completely separate the EEG signal from the EEG signal. The signal will be mixed with additional noise signals, which increases the difficulty of subsequent signal processing and analysis. However, the technical solution of the present invention only removes artifacts similar to baseline drift caused by high-amplitude electro-oculograms, and retains most of the detailed information of the signal. Therefore, it is beneficial to the subsequent processing of the EEG signal analysis method based on the time domain.

The various technical features of the above-mentioned embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the various technical features in the above-mentioned embodiments are not described. However, as long as there is no contradiction in the combination of these technical features, should be considered as within the scope of this specification.

The above-mentioned embodiments only express several implementation modes of the present invention, and the descriptions thereof are relatively specific and detailed, but should not be construed as limiting the patent scope of the invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention should be based on the appended claims.

Claims (10)

1. the equipment of eye electricity artefact is removed in a kind of sleep state analysis, it is characterised in that including：Electrode for encephalograms, eye electricity electrode, Reference electrode, analog-digital converter, filter circuit and processor；

The electrode for encephalograms, eye electricity electrode, reference electrode connect analog-digital converter, and pass sequentially through the analog-digital converter respectively Processor is connected to filter circuit；

The electrode for encephalograms is used to detect EEG signals of the user in sleep；The eye electricity electrode is being slept for gathering user In electro-ocular signal；

Electro-ocular signal and EEG signals are converted to data signal by the analog-digital converter, the filter circuit to electro-ocular signal and EEG signals are inputted to processor after carrying out low frequency filtering；

The processor, for carrying out empirical mode decomposition to the filtered EEG signals of every frame, it is broken down into several Modular function is levied, calculates the coefficient correlation between each intrinsic mode functions and the electro-ocular signal of synchronization；Search and delete correlation Coefficient is more than the intrinsic mode functions of predetermined threshold value and the intrinsic mode functions that coefficient correlation is maximum, utilizes remaining intrinsic mode functions weight Build every frame EEG signals.

2. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 1, it is characterised in that the brain electricity Electrode is arranged on the forehead position of user；The reference electrode is arranged on the ear-lobe of user；The eye electricity electrode is arranged on canthus Position；The signal of the filter circuit output 0-256Hz frequency ranges.

3. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 1, it is characterised in that the processing Device, for the electro-ocular signal and EEG signals according to setting frame length collection user；Empirical modal is carried out to the frame EEG signals Decompose, be broken down into several intrinsic mode functions, obtain intrinsic mode functions set；The intrinsic mode functions set is calculated respectively Each intrinsic mode functions and synchronization electro-ocular signal between coefficient correlation；Find out coefficient correlation and be more than predetermined threshold value Intrinsic mode functions and the maximum intrinsic mode functions of coefficient correlation, and it is deleted from intrinsic mode functions set；Using intrinsic Remaining intrinsic mode functions rebuild the frame EEG signals in modular function set.

4. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 3, it is characterised in that described intrinsic Modular function set includes equation below：

In formula, EEG_originalRepresent EEG signals, imf_iI-th of intrinsic mode functions is represented, Re represents residual error function.

5. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 1, it is characterised in that the processing Device rebuilds EEG signals using equation below：

In formula, EEG_pureThe EEG signals rebuild are represented, corrcoef represents coefficient correlation, imf_iRepresent i-th of eigen mode letter Number, EOG represent electro-ocular signal, corrcoef_maxMaximum coefficient correlation is represented, thre represents default correlation coefficient threshold.

6. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 3, it is characterised in that the processing Device, it is additionally operable in the intrinsic mode functions that coefficient correlation is more than predetermined threshold value, calculates the Euclidean of intrinsic mode functions and electro-ocular signal Distance, the minimum intrinsic mode functions of Euclidean distance are rejected from intrinsic mode functions set.

7. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 3, it is characterised in that the processing Device, it is additionally operable in the intrinsic mode functions that coefficient correlation is more than predetermined threshold value, calculates the cosine of intrinsic mode functions and electro-ocular signal Distance, the minimum intrinsic mode functions of COS distance are rejected from intrinsic mode functions set.

8. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 3, it is characterised in that the processing Device, it is additionally operable to before empirical mode decomposition is carried out to EEG signals, EEG signals frame is divided into several time windows, and it is right The EEG signals of each time window carry out intrinsic mode functions decomposition；It is and after EEG signals are rebuild that each time window is salty The EEG signals built merge, and obtain EEG signals frame.

9. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 3, it is characterised in that the place Device is managed, is additionally operable to when rebuilding EEG signals, according to putting in order for intrinsic mode functions, selects position in intrinsic mode functions set Several forward intrinsic mode functions carry out reconstruction EEG signals.

10. the equipment of eye electricity artefact is removed in sleep state analysis according to claim 8, it is characterised in that described pre- If threshold value is 0.5, the frame length is 30s, and the time window length is 5s or 10s.