CN109512394B - Multi-channel evoked potential detection method and system based on independent component analysis - Google Patents
Multi-channel evoked potential detection method and system based on independent component analysis Download PDFInfo
- Publication number
- CN109512394B CN109512394B CN201811484564.1A CN201811484564A CN109512394B CN 109512394 B CN109512394 B CN 109512394B CN 201811484564 A CN201811484564 A CN 201811484564A CN 109512394 B CN109512394 B CN 109512394B
- Authority
- CN
- China
- Prior art keywords
- evoked
- independent component
- energy gain
- evoked potential
- channel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 230000000763 evoking effect Effects 0.000 title claims abstract description 155
- 238000001514 detection method Methods 0.000 title claims abstract description 71
- 238000012880 independent component analysis Methods 0.000 title claims abstract description 51
- 230000000638 stimulation Effects 0.000 claims abstract description 75
- 238000000034 method Methods 0.000 claims abstract description 36
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 238000012360 testing method Methods 0.000 claims abstract description 13
- 238000004364 calculation method Methods 0.000 claims abstract description 10
- 238000012544 monitoring process Methods 0.000 claims description 18
- 238000012216 screening Methods 0.000 claims description 16
- 238000012935 Averaging Methods 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 11
- 230000004044 response Effects 0.000 claims description 10
- 238000012353 t test Methods 0.000 claims description 10
- 238000000692 Student's t-test Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 8
- 238000001914 filtration Methods 0.000 claims description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 230000008030 elimination Effects 0.000 claims description 5
- 238000003379 elimination reaction Methods 0.000 claims description 5
- 238000007476 Maximum Likelihood Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 4
- 230000000284 resting effect Effects 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 230000001052 transient effect Effects 0.000 claims description 4
- 210000004556 brain Anatomy 0.000 claims description 2
- 230000035945 sensitivity Effects 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 5
- 230000003252 repetitive effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 230000001537 neural effect Effects 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000010304 firing Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000005381 potential energy Methods 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/377—Electroencephalography [EEG] using evoked responses
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Psychiatry (AREA)
- Molecular Biology (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychology (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
本发明公开了基于独立成分分析的多通道诱发电位检测方法及系统，该方法包括：采集被测试者在重复刺激作用下的多通道脑电信号并进行预处理；对预处理后的多通道脑电信号进行独立成分分析，获得多个独立成分；以刺激单元的起始时刻为基准，获得每个刺激单元的独立成本样本；对独立成分样本进行时域的叠加平均，获得每个独立成分的平均诱发电位；按照设定的准则对独立成分样本进行筛选剔除；选取检测样本，计算检测样本的总诱发能量增益和检测样本的平均样本的总诱发能量增益，判断选取的检测样本是否存在刺激单元所引发的诱发电位。本发明计算速度快、效率高，检测灵敏度高，可广泛应用于微弱诱发电位的检测领域。
The invention discloses a multi-channel evoked potential detection method and system based on independent component analysis. The method comprises: collecting multi-channel EEG signals of a subject under repeated stimulation and preprocessing; Perform independent component analysis on the electrical signal to obtain multiple independent components; take the start time of the stimulation unit as the benchmark to obtain an independent cost sample of each stimulation unit; perform a time-domain superposition average on the independent component samples to obtain the cost of each independent component. Average evoked potential; screen out the independent component samples according to the set criteria; select the test sample, calculate the total evoked energy gain of the test sample and the total evoked energy gain of the average sample of the test sample, and judge whether the selected test sample has a stimulation unit evoked potentials. The invention has fast calculation speed, high efficiency and high detection sensitivity, and can be widely used in the field of weak evoked potential detection.
Description
Technical Field
The invention relates to the field of biomedical information processing, in particular to a multichannel evoked potential detection method and a multichannel evoked potential detection system based on independent component analysis.
Background
In the process of collecting and processing biomedical signals, compared with background electroencephalogram signals and interference signals of 20-150 mu V, evoked potential signals are very weak, and are mainly in the range of 0.1-10 mu V. The traditional method generally adopts a superposition averaging technology to remove background electroencephalograms and interference signals, and in the mode, the number of times of superposition averaging needs to be determined according to the signal-to-noise ratio of the recorded background electroencephalograms, and the average number of times of dozens or even thousands of times is usually needed to judge whether the signal to be detected contains evoked potentials, so that long time is needed, and the efficiency is low.
Noun interpretation
ICA: the full name Independent Component Analysis, which represents Independent Component Analysis, is a very effective data Analysis tool proposed in recent years, and is mainly used for extracting original Independent signals from mixed data.
Unilateral t-test: a common statistical method in statistics, also known as the single population t-test, is used to test whether a sample mean differs significantly from a known population mean.
Disclosure of Invention
In order to solve the above technical problems, an object of the present invention is to provide a fast and efficient multi-channel evoked potential monitoring method and system based on independent component analysis.
In one aspect, an embodiment of the present invention provides a multichannel evoked potential monitoring method based on independent component analysis, including the following steps:
collecting multi-channel electroencephalogram signals of a tested person under the action of repeated stimulation; the repeated stimulation is formed by repeatedly arranging a plurality of identical transient stimulation units in series;
preprocessing the acquired multi-channel electroencephalogram signals;
carrying out independent component analysis on the preprocessed multi-channel electroencephalogram signals to obtain a plurality of independent components;
taking the initial time of the stimulation unit as a reference, and taking a period of time before and after the reference to form a time window of signal segmentation; segmenting the independent components obtained after the independent component analysis according to a time window so as to obtain independent component samples of each stimulation unit;
carrying out time domain superposition averaging on the independent component samples to obtain an average evoked potential of each independent component;
after the evoked energy gain of the average evoked potential of each independent component is calculated, independent component samples are screened and removed according to a set criterion;
selecting a detection sample from the independent component samples, calculating the total evoked energy gain of the detection sample and the total evoked energy gain of the average sample of the detection sample, and judging whether the selected detection sample has the evoked potential caused by the stimulation unit.
Further, the step of selecting a detection sample from the independent component samples, calculating a total evoked energy gain of the detection sample and a total evoked energy gain of an average sample of the detection sample, and determining whether the selected detection sample has an evoked potential caused by the stimulus unit specifically includes:
selecting a plurality of independent component samples as detection samples, and calculating the total evoked energy gain of each detection sample;
performing superposition averaging on the selected detection samples, and then calculating the total induced energy gain of the averaged average samples;
and comparing the total evoked energy gain of the detection sample with the total evoked energy gain of the average sample, and judging whether the evoked potential caused by the stimulation unit exists in the detection sample.
Further, in the step of comparing the total evoked energy gain of the detection sample with the total evoked energy gain of the average sample and judging whether the evoked potential caused by the stimulation unit exists in the detection sample, a modified unilateral t-test method is adopted for test judgment.
Further, in the step of preprocessing the acquired multi-channel electroencephalogram signals, the preprocessing includes at least one of filtering processing, interpolation processing and abnormal data elimination.
Further, in the step of obtaining a plurality of independent components after performing independent component analysis on the preprocessed multi-channel electroencephalogram signals, independent component analysis is performed by adopting an information maximization ICA algorithm or a maximum likelihood ICA algorithm.
Further, the calculation formula of the evoked energy gain is as follows:
where EER denotes the evoked energy gain in dB, x (τ) denotes the signal to be calculated, τ -0 denotes the starting time of the stimulation unit, Δ t1B-a denotes spineA time before the initiation of the excitation unit, specifically representing the rest period of the evoked response, Δ t2D-c denotes the period of time after the stimulation unit has started the onset of evoked potentials.
Further, after the evoked energy gain of the average evoked potential of each independent component is calculated, the step of screening and rejecting independent component samples according to a set criterion is specifically as follows:
and after the evoked energy gain of the average evoked potential of each independent component is calculated, screening out the independent component corresponding to the evoked energy gain smaller than a preset threshold value as an interference component to remove.
Further, the calculation formula of the total evoked energy gain is as follows:
wherein GEER represents the total induced energy gain in dB, xj(τ) represents the signal to be calculated, j ═ 1,2, …, p, the serial number of the individual component samples remaining after screening, τ ═ 0 represents the starting time of the stimulation unit, Δ t1B-a denotes the moment before the start of the stimulation unit, in particular the resting period of the evoked response, Δ t2D-c denotes the period of time after the stimulation unit has started the onset of evoked potentials.
In another aspect, an embodiment of the present invention provides a multi-channel evoked potential monitoring system based on independent component analysis, including:
at least one processor;
at least one memory for storing at least one program;
when the at least one program is executed by the at least one processor, the at least one program causes the at least one processor to implement a multi-channel evoked potential monitoring method based on independent component analysis according to an embodiment of the present invention.
In this embodiment, after acquiring and preprocessing a multi-channel electroencephalogram signal, performing independent component analysis to obtain a plurality of independent components corresponding to the multi-channel electroencephalogram signal, so as to obtain an independent cost sample of each stimulation unit based on the starting time of the stimulation unit, performing time-domain superposition averaging on the independent component samples to obtain an average evoked potential of each independent component, screening and removing the independent components based on an evoked energy gain of the average evoked potential of each independent component according to a set criterion, selecting a plurality of independent component samples as detection samples from the independent component samples, calculating a total evoked energy gain of the detection samples and a total evoked energy gain of the average sample of the detection samples, and thus determining whether the selected detection samples have evoked potentials caused by the stimulation units Screening and rejecting can detect whether any detection sample contains evoked potential, the calculation speed is high, the efficiency is high, the detection sensitivity is high, the problems of long time consumption and low efficiency of the existing detection method are solved, and whether the signal to be detected contains evoked potential can be quickly detected.
Drawings
The invention is further illustrated by the following figures and examples.
FIG. 1 is a flow chart of a multi-channel evoked potential monitoring method based on independent component analysis in accordance with an exemplary embodiment of the present invention;
FIG. 2 is a schematic representation of the average evoked potentials of the individual components obtained in an embodiment of the present invention;
FIG. 3 is a schematic diagram of an evoked potential energy waveform and energy gain in an embodiment of the present invention;
FIG. 4 is an electronic block diagram of an independent component analysis based multi-channel evoked potential monitoring system in accordance with an implementation of the present invention.
Detailed Description
The step numbers in the embodiments of the present invention are set for convenience of illustration only, the order between the steps is not limited at all, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.
Method embodiment one
Referring to fig. 1, the present embodiment provides a multi-channel evoked potential monitoring method based on independent component analysis, which is executed by a microprocessor, an embedded processor or a computer terminal, and comprises the following steps:
s1, collecting multi-channel electroencephalogram signals of the testee under the action of repeated stimulation; the repetitive stimulus is composed of a plurality of identical transient stimulus units arranged in series, and in this embodiment, the repetitive stimulus is composed of N series of stimulus units.
S2, preprocessing the acquired multi-channel electroencephalogram signals; in this step, the preprocessing includes at least one of filtering, interpolation, and elimination of abnormal data. The filtering processing is as follows: the band-pass filtering is carried out on the multi-channel electroencephalogram signals, and the specific pass band range can be set according to the actual signal bandwidth. The interpolation processing means: according to the statistical characteristics, after detecting the electroencephalogram signals of a plurality of channels with abnormality, carrying out interpolation processing on the electroencephalogram signals; the statistical characteristics refer to the correlation, variance and the like among the electroencephalogram signals of each channel. The abnormal data elimination means that: detecting abnormal observation signals of a single stimulation unit according to the statistical characteristics, and marking the abnormal observation signals as invalid values; statistical properties here refer to the deviation, variance, etc. between the mean of the evoked responses of a single stimulation unit and the mean of the evoked responses of all stimulation units of the channel. By preprocessing the multichannel electroencephalogram signals, obviously abnormal data and invalid data can be removed, and the calculation amount of the subsequent data processing process is reduced.
S3, carrying out independent component analysis on the preprocessed multichannel electroencephalogram signals to obtain a plurality of independent components; in the step, an information maximization ICA algorithm or a maximum likelihood ICA algorithm is adopted to carry out independent component analysis, and a plurality of obtained independent components comprise components such as signals with different proportions, artifact interference and the like.
S4, taking the starting time of the stimulation units as a reference, obtaining independent component samples of each stimulation unit; the method comprises the following specific steps: the method specifically comprises the following steps: taking the initial time of the stimulation unit as a reference, and taking a period of time before and after the reference to form a time window of signal segmentation; and segmenting the independent components obtained after the independent component analysis according to the time window so as to obtain the independent component sample of each stimulation unit.
S5, carrying out time domain superposition averaging on the independent component samples to obtain an average evoked potential of each independent component; the method comprises the following specific steps: and segmenting each obtained independent component according to the starting time of each stimulation unit which is repeatedly stimulated, aligning the independent component samples obtained by segmentation according to the starting time of the stimulation units, and calculating the average value of all the independent component samples to be used as the average evoked potential of the independent component.
S6, after calculating the evoked energy gain of the average evoked potential of each independent component, screening and rejecting independent component samples according to a set criterion;
s7, selecting a detection sample from the independent component samples, calculating the total evoked energy gain of the detection sample and the total evoked energy gain of the average sample of the detection sample, and judging whether the selected detection sample has an evoked potential caused by the stimulation unit.
In this embodiment, after acquiring and preprocessing a multi-channel electroencephalogram signal, performing independent component analysis to obtain a plurality of independent components corresponding to the multi-channel electroencephalogram signal, so as to obtain independent cost samples of each stimulation unit based on the starting time of the stimulation unit, performing time-domain superposition averaging on the independent component samples to obtain an average evoked potential of each independent component, screening and removing the independent component samples according to a set criterion after the evoked energy gain of the average evoked potential of each independent component is based on, and then selecting a plurality of independent component samples as detection samples from the independent component samples to calculate a total evoked energy gain of the detection samples and a total evoked energy gain of the average sample of the detection samples, thereby determining whether the selected detection samples have evoked potentials caused by the stimulation units Screening and rejecting can detect whether any detection sample contains evoked potential, the calculation speed is high, the efficiency is high, the detection sensitivity is high, and whether the signal to be detected contains evoked potential can be rapidly detected.
In a further preferred embodiment, in step S6, the formula for calculating the induced energy gain is:
where EER denotes the evoked energy gain in dB, x (τ) denotes the signal to be calculated, τ denotes the time, τ -0 denotes the starting time of the stimulation unit, Δ t1B-a denotes the moment before the start of the stimulation unit, in particular the resting period of the evoked response, Δ t2D-c denotes the period of time after the stimulation unit has started the onset of evoked potentials.
Evoked energy gain EER essentially represents the energy gain of neural activity before and after stimulation. In the calculation formula, x (τ) represents the signal to be calculated, and in the present step, x (τ) refers to the average evoked potential of the individual components. Since the correlation between the average evoked potential of the independent components and the conventional evoked potential waveform is not clear, and the ratio of the included evoked components, spontaneous components, and artifact interference is unknown, for a section of signal x (τ) that may include an evoked potential, the signal components and energy distribution change due to the stimulation that may trigger the neural synchronization activity, and therefore, the evoked response energy gain can be expressed by the formula of the evoked energy gain EER defined in this step. The extracted independent components are screened based on the evoked energy gain EER, so that the evoked potential components and the gain of the enhanced reaction can be more effectively highlighted in the subsequent treatment process, and the detection capability of the evoked potential is obviously improved.
Further preferably, the criterion set in step S6 is a set criterion for screening the independent components, and for example, the independent components having an induced energy gain within a certain interval are screened and removed. Specifically, in step S6, it specifically is:
and after calculating the induced energy gain EER of the average induced potential of each independent component, screening out the independent component corresponding to the induced energy gain EER smaller than a preset threshold value as an interference component to remove.
For example, a preset threshold is set to 3dB, the independent component having an EER smaller than the preset threshold is determined as an interference component and discarded, and the remaining independent component is determined as a signal component, i.e., a valid independent component.
The smaller the evoked energy gain EER, the less the energy of the evoked potential contained, or the greater the disturbance, based on which the independent component of the larger disturbance can be removed. In the process of filtering and rejecting the independent components, various customized rules may be adopted for filtering and rejecting, for example, step S6, and may also be:
after the evoked energy gain EER of the average evoked potential of each independent component is calculated, all the calculated evoked energy gains EER are sequenced, and thus the independent component corresponding to the evoked energy gain EER with the lowest numerical value and the preset proportion is screened out and removed as an interference component. For example, after sorting, the independent component corresponding to the EER with the lowest value of 5% is taken as an interference component to be removed.
Because the acquired multi-channel electroencephalogram signals can be regarded as different weight mixtures of source signals from different brain areas, interference and the like, after independent component analysis is carried out in the step S3, the source signals can be separated from the multi-channel electroencephalogram signals to be detected to obtain a plurality of source signals, which are called as independent components in the application. In step S6, by removing the independent component with low signal-to-noise ratio, the number of subsequent calculations can be reduced, and the detection efficiency of the method can be improved.
In a further preferred embodiment, in step S7, the total induced energy gain is calculated by the following formula:
wherein GEER represents the total induced energy gain in dB, xj(τ) represents the signal to be calculated, j ═ 1,2, …, p, the number of the individual component samples remaining after screening, τ represents the time, τ ═ 0 represents the starting time of the stimulation unit, Δ t represents the number of the individual component samples remaining after screening, and1b-a denotes the moment before the start of the stimulation unit, in particular the resting period of the evoked response, Δ t2D-c denotes thornThe time period during which the evoked potential begins to appear after the firing of the cells.
The embodiment enhances the reserved independent components by defining the measurement parameter of the total evoked energy gain, and is more convenient for realizing the detection of the evoked potential.
Further as a preferred embodiment, the step S7 specifically includes:
s71, selecting a plurality of independent component samples as detection samples, and calculating the total induced energy gain of each detection sampleGEER j(j ═ 1, 2.., n), where n denotes the number of samples tested. In this step, x required for computing the GERj(τ) represents the individual component samples after screening of the test sample.
S72, performing superposition averaging on the selected detection samples, and calculating the total induced energy gain of the averaged average samples; in this step, after n detection samples are subjected to time domain superposition averaging to obtain an average sample, the total induced energy gain is calculated and recorded as mgER. In the step, the average sample after the superposition average of the selected detection samples is used as x required for calculating the GEERj(τ) is calculated by substituting the total induced energy gain into the calculation formula.
And S73, comparing the total evoked energy gain of the detection sample with the total evoked energy gain of the average sample, and judging whether the evoked potential caused by the stimulation unit exists in the detection sample. Specifically, whether the evoked potential caused by the stimulation unit exists in the detection sample is judged by judging whether the sum of the total evoked energy gain of the detection sample and the total evoked energy gain of the average sample is significantly larger than 0, and if the sum is larger than 0, the evoked potential caused by the stimulation unit exists in the detection sample is judged, otherwise, the sum is judged to be absent.
In a further preferred embodiment, in step S73, the check judgment is performed by a modified one-sided t-test method. The method specifically comprises the following steps: taking the total evoked energy gain value of each detection sample as the sample value of the t-test; taking the total induced energy gain of the average sample of the selected detection samples as the offset value of the sample of the t-test; a one-sided t-test is used to detect if the sum of the sample value and the offset value is greater than 0. Utensil for cleaning buttockBody pass judgment sample setGEER j(ii) whether + mGEER } is significantly greater than 0 is examined (p)<0.001). By adopting a modified unilateral t-test method for test judgment, whether the evoked potential exists can be accurately and quickly detected.
Method embodiment two
The embodiment is a detailed example of the first embodiment of the method, and particularly adopts a NeuroScan (synthesis amps2) electroencephalogram recording system to acquire multi-channel electroencephalogram signals of a tested person under the action of repeated stimulation. In this embodiment, the repetitive stimulation is composed of a plurality of identical transient stimulation units arranged in series and repeatedly, specifically, the repetitive stimulation composed of short pure tones with duration of 50ms is adopted, the stimulation interval is 1.01s, and the number of stimulation is about 500. In the test process, the placement mode of the test electrodes is as follows: adopting an international 10-20 system, 36 positions of the electrodes are selected, wherein the 36 positions are respectively as follows: FT7, FC5, FC3, FC1, FCZ, FC2, FC4, FC6, FT8, T7, C5, C3, C1, CZ, C2, C4, C6, T8, TP7, CP5, CP3, CP1, CPZ, CP2, CP4, CP6, TP8, P7, P5, P3, P1, PZ, P2, P4, P6, P8, wherein the reference electrode is at the midpoint of CZ and CPZ and Ground (GND) is at the midpoint of Fz and FpZ.
The method mainly comprises the following steps:
step 1, outputting repeated stimulation to a tested person through a professional earphone (such as an ER-3 plug-in earphone), recording observation signals of 36 channels by an acquisition system, and simultaneously recording output time of the repeated stimulation to be used as multi-channel electroencephalogram signals.
And 2, after preprocessing of filtering processing, interpolation processing, abnormal data elimination and the like is carried out on the acquired multi-channel electroencephalogram signals, independent component analysis is carried out on the preprocessed multi-channel electroencephalogram signals by adopting a maximum likelihood ICA algorithm, and 30 independent components are obtained through decomposition.
And 3, segmenting each independent component according to the starting time of each stimulation unit, specifically segmenting according to-400 ms to +600ms to obtain 565 independent component samples, aligning the independent component samples, and performing time domain superposition averaging to obtain the average evoked potential of each independent component. In this example, the average evoked potentials of the individual components obtained are shown in FIG. 2.
Step 5, calculating the obtained EERiSorting in an ascending manner, determining that the EER selection threshold is 3dB, and judging the independent component with the EER smaller than the threshold as the interference component to be discarded.
Step 6, arbitrarily selecting n (═ 11) detection samples, and calculatingGEER jJ is 1,2, …,11, and then the mger of n consecutive stimulation units is calculated. Performing unilateral t test, and obtaining a test resultGEER j+ mGEER } is significantly greater than 0 (p)<0.001), the presence of evoked potential is judged. In this embodiment, the total induced energy gain of the sample is detectedGEER jAs shown in fig. 3, line a represents the average energy at rest before stimulation, line b represents the average energy of the signal after stimulation, and it can be seen from fig. 3 that the average energy of the signal after stimulation is significantly greater than the average energy at rest before stimulation.
System embodiment
Referring to fig. 4, the present embodiment provides a multi-channel evoked potential monitoring system based on independent component analysis, including:
at least one processor 100;
at least one memory 200 for storing at least one program;
when executed by the at least one processor 100, cause the at least one processor 100 to implement the independent component analysis based multi-channel evoked potential monitoring method.
The multi-channel evoked potential monitoring system based on independent component analysis of the embodiment can execute the multi-channel evoked potential monitoring method based on independent component analysis provided by the method embodiment of the invention, can execute any combination implementation steps of the method embodiment, and has corresponding functions and beneficial effects of the method.
While the invention has been described with reference to a preferred embodiment, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (9)
1. The multichannel evoked potential detection method based on independent component analysis is characterized by comprising the following steps:
collecting multi-channel electroencephalogram signals of a tested person under the action of repeated stimulation; the repeated stimulation is formed by repeatedly arranging a plurality of identical transient stimulation units in series;
preprocessing the acquired multi-channel electroencephalogram signals;
carrying out independent component analysis on the preprocessed multi-channel electroencephalogram signals to obtain a plurality of independent components;
taking the initial time of the stimulation unit as a reference, and taking a period of time before and after the reference to form a time window of signal segmentation; segmenting the independent components obtained after the independent component analysis according to a time window so as to obtain independent component samples of each stimulation unit;
carrying out time domain superposition averaging on the independent component samples to obtain an average evoked potential of each independent component;
after the evoked energy gain of the average evoked potential of each independent component is calculated, independent component samples are screened and removed according to a set criterion;
selecting a detection sample from the independent component samples, calculating the total evoked energy gain of the detection sample and the total evoked energy gain of the average sample of the detection sample, and judging whether the selected detection sample has the evoked potential caused by the stimulation unit.
2. The multi-channel evoked potential monitoring method based on independent component analysis according to claim 1, wherein the step of selecting a detection sample from independent component samples, calculating a total evoked energy gain of the detection sample and a total evoked energy gain of an average sample of the detection sample, and determining whether the selected detection sample has an evoked potential induced by a stimulation unit specifically comprises:
selecting a plurality of independent component samples as detection samples, and calculating the total evoked energy gain of each detection sample;
performing superposition averaging on the selected detection samples, and then calculating the total induced energy gain of the averaged average samples;
and comparing the total evoked energy gain of the detection sample with the total evoked energy gain of the average sample, and judging whether the evoked potential caused by the stimulation unit exists in the detection sample.
3. The multi-channel evoked potential monitoring method based on independent component analysis according to claim 2, wherein the step of comparing the total evoked energy gain of the test sample with the total evoked energy gain of the average sample to determine whether the evoked potential caused by the stimulation unit exists in the test sample is performed by using a modified single-sided t-test method.
4. The multi-channel evoked potential detection method based on independent component analysis according to claim 1, wherein in the step of preprocessing the acquired multi-channel brain electrical signals, the preprocessing includes at least one of filtering processing, interpolation processing and abnormal data elimination.
5. The multi-channel evoked potential detection method based on independent component analysis according to claim 1, wherein in the step of obtaining a plurality of independent components after the independent component analysis of the preprocessed multi-channel electroencephalogram signal, the independent component analysis is performed by using an information maximization ICA algorithm or a maximum likelihood ICA algorithm.
6. The multi-channel evoked potential monitoring method based on independent component analysis according to claim 1, wherein the calculation formula of the evoked energy gain is as follows:
where EER denotes the evoked energy gain in dB, x (τ) denotes the signal to be calculated, τ -0 denotes the starting time of the stimulation unit, Δ t1B-a denotes the moment before the start of the stimulation unit, in particular the resting period of the evoked response, Δ t2D-c denotes the period of time after the stimulation unit has started the onset of evoked potentials.
7. The multi-channel evoked potential monitoring method based on independent component analysis according to claim 1, wherein the step of screening and rejecting independent component samples according to a set criterion after calculating the evoked energy gain of the average evoked potential of each independent component is specifically as follows:
and after the evoked energy gain of the average evoked potential of each independent component is calculated, screening out independent component samples corresponding to the evoked energy gain smaller than a preset threshold value as interference components to be removed.
8. The multi-channel evoked potential monitoring method based on independent component analysis according to claim 1, wherein the total evoked energy gain is calculated by the formula:
wherein GEER represents the total induced energy gain in dB, xj(τ) represents the signal to be calculated, j ═ 1,2, …, p, the serial number of the individual component samples remaining after screening, τ ═ 0 represents the starting time of the stimulation unit, Δ t1B-a represents a time before the stimulation unit is initiated, in particular an evoked response rest period,Δt2d-c denotes the period of time after the stimulation unit has started the onset of evoked potentials.
9. Multichannel evoked potential monitoring system based on independent component analysis, characterized by comprising:
at least one processor;
at least one memory for storing at least one program;
when executed by the at least one processor, cause the at least one processor to implement the independent component analysis based multi-channel evoked potential monitoring method according to any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811484564.1A CN109512394B (en) | 2018-12-06 | 2018-12-06 | Multi-channel evoked potential detection method and system based on independent component analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811484564.1A CN109512394B (en) | 2018-12-06 | 2018-12-06 | Multi-channel evoked potential detection method and system based on independent component analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109512394A CN109512394A (en) | 2019-03-26 |
| CN109512394B true CN109512394B (en) | 2021-07-13 |
Family
ID=65794280
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811484564.1A Expired - Fee Related CN109512394B (en) | 2018-12-06 | 2018-12-06 | Multi-channel evoked potential detection method and system based on independent component analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109512394B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201005696Y (en) * | 2007-07-17 | 2008-01-16 | 中国人民解放军南京军区南京总医院 | Brain Evoked Electrical Signal Tester Based on Independent Component Analysis Technology |
| CN103340623A (en) * | 2013-06-27 | 2013-10-09 | 南方医科大学 | Method for extracting evoked potentials under high stimulation ratio |
| CN106344011A (en) * | 2016-10-21 | 2017-01-25 | 电子科技大学 | Evoked electroencephalogram signal extraction method based on factor analysis |
| CN107280665A (en) * | 2017-06-09 | 2017-10-24 | 上海海神医疗电子仪器有限公司 | Brain evoked potential signal acquisition method |
| CN108564011A (en) * | 2017-08-01 | 2018-09-21 | 南京邮电大学 | A kind of personal identification method that normal form being presented based on brain electricity Rapid Speech |
| CN108784693A (en) * | 2018-06-15 | 2018-11-13 | 北京理工大学 | P300 single-trial extraction technologies based on independent component analysis and Kalman smoothing |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080167571A1 (en) * | 2006-12-19 | 2008-07-10 | Alan Gevins | Determination of treatment results prior to treatment or after few treatment events |
-
2018
- 2018-12-06 CN CN201811484564.1A patent/CN109512394B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201005696Y (en) * | 2007-07-17 | 2008-01-16 | 中国人民解放军南京军区南京总医院 | Brain Evoked Electrical Signal Tester Based on Independent Component Analysis Technology |
| CN103340623A (en) * | 2013-06-27 | 2013-10-09 | 南方医科大学 | Method for extracting evoked potentials under high stimulation ratio |
| CN106344011A (en) * | 2016-10-21 | 2017-01-25 | 电子科技大学 | Evoked electroencephalogram signal extraction method based on factor analysis |
| CN107280665A (en) * | 2017-06-09 | 2017-10-24 | 上海海神医疗电子仪器有限公司 | Brain evoked potential signal acquisition method |
| CN108564011A (en) * | 2017-08-01 | 2018-09-21 | 南京邮电大学 | A kind of personal identification method that normal form being presented based on brain electricity Rapid Speech |
| CN108784693A (en) * | 2018-06-15 | 2018-11-13 | 北京理工大学 | P300 single-trial extraction technologies based on independent component analysis and Kalman smoothing |
Non-Patent Citations (1)
| Title |
|---|
| Processing and analysis of magnetoencephalographic data based on independent component analysis;Wang bin et al;《中国临床康复》;20050728;第9卷(第28期);254-256 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109512394A (en) | 2019-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210290141A1 (en) | Electrocardiogram signal detection | |
| US8112147B2 (en) | Method and apparatus for generating determination indexes for identifying ECG interfering signals | |
| CN110338786B (en) | A method, system, device and medium for identifying and classifying epileptiform discharges | |
| US20090177101A1 (en) | Method of fetal and maternal ecg identification across multiple epochs | |
| CN110575164B (en) | Method for removing artifacts of electroencephalogram signal and computer-readable storage medium | |
| CN108814590A (en) | A kind of detection method and its ecg analysis method of Electrocardiograph QRS Wave group | |
| CN106236080B (en) | A method for eliminating EMG noise in EEG signal based on multi-channel | |
| CN107811631A (en) | Electrocardiosignal method for evaluating quality | |
| US6574573B1 (en) | Spectrum analysis and display method time series | |
| CN110070013B (en) | An Intelligent Evaluation Method for Noise Pollution Degree and Category of ECG Signals | |
| JPH0924031A (en) | Single nerve action potential measuring device | |
| CN104644160A (en) | Electrocardiogram pseudo-difference signal identification method and device | |
| CN103720471B (en) | A Method of Removing Oculograph Artifacts Based on Factor Analysis | |
| CN114587381A (en) | Spike detection method based on multi-channel EEG intelligent screening and weighted sample generation | |
| CN105574348A (en) | Heartbeat period acquiring method and device based on BCG signal | |
| CN109512394B (en) | Multi-channel evoked potential detection method and system based on independent component analysis | |
| CN102988041A (en) | Selective cardiac-magnetic signal averaging method in signal noise suppression | |
| CN104983415B (en) | The heart claps the method and device of identification | |
| CN108694355A (en) | A signal processing method and device | |
| CN111887811B (en) | Brain abnormal discharge detection method and system based on electroencephalogram signal characteristics | |
| EP0848335B1 (en) | Method and apparatus for QRS detection | |
| CN114027854B (en) | BECT spike detection method based on optimal template matching and morphological feature extraction | |
| CN110507299A (en) | Heart rate signal detection device and method | |
| CN116942174A (en) | Fetal ECG signal extraction method based on template subtraction and independent component analysis | |
| Parsaei et al. | A method for detecting and editing MUPTs contaminated by false classification errors during EMG signal decomposition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210713 Termination date: 20211206 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |