CN113545791A - Background activity automatic identification method and system based on resting electroencephalogram - Google Patents
Background activity automatic identification method and system based on resting electroencephalogram Download PDFInfo
- Publication number
- CN113545791A CN113545791A CN202110819422.1A CN202110819422A CN113545791A CN 113545791 A CN113545791 A CN 113545791A CN 202110819422 A CN202110819422 A CN 202110819422A CN 113545791 A CN113545791 A CN 113545791A
- Authority
- CN
- China
- Prior art keywords
- area
- electroencephalogram
- module
- average
- average time
- 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.)
- Pending
Links
- 230000000694 effects Effects 0.000 title claims abstract description 35
- 238000000034 method Methods 0.000 title claims abstract description 31
- 230000000284 resting effect Effects 0.000 title claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 18
- 238000004458 analytical method Methods 0.000 claims abstract description 17
- 230000000007 visual effect Effects 0.000 claims abstract description 17
- 238000012795 verification Methods 0.000 claims abstract description 6
- 210000004556 brain Anatomy 0.000 claims description 45
- 230000002146 bilateral effect Effects 0.000 claims description 37
- 230000005856 abnormality Effects 0.000 claims description 33
- 238000004364 calculation method Methods 0.000 claims description 19
- 230000002159 abnormal effect Effects 0.000 claims description 18
- 238000003745 diagnosis Methods 0.000 claims description 17
- 238000013500 data storage Methods 0.000 claims description 14
- 238000013523 data management Methods 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 13
- 238000007781 pre-processing Methods 0.000 claims description 13
- 230000002123 temporal effect Effects 0.000 claims description 11
- 239000003086 colorant Substances 0.000 claims description 4
- 230000004424 eye movement Effects 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 230000001936 parietal effect Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract description 2
- 239000000284 extract Substances 0.000 abstract description 2
- 238000012800 visualization Methods 0.000 description 7
- 230000001314 paroxysmal effect Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 3
- 230000005611 electricity Effects 0.000 description 2
- 206010015037 epilepsy Diseases 0.000 description 2
- 230000001037 epileptic effect Effects 0.000 description 2
- 230000001787 epileptiform Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 208000020401 Depressive disease Diseases 0.000 description 1
- 208000012902 Nervous system disease Diseases 0.000 description 1
- 208000018737 Parkinson disease Diseases 0.000 description 1
- 208000013738 Sleep Initiation and Maintenance disease Diseases 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 208000026106 cerebrovascular disease Diseases 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000000537 electroencephalography Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000005714 functional activity Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 206010022437 insomnia Diseases 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000004393 prognosis Methods 0.000 description 1
- 210000004761 scalp Anatomy 0.000 description 1
- 238000006467 substitution 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]
- A61B5/372—Analysis of electroencephalograms
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Psychology (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Psychiatry (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a background activity automatic identification method and a system based on resting state electroencephalogram, which automatically carry out regional identification on electroencephalogram signals of eye opening and eye closing states of a patient through the system, respectively extract electroencephalogram characteristics of each frequency of each region, analyze and compare the electroencephalogram characteristics by referring to a standardized range, and finally output an identification and analysis result as a visual report, thereby facilitating quick manual verification, editing and modifying on line when necessary, realizing the high efficiency of electroencephalogram automatic interpretation, ensuring the accuracy of result analysis, providing possibility for efficiently and quickly interpreting high-dimensional data information from original two-dimensional electroencephalogram, and providing a potential technical means for non-electroencephalogram professionals who are not familiar with EEG data analysis and processing. The invention can realize the automation of the identification process and improve the identification efficiency, thereby reducing the heavy and repeated mechanized labor, facilitating the medical staff to input the analysis and judgment of the result report, improving the working efficiency and saving the labor cost.
Description
Technical Field
The invention relates to the field of resting state electroencephalogram analysis and identification, in particular to a method and a system for automatically identifying background activities based on resting state electroencephalogram.
Background
Electroencephalogram (EEG) is a non-invasive, non-invasive technique for collecting electrical signals from neurons in the brain, which can reflect the electrical activity of the brain neurons in their cluster points with extremely high time accuracy, and has the advantages of high time resolution, short examination time, low cost, convenience and easy implementation, etc., compared with invasive detection methods which consume long time for examination. In recent years, electroencephalograms become powerful tools for evaluating changes of brain functional activities, along with the development of brain functional imaging technology, electroencephalogram signals have close relations with nervous system diseases such as epilepsy, cerebrovascular diseases, insomnia, anxiety depression, parkinson disease and the like, and high-dimensional electroencephalogram characteristics are extracted through energy and brain connection analysis of an electrode level and a source level, so that the electroencephalogram has certain potential for auxiliary diagnosis, curative effect prediction and prognosis judgment of partial diseases.
Limitations of artificial processing of electroencephalogram signals: the interpretation of the background activities of the electroencephalogram signals needs to have professional training for more than several years and certain clinical experience, the characteristic extraction of the electroencephalogram signals needs to be familiar with software operation procedures of Matlab and EEGlab, the electroencephalogram analysis interpretation cannot be widely popularized, and the problem of limitation of users is solved by the one-key automatic processing and analyzing system.
Manual identification consumes time and labor: electroencephalography is most commonly used in clinical applications to aid in the diagnosis of epilepsy. After the electroencephalogram of the patient is collected, the background activity of the electroencephalogram and epileptic waves are identified manually. Epileptic waves have a characteristic good identification, but identification of background activity tends to be labor intensive.
The manual identification of the electroencephalogram requires the electroencephalogram specialist to carefully perform identification and examination on data every 10 seconds, generally, the electroencephalogram recorded for 8 hours takes several hours, and the electroencephalogram doctor with rich experience needs to be cultured for a long time. The electroencephalogram data volume is large, the processing time is long, the requirements on hardware and operators are high, and the current manual EEG processing mode limits the data analysis efficiency.
Manual identification is subjective and inconsistent: because the judgment of EEG results lacks objective quantitative indicators, it is largely influenced by the level of personal experience and comprehensive analysis capabilities of doctors and technicians. The reader's understanding of the diagnostic criteria is inconsistent and often has varying degrees of subjectivity and propensity to cause inconsistency in the determination of EEG results. The domestic report shows that the coincidence rate of the EEG results judged by senior citizens who took over 10 years from the EEG specialty is 73%, and that between senior citizens and senior citizens who took less than 5 years from the present specialty is 64%.
Machine learning classification and identification based on electroencephalogram signal features has high efficiency, but lacks uniform specific feature standards, and has poor generalization capability: the method has the advantages that the original brain electricity is subjected to time-frequency conversion, the characteristics are extracted, the machine learning model is combined for training, the recognition efficiency is high, the accuracy, the specificity and the generalization capability need large sample verification, and the recognition result is greatly influenced by the sample trained by the model.
Accordingly, the prior art is deficient and needs improvement.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a background activity automatic identification method and system based on resting state electroencephalogram.
The technical scheme of the invention is as follows: the background activity automatic identification method based on the resting state electroencephalogram comprises the following steps:
step 1: acquiring a resting state electroencephalogram rsEEG signal of a patient by utilizing electroencephalogram signal acquisition equipment;
step 2: preprocessing resting state electroencephalogram rsEEG signal data of a patient in an eye closing state and an eye opening state for 3min respectively by adopting an ARTIST full-automatic denoising method;
and step 3: respectively calculating and extracting three electroencephalogram characteristics of average time ratio, average amplitude and bilateral amplitude difference of different brain areas of each frequency band;
and 4, step 4: analyzing and interpreting the extracted electroencephalogram characteristics, and sequentially and respectively judging in the characteristic conditions of severe background abnormality, moderate background abnormality, mild background abnormality, boundary electroencephalogram and normal electroencephalogram;
and 5: automatically generating a visual report according to the judgment result, presenting a conclusion of electroencephalogram background activity identification, and the quantified electroencephalogram characteristics and reference ranges of different brain areas of each frequency band;
step 6: and (5) carrying out manual quick verification according to the visual report generated in the step (5), and modifying and editing the report content with deviation by manual work if necessary.
Further, the specific steps of step 2 are as follows:
step 2.1: respectively keeping the eyes of the patient in a closed state and an open state for a period of time, and respectively selecting 3min of resting state electroencephalogram rsEEG signal data from the collected signals;
step 2.2: removing direct current drift in resting state electroencephalogram rsEEG signal data;
step 2.2: removing eye movement interference in signal data selected by the eye closing state and the eye opening state respectively;
step 2.3: reducing the sampling rate to 250Hz, adjusting the band-pass filtering to 1-45Hz, and replacing a bad channel;
step 2.4: segmenting data, wherein the time length of each segment of data is 2 s;
step 2.5: removing the segmented bad channel and performing interpolation;
step 2.6: and (4) after analyzing the independent components, removing pseudo-error components and adopting an average reference value.
Further, the specific steps of step 3 are as follows:
step 3.1: dividing the whole brain into four areas, wherein the first area comprises a frontal area and an anterior temporal area, the second area comprises a central area and a medial temporal area, the third area comprises a parietal area, and the fourth area comprises an occipital area and a posterior temporal area;
step 3.2: according to a frequency calculation method, signals with different frequencies appearing in each channel of an electroencephalogram along with time are identified, the frequencies are divided into five frequency bands which are respectively alpha waves, beta waves, gamma waves, delta waves and theta waves, the identified waves with different frequencies are marked by different colors, and then the time length of the waves appearing in each frequency band is compared with the total time length of the signals, namely the average time ratio of each frequency band;
step 3.3: calculating the peak-peak value of each identified brain wave, storing the peak-peak value as the amplitude of each wave, wherein the average peak-peak value of all brain waves under each frequency band is the average amplitude of the frequency band;
step 3.4: dividing the difference value of the average amplitude values of a certain frequency band in the left hemisphere and the right hemisphere by the low value of the average amplitude values of the same frequency band in the left hemisphere and the right hemisphere by taking the brain center line as a boundary, reflecting the symmetry of the left hemisphere and the right hemisphere, and obtaining the bilateral amplitude difference;
step 3.5: and (4) extracting the average time ratio, the average amplitude and the two-side amplitude difference electroencephalogram characteristics of each region in each frequency band according to the steps 3.2-3.4.
Further, the diagnosis criteria of the normal electroencephalogram in step 4 are as follows:
the average time of delta wave in each region accounts for 0-2%, and the average amplitude is 0-50 muV;
the average time proportion of theta wave in each zone is 0-15%, and the average amplitude is 0-50 muV;
the average time proportion of alpha waves in the first area and the second area is 40-100%, the average time proportion in the third area is 45-100%, the average time proportion in the fourth area is 50-100%, and the average amplitude of each area is 0-100 μ V;
fourthly, the average time proportion of the beta wave in each area is 0 to 40 percent, and the average amplitude is 0 to 20 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 0-30%, the bilateral amplitude difference in the third area is 0-40%, and the bilateral amplitude difference in the fourth area is 0-100%.
Further, the diagnostic criteria of the bounded electroencephalogram in the step 4 are as follows:
the restricted electroencephalogram is characterized by any one of the following abnormal manifestations:
the average time of delta wave in each region accounts for 2-5%, and the average amplitude is more than 50 muV;
the average time proportion of theta wave in each area is 15-30%, and the average amplitude is more than 50 mu V;
the average time proportion of alpha waves in the first area and the second area is 10-40%, the average time proportion in the third area is 10-45%, the average time proportion in the fourth area is 10-50%, and the average amplitude of each area is more than 100 mu V;
the average time proportion of the beta wave in each area is 40-100%, and the average amplitude is 20-50 μ V;
the amplitude difference of the two sides of each frequency band in the first area and the second area is 30-50%, the amplitude difference of the two sides of the third area is 40-60%, and the amplitude difference of the two sides of the fourth area is 100-150%.
Further, the diagnosis criteria of mild background abnormality in step 4 are as follows:
the mild background abnormality is any one of the following abnormal manifestations:
the average time of delta waves in each region accounts for 5-10%;
the average time of theta wave in each zone accounts for 30-75%;
③ 5 to 10 percent of the average time of each area of the alpha wave;
the average amplitude of the beta wave in each area is more than 50 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 50-75%, the bilateral amplitude difference in the third area is 60-85%, and the bilateral amplitude difference in the fourth area is 150-200%.
Further, the diagnosis criteria of moderate background abnormality in step 4 are as follows:
the abnormal expression of any one of the following is moderate background abnormality:
the average time of delta waves in each region accounts for 10-95%;
the average time of theta wave in each zone accounts for 75-95%;
③ the average time of each area of the alpha wave is 0 to 5 percent;
and the bilateral amplitude difference of each frequency band in the first area and the second area is more than 75%, the bilateral amplitude difference in the third area is more than 85%, and the bilateral amplitude difference in the fourth area is more than 200%.
Further, the diagnosis criteria of severe background abnormality in the step 4 are as follows:
the severe background abnormality is represented by any one of the following abnormalities:
the average time of delta waves in each region accounts for 95-100%;
the average time of theta wave in each zone is 95-100%.
The invention also provides a background activity automatic identification system, which comprises: interconnect's data management module, data storage module and data processing analysis module, the data management module includes: the data editing and modifying module and the information input module, the data processing and analyzing module comprises: preprocessing module, frequency identification and color identification module, brain district division module, feature extraction calculation module, comparison module and visual module, data management module, preprocessing module, frequency identification and color identification module and brain district division module are connected with feature extraction calculation module, feature extraction calculation module is connected with comparison module, comparison module is connected with visual module, visual module is connected with data storage module, data storage module is connected with data editing and modifying module.
Furthermore, the data management module is a computer group, the data storage module is a disk array memory, and the data processing and analyzing module is a processor.
By adopting the scheme, the electroencephalogram characteristics of different brain areas and different frequencies are automatically calculated and extracted through the system, the electroencephalogram characteristics comprise average time ratio, average amplitude and bilateral amplitude difference, the electroencephalogram characteristics are analyzed and compared according to a standardized range, and finally, a visual report is output, so that classification of electroencephalogram background activities is realized. The system can be manually checked quickly, and can be modified and edited on line when necessary, the accuracy of result analysis is guaranteed while the automatic identification efficiency of the system is met, and the application flexibility of the system is increased, so that the possibility of efficiently extracting high-dimensional data information from the original two-dimensional electroencephalogram is provided, and a potential electroencephalogram analysis technical means is provided for non-electroencephalogram professional doctors and scientific researchers. Compared with the traditional manual identification processing mode, the method realizes the automation of the electroencephalogram background activity identification process, separates electroencephalographs from the complex and time-consuming background activity identification and description work, so that report result auditing and paroxysmal abnormal activity deep interpretation are invested, the working efficiency is improved, and the labor cost is saved.
Drawings
FIG. 1 is a block flow diagram of the present invention.
Fig. 2 is a schematic diagram of brain electrical partition.
FIG. 3 is a schematic diagram of color mark recognition of different frequencies of electroencephalogram signals.
Fig. 4 shows the average time ratio, average amplitude, and bilateral amplitude difference of each frequency band wave in each brain region and their reference ranges.
FIG. 5 is a table of automated electroencephalogram report grading reference criteria.
Fig. 6 is a system connection block diagram of the present invention.
Detailed Description
The invention is described in detail below with reference to the figures and the specific embodiments.
Referring to fig. 1, the present invention provides a background activity automatic identification method based on resting state electroencephalogram, including the following steps:
step 1: and acquiring a resting state electroencephalogram rsEEG signal of the patient by utilizing electroencephalogram signal acquisition equipment.
Step 2: and (3) preprocessing resting state electroencephalogram rsEEG signal data of the patient in a closed eye state and an open eye state for 3min respectively by adopting an ARTIST full-automatic denoising method. The method comprises the following specific steps:
step 2.1: respectively keeping the eyes of the patient in a closed state and an open state for a period of time, and respectively selecting 3min of resting state electroencephalogram rsEEG signal data from the collected signals;
step 2.2: removing direct current drift in resting state electroencephalogram rsEEG signal data;
step 2.2: removing eye movement interference in signal data selected by the eye closing state and the eye opening state respectively;
step 2.3: reducing the sampling rate to 250Hz, adjusting the band-pass filtering to 1-45Hz, and replacing a bad channel;
step 2.4: segmenting data, wherein the time length of each segment of data is 2 s;
step 2.5: removing the segmented bad channel and performing interpolation;
step 2.6: and (4) after analyzing the independent components, removing pseudo-error components and adopting an average reference value.
And selecting resting state electroencephalogram rsEEG signal data of 3min for keeping the eyes of the patient in a closed state and an eyes of the patient in an open state, so that the stability of the signal data is ensured, and the subsequent identification is facilitated. Clutter interference in the selected signal data is removed by adopting an ARTIST full-automatic denoising method, so that the identification accuracy of the signal data is improved.
And step 3: and respectively extracting three electroencephalogram characteristics of average time ratio, average amplitude and bilateral amplitude difference of different brain areas of each frequency band. The method comprises the following specific steps:
step 3.1: referring to fig. 2, the whole brain is divided into four regions, wherein the first region includes a frontal region and an anterior temporal region, the second region includes a central region and a medial temporal region, the third region includes a parietal region, and the fourth region includes an occipital region and a posterior temporal region. The frontal channel includes leftFchan (F7, F3) and rightFchan (F8, F4). The anterior temporal region includes FT7, FT 8. The central zone channel includes leftcch (C3) and rightcch (C4). The medial temporal region includes T7, T8. The top channel includes leftPchan (P3) and rightPchan (P4). The occipital channels include leftOchan (O1) and rightOchan (O2). The posterior head portion includes the occipital region and the anterior head portion includes the other regions described above except the occipital region.
Step 3.2: referring to fig. 3, according to the frequency calculation method, signals with different frequencies appearing in each channel of the electroencephalogram over time are identified, the frequencies are divided into five frequency bands, namely, α waves, β waves, γ waves, δ waves and θ waves, the identified waves with different frequencies are marked with different colors, and then the time length of the waves appearing in each frequency band is compared with the total time length of the signals, which is the average time ratio of each frequency band. Wherein the frequency range of the delta wave is 1-4Hz, the frequency range of the theta wave is 4-8Hz, the frequency range of the alpha wave is 8-13Hz, the frequency range of the beta wave is 13-30Hz, the frequency range of the gamma wave is 30-45Hz, and the waves with the frequency more than 45Hz are not included in the calculation. Frequency refers to the number of repetitions of the same periodic waveform in a second, or a wave that is a fraction of a second in Hz or cycles/second (c/s), and the frequency range for scalp EEG analysis is typically 0.1-100Hz, dividing the EEG frequency into five bands, delta, theta, alpha, beta, and gamma. The measurement of frequency is from the trough of any brain wave to the next trough, or from any peak to the next peak.
Specifically, on the preprocessed time series signal, the frequency band range to which each visible wave belongs is calculated: the code automatically retrieves the time taken for each wave, i.e. the length of time from trough-peak-trough. The wave trough is defined as the local minimum value appearing first, the wave crest is the maximum value appearing later, and when the next wave trough appears and the amplitude of the descending branch is larger than 1/2 of the ascending branch, a brain wave is judged. When the time occupied by brain waves is 0.25s-1s, it is marked as delta waves. When the time occupied by brain waves is between 0.125s and 0.25s, the mark is theta waves. Similarly, alpha waves are between 0.076s and 0.125s, beta waves are between 33ms and 76ms, and gamma waves are between 22ms and 33 ms. Waves with times less than 22ms do not fit into the calculation. The identified waves of different frequencies are marked with different colors. And then, comparing the time length of the wave appearing in each frequency band with the total time length of the signal, namely, the average time ratio of each frequency band.
Step 3.3: and calculating the peak-peak value of each identified brain wave, storing the peak-peak value as the amplitude of each wave, and obtaining the average peak-peak value of all brain waves in each frequency band as the average amplitude of the frequency band.
The amplitude is a voltage for describing brain waves, the potential difference between any two electrodes is measured in microvolts (μ V) (1 μ V is 10-6V), the height of the voltage is measured by scaling the voltage with an amplifier, and the voltage value can be determined by the height (mm) of the brain waves.
The peak-to-peak value of each identified brain wave is calculated and stored as the amplitude of each wave. The average peak-peak value of all brain waves under each frequency band is the average amplitude of the frequency band. For example, 3 θ waves are identified within 10s, and their peak-to-peak values are 20 μ V, 25 μ V, and 30 μ V, respectively, so that the average amplitude of the θ wave band is 25 μ V.
Step 3.4: dividing the difference value of the average amplitude values of a certain frequency band in the left hemisphere and the right hemisphere by the low value of the average amplitude values of the same frequency band in the left hemisphere and the right hemisphere by taking the brain center line as a boundary, and reflecting the symmetry of the left hemisphere and the right hemisphere, namely the difference of the amplitude values on both sides;
step 3.5: referring to fig. 4, the average time ratio, the average amplitude and the bilateral amplitude difference electroencephalogram characteristics of each region in each frequency band are extracted according to the steps 3.2 to 3.4.
And 4, step 4: referring to fig. 5, the extracted electroencephalogram features are analyzed and interpreted, and sequentially and respectively distinguished in the conditions of severe background abnormality, moderate background abnormality, mild background abnormality, borderline electroencephalogram, and normal electroencephalogram features.
The electroencephalogram diagnosis system is based on an adult electroencephalogram diagnosis reference standard, and takes three characteristics of average time ratio, average amplitude and bilateral amplitude difference as core parameters to automatically interpret electroencephalograms, firstly, whether the acquired electroencephalograms of patients are normal electroencephalograms is interpreted, and if the acquired electroencephalograms are not normal, whether severe background abnormality, moderate background abnormality, mild background abnormality and borderline electroencephalogram characteristic conditions are sequentially interpreted.
Further, the diagnosis criteria of the normal electroencephalogram in step 4 are as follows:
the average time of delta wave in each region accounts for 0-2%, and the average amplitude is 0-50 muV;
the average time proportion of theta wave in each zone is 0-15%, and the average amplitude is 0-50 muV;
the average time proportion of alpha waves in the first area and the second area is 40-100%, the average time proportion in the third area is 45-100%, the average time proportion in the fourth area is 50-100%, and the average amplitude of each area is 0-100 μ V;
fourthly, the average time proportion of the beta wave in each area is 0 to 40 percent, and the average amplitude is 0 to 20 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 0-30%, the bilateral amplitude difference in the third area is 0-40%, and the bilateral amplitude difference in the fourth area is 0-100%. Further, the diagnostic criteria of the bounded electroencephalogram in the step 4 are as follows:
the restricted electroencephalogram is characterized by any one of the following abnormal manifestations:
the average time of delta wave in each region accounts for 2-5%, and the average amplitude is more than 50 muV;
the average time proportion of theta wave in each area is 15-30%, and the average amplitude is more than 50 mu V;
the average time proportion of alpha waves in the first area and the second area is 10-40%, the average time proportion in the third area is 10-45%, the average time proportion in the fourth area is 10-50%, and the average amplitude of each area is more than 100 mu V;
the average time proportion of the beta wave in each area is 40-100%, and the average amplitude is 20-50 μ V;
the amplitude difference of the two sides of each frequency band in the first area and the second area is 30-50%, the amplitude difference of the two sides of the third area is 40-60%, and the amplitude difference of the two sides of the fourth area is 100-150%. Further, the diagnosis criteria of mild background abnormality in step 4 are as follows:
the mild background abnormality is any one of the following abnormal manifestations:
the average time of delta waves in each region accounts for 5-10%;
the average time of theta wave in each zone accounts for 30-75%;
③ 5 to 10 percent of the average time of each area of the alpha wave;
the average amplitude of the beta wave in each area is more than 50 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 50-75%, the bilateral amplitude difference in the third area is 60-85%, and the bilateral amplitude difference in the fourth area is 150-200%. Further, the diagnosis criteria of moderate background abnormality in step 4 are as follows:
the abnormal expression of any one of the following is moderate background abnormality:
the average time of delta waves in each region accounts for 10-95%;
the average time of theta wave in each zone accounts for 75-95%;
③ the average time of each area of the alpha wave is 0 to 5 percent;
and the bilateral amplitude difference of each frequency band in the first area and the second area is more than 75%, the bilateral amplitude difference in the third area is more than 85%, and the bilateral amplitude difference in the fourth area is more than 200%.
Further, the diagnosis criteria of severe background abnormality in the step 4 are as follows:
the severe background abnormality is represented by any one of the following abnormalities:
the average time of delta waves in each region accounts for 95-100%;
the average time of theta wave in each zone is 95-100%.
And 5: automatically generating a visual report according to the judgment result, presenting a conclusion of electroencephalogram background activity identification, and the quantified electroencephalogram characteristics and reference ranges of different brain areas of each frequency band;
the system only carries out hierarchical identification on the adult EEG background abnormity, but does not comprise paroxysmal abnormity (epileptiform discharge), and the epileptiform discharge needs a professional in artificial verification to individually and specifically describe local paroxysmal abnormal activity. Comparing the extracted value with a quantitative reference range of the classification recognition of the electroencephalogram background activity and classifying: (1) normal electroencephalogram, all values listed in the table need to be met; (2) otherwise, only any one of the values listed in the table need be met. And then automatically generating a visual report according to calculation comparison, carrying out hierarchical identification on the electroencephalogram background activity, describing the report content in detail, displaying the automatically calculated individualized electroencephalogram characteristics in a digital form, listing normal reference values, wherein red represents an increase of the normal value, and blue represents a decrease of the normal value.
Step 6: and (5) carrying out manual quick verification according to the visual report generated in the step (5), and modifying and editing the report content with deviation by manual work if necessary. Electroencephalographs can perform rapid manual examination according to the visualization report result, online modification and adjustment can be performed on report contents when necessary, and paroxysmal abnormal activities can be specifically described, so that the flexibility and the convenience of the system application are improved.
According to experiments, compared with the manual identification result, in 124 abnormal electroencephalogram reports, 108 abnormal electroencephalogram reports are manually identified, and 124 abnormal electroencephalogram reports are identified by the system; of 170 patients who were manually reported as normal electroencephalograms, 170 were manually identified as normal and 169 were identified as normal by the system, and then analysis 1 was conducted to identify the cause of non-normality mainly due to the imperfect pre-treatment of eye movement artifacts, and 100% was identified as normal after re-pretreatment.
Sensitivity (SEN): the true positive rate, the percentage of actual abnormal patients diagnosed, is calculated as follows:
manual identification: (diagnosis positive patients by manual identification ÷ actual total positive patients) × (108 ÷ 124) × (100)% -87.10%.
Automatic identification: (automatically identified diagnostic positive patients ÷ actual total positive patients) × (124 ÷ 124) × (100%).
Specificity (SPE): the true negative rate, the actual disease-free is correctly judged as the percentage of disease-free according to the diagnostic standard, and the calculation formula is as follows:
manual identification: (human-identified diagnostic negative patients ÷ actual total negative patients) × (170 ÷ 170) × (100%).
Automatic identification: (automatically identified diagnostic negative patients ÷ actual total negative patients) × (169 ÷ 170) × (100%) 99.4%.
Equilibrium accuracy (BAC): the influence of the bias condition of the positive and negative sample numbers in the signal set on the accuracy is considered, and the calculation formula is as follows:
manual identification: 1/2 × (87.10% + 100%) 100% ═ 93.6%.
Automatic identification: 1/2: (100% + 99.4%) 100% ═ 99.7%.
From the data, the result of the automatic analysis of the system has better overall effect in the classification of normal and abnormal brain electricity, and has good sensitivity, specificity and balance accuracy index. Meanwhile, the data prove that the automatic identification method provided by the invention is obviously improved in the aspect of the accuracy of automatic identification compared with manual identification, and the effectiveness of the automatic identification method provided by the invention is proved.
Referring to fig. 6, the present invention further provides an automatic background activity recognition system, which includes: interconnect's data management module, data storage module and data processing analysis module, the data management module includes: the data editing and modifying module and the information input module, the data processing and analyzing module comprises: the device comprises a preprocessing module, a frequency identification and color identification module, a brain region division module, a feature extraction and calculation module, a comparison module and a visualization module. The system comprises a data management module, a preprocessing module, a frequency identification and color identification module and a brain region division module, wherein the data management module, the preprocessing module, the frequency identification and color identification module and the brain region division module are connected with a feature extraction calculation module, the feature extraction calculation module is connected with a comparison module, the comparison module is connected with a visualization module, the visualization module is connected with a data storage module, and the data storage module is connected with a data editing and modifying module.
The data management module is a computer set, the data storage module is a disk array memory, and the data processing and analyzing module is a processor.
The information entry module can be convenient for medical personnel to look over patient information to edit patient's basic information, such as outpatient service number, name, disease type. If a newly added patient exists, the relevant information of the patient can be input through the information input module.
The data processing and analyzing module can realize the processing of the electroencephalogram data. The preprocessing module can be used for preprocessing the resting state electroencephalogram rsEEG signal data, removing interference segments in the signal data and improving the identification accuracy. The frequency identification and color identification module can identify and identify signals with different frequencies appearing in each channel of the electroencephalogram along with time. The brain region dividing module can divide the whole electroencephalogram into four brain regions. The output results of the preprocessing module, the frequency identification and color identification module and the brain area division module are used as input data of the characteristic extraction and calculation module, so that three electroencephalogram characteristics of average time ratio, average amplitude and bilateral amplitude difference of different brain areas of each frequency band are extracted and calculated. And then the output result of the feature extraction and calculation module is input into the comparison module and then is sent to the visualization module, so that the comparison and identification result is output as a visualization report, and the visualization report is stored in the data storage module. The medical staff can check the visual report stored in the data storage module through the computer set, and meanwhile, the report content can be edited and modified when necessary through the data editing and modifying module to quickly and manually check the report, so that the accuracy and the authority of the report are improved.
In conclusion, the electroencephalogram characteristic extraction method automatically calculates and extracts electroencephalogram characteristics of different brain areas and different frequencies through a system, wherein the electroencephalogram characteristics comprise average time ratio, average amplitude and amplitude difference of two sides, the electroencephalogram characteristics are analyzed and compared according to a standardized range, and finally a visual report is output, so that electroencephalogram background activity grading is realized. The system can be manually checked quickly, and can be modified and edited on line when necessary, the accuracy of result analysis is guaranteed while the automatic identification efficiency of the system is met, and the application flexibility of the system is increased, so that the possibility of efficiently extracting high-dimensional data information from the original two-dimensional electroencephalogram is provided, and a potential electroencephalogram analysis technical means is provided for non-electroencephalogram professional doctors and scientific researchers. Compared with the traditional manual identification processing mode, the method realizes the automation of the electroencephalogram background activity identification process, separates electroencephalographs from the complex and time-consuming background activity identification and description work, so that report result auditing and paroxysmal abnormal activity deep interpretation are invested, the working efficiency is improved, and the labor cost is saved.
The present invention is not limited to the above preferred embodiments, and any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A background activity automatic identification method based on resting state electroencephalogram is characterized by comprising the following steps:
step 1: acquiring a resting state electroencephalogram rsEEG signal of a patient by utilizing electroencephalogram signal acquisition equipment;
step 2: preprocessing resting state electroencephalogram rsEEG signal data of a patient in an eye closing state and an eye opening state for 3min respectively by adopting an ARTIST full-automatic denoising method;
and step 3: respectively calculating and extracting three electroencephalogram characteristics of average time ratio, average amplitude and bilateral amplitude difference of different brain areas of each frequency band;
and 4, step 4: analyzing and interpreting the extracted electroencephalogram characteristics, and sequentially and respectively judging in the characteristic conditions of severe background abnormality, moderate background abnormality, mild background abnormality, boundary electroencephalogram and normal electroencephalogram;
and 5: automatically generating a visual report according to the judgment result, presenting a conclusion of electroencephalogram background activity identification, and the quantified electroencephalogram characteristics and reference ranges of different brain areas of each frequency band;
step 6: and (5) carrying out manual quick verification according to the visual report generated in the step (5), and modifying and editing the report content with deviation by manual work if necessary.
2. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, characterized in that the specific steps of the step 2 are as follows:
step 2.1: respectively keeping the eyes of the patient in a closed state and an open state for a period of time, and respectively selecting 3min of resting state electroencephalogram rsEEG signal data from the collected signals;
step 2.2: removing direct current drift in resting state electroencephalogram rsEEG signal data;
step 2.2: removing eye movement interference in signal data selected by the eye closing state and the eye opening state respectively;
step 2.3: reducing the sampling rate to 250Hz, adjusting the band-pass filtering to 1-45Hz, and replacing a bad channel;
step 2.4: segmenting data, wherein the time length of each segment of data is 2 s;
step 2.5: removing the segmented bad channel and performing interpolation;
step 2.6: and (4) after analyzing the independent components, removing pseudo-error components and adopting an average reference value.
3. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, characterized in that the specific steps of step 3 are as follows:
step 3.1: dividing the whole brain into four areas, wherein the first area comprises a frontal area and an anterior temporal area, the second area comprises a central area and a medial temporal area, the third area comprises a parietal area, and the fourth area comprises an occipital area and a posterior temporal area;
step 3.2: according to a frequency calculation method, signals with different frequencies appearing in each channel of an electroencephalogram along with time are identified, the frequencies are divided into five frequency bands which are respectively alpha waves, beta waves, gamma waves, delta waves and theta waves, the identified waves with different frequencies are marked by different colors, and then the time length of the waves appearing in each frequency band is compared with the total time length of the signals, namely the average time ratio of each frequency band;
step 3.3: and calculating the peak-peak value of each identified brain wave, storing the peak-peak value as the amplitude of each wave, and obtaining the average peak-peak value of all brain waves in each frequency band as the average amplitude of the frequency band.
Step 3.4: dividing the difference value of the average amplitude values of a certain frequency band in the left hemisphere and the right hemisphere by the low value of the average amplitude values of the same frequency band in the left hemisphere and the right hemisphere by taking the brain center line as a boundary, reflecting the symmetry of the left hemisphere and the right hemisphere, and obtaining the bilateral amplitude difference;
step 3.5: and (4) extracting the average time ratio, the average amplitude and the two-side amplitude difference electroencephalogram characteristics of each region in each frequency band according to the steps 3.2-3.4.
4. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, wherein the diagnosis criteria of the normal electroencephalogram in the step 4 are as follows:
the average time of delta wave in each region accounts for 0-2%, and the average amplitude is 0-50 muV;
the average time proportion of theta wave in each zone is 0-15%, and the average amplitude is 0-50 muV;
the average time proportion of alpha waves in the first area and the second area is 40-100%, the average time proportion in the third area is 45-100%, the average time proportion in the fourth area is 50-100%, and the average amplitude of each area is 0-100 μ V;
fourthly, the average time proportion of the beta wave in each area is 0 to 40 percent, and the average amplitude is 0 to 20 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 0-30%, the bilateral amplitude difference in the third area is 0-40%, and the bilateral amplitude difference in the fourth area is 0-100%.
5. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, wherein the diagnostic criteria of the boundary electroencephalogram in the step 4 are as follows:
the restricted electroencephalogram is characterized by any one of the following abnormal manifestations:
the average time of delta wave in each region accounts for 2-5%, and the average amplitude is more than 50 muV;
the average time proportion of theta wave in each area is 15-30%, and the average amplitude is more than 50 mu V;
the average time proportion of alpha waves in the first area and the second area is 10-40%, the average time proportion in the third area is 10-45%, the average time proportion in the fourth area is 10-50%, and the average amplitude of each area is more than 100 mu V;
the average time proportion of the beta wave in each area is 40-100%, and the average amplitude is 20-50 μ V;
the amplitude difference of the two sides of each frequency band in the first area and the second area is 30-50%, the amplitude difference of the two sides in the third area is 40-60%, and the amplitude difference of the two sides in the fourth area is 100-150%.
6. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, wherein the diagnosis criteria of mild background abnormality in the step 4 are as follows:
the mild background abnormality is any one of the following abnormal manifestations:
the average time of delta waves in each region accounts for 5-10%;
the average time of theta wave in each zone accounts for 30-75%;
③ 5 to 10 percent of the average time of each area of the alpha wave;
the average amplitude of the beta wave in each area is more than 50 MuV;
the bilateral amplitude difference of each frequency band in the first area and the second area is 50-75%, the bilateral amplitude difference in the third area is 60-85%, and the bilateral amplitude difference in the fourth area is 150-200%.
7. The method for automatically identifying background activity based on resting state electroencephalogram according to claim 1, wherein the diagnosis criteria of moderate background abnormality in the step 4 are as follows:
the abnormal expression of any one of the following is moderate background abnormality:
the average time of delta waves in each region accounts for 10-95%;
the average time of theta wave in each zone accounts for 75-95%;
③ the average time of each area of the alpha wave is 0 to 5 percent;
and the bilateral amplitude difference of each frequency band in the first area and the second area is more than 75%, the bilateral amplitude difference in the third area is more than 85%, and the bilateral amplitude difference in the fourth area is more than 200%.
8. The method for automatically identifying background activity based on resting electroencephalogram according to claim 1, wherein the diagnosis standard of severe background abnormality in the step 4 is as follows:
the severe background abnormality is represented by any one of the following abnormalities:
the average time of delta waves in each region accounts for 95-100%;
the average time of theta wave in each zone is 95-100%.
9. An automatic background activity recognition system, comprising: interconnect's data management module, data storage module and data processing analysis module, the data management module includes: the data editing and modifying module and the information input module, the data processing and analyzing module comprises: preprocessing module, frequency identification and color identification module, brain district division module, feature extraction calculation module, comparison module and visual module, data management module, preprocessing module, frequency identification and color identification module and brain district division module are connected with feature extraction calculation module, feature extraction calculation module is connected with comparison module, comparison module is connected with visual module, visual module is connected with data storage module, data storage module is connected with data editing and modifying module.
10. The system of claim 9, wherein the data management module is a computer group, the data storage module is a disk array memory, and the data processing and analyzing module is a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110819422.1A CN113545791A (en) | 2021-07-20 | 2021-07-20 | Background activity automatic identification method and system based on resting electroencephalogram |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110819422.1A CN113545791A (en) | 2021-07-20 | 2021-07-20 | Background activity automatic identification method and system based on resting electroencephalogram |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113545791A true CN113545791A (en) | 2021-10-26 |
Family
ID=78103555
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110819422.1A Pending CN113545791A (en) | 2021-07-20 | 2021-07-20 | Background activity automatic identification method and system based on resting electroencephalogram |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113545791A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102715903A (en) * | 2012-07-09 | 2012-10-10 | 天津市人民医院 | Method for extracting electroencephalogram characteristic based on quantitative electroencephalogram |
| CN108143411A (en) * | 2017-12-13 | 2018-06-12 | 东南大学 | A kind of tranquillization state brain electricity analytical system towards Autism Diagnostic |
| US20190159712A1 (en) * | 2009-03-30 | 2019-05-30 | Donald H. Marks | Brain function decoding process and system |
| CN110074783A (en) * | 2019-05-17 | 2019-08-02 | 杭州师范大学 | The cerebral cortex excitability of transcranial magnetic stimulation induction signal and imaging and quantization method |
| CN112927793A (en) * | 2021-01-26 | 2021-06-08 | 天津大学 | Consciousness disturbance electroencephalogram data analysis system based on Web |
-
2021
- 2021-07-20 CN CN202110819422.1A patent/CN113545791A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20190159712A1 (en) * | 2009-03-30 | 2019-05-30 | Donald H. Marks | Brain function decoding process and system |
| CN102715903A (en) * | 2012-07-09 | 2012-10-10 | 天津市人民医院 | Method for extracting electroencephalogram characteristic based on quantitative electroencephalogram |
| CN108143411A (en) * | 2017-12-13 | 2018-06-12 | 东南大学 | A kind of tranquillization state brain electricity analytical system towards Autism Diagnostic |
| CN110074783A (en) * | 2019-05-17 | 2019-08-02 | 杭州师范大学 | The cerebral cortex excitability of transcranial magnetic stimulation induction signal and imaging and quantization method |
| CN112927793A (en) * | 2021-01-26 | 2021-06-08 | 天津大学 | Consciousness disturbance electroencephalogram data analysis system based on Web |
Non-Patent Citations (2)
| Title |
|---|
| 党鸽;石雪;蔡敏;苏晓琳;陈思言;曾思琳;兰小勇;付学军;邹良玉;郭毅;: "基于静息态脑电图的卒中后抑郁脑网络异常特征研究", 中国卒中杂志, no. 05, 20 May 2020 (2020-05-20) * |
| 方舟;姚柳叶;钱志余;: "EEG多参数可视化系统研究", 生命科学仪器, no. 06, 25 December 2019 (2019-12-25), pages 72 - 74 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chu et al. | Spatiotemporal EEG microstate analysis in drug-free patients with Parkinson's disease | |
| Gevins et al. | Automated analysis of the electrical activity of the human brain (EEG): A progress report | |
| CN111772585B (en) | Data extraction method for epileptic abnormal group activities in interval intracranial electroencephalogram signals | |
| CN111340142B (en) | Epilepsia magnetoencephalogram spike automatic detection method and tracing positioning system | |
| Dorfman et al. | AAEE minimonograph# 29: Automatic quantitative electromyography | |
| WO2007138579B1 (en) | Neuropsychological spatiotemporal pattern recognition | |
| CN113768519B (en) | Method for analyzing consciousness level of patient based on deep learning and resting state electroencephalogram data | |
| CN112220485A (en) | Epilepsia electroencephalogram/magnetoencephalogram abnormal wave automatic detection method and positioning system | |
| WO2022135449A1 (en) | Interictal epileptiform discharge activity detection apparatus and method for epileptic patient | |
| CN111067513B (en) | Sleep quality detection key brain area judgment method based on characteristic weight self-learning | |
| Quitadamo et al. | EPINETLAB: a software for seizure-onset zone identification from intracranial EEG signal in epilepsy | |
| CN112057087A (en) | Method and device for evaluating autonomic nerve function of high-risk schizophrenic population | |
| Zhu et al. | Altered eeg oscillatory brain networks during music-listening in major depression | |
| Bai et al. | Network analysis of magnetoencephalogram signals in schizophrenia patients when viewing emotional facial stimuli | |
| Yang et al. | Depression detection based on analysis of EEG signals in multi brain regions | |
| CN116616771B (en) | Multichannel simple mental state detection method, device and system | |
| CN113288174A (en) | Method for detecting cognitive function of schizophrenic patient | |
| CN113208629A (en) | Alzheimer disease screening method and system based on EEG signal | |
| CN113545791A (en) | Background activity automatic identification method and system based on resting electroencephalogram | |
| Kulkarni | Support vector machine based alzheimer’s disease diagnosis using synchrony features | |
| CN111281382A (en) | Feature extraction and classification method based on electroencephalogram signals | |
| CN116211306A (en) | Psychological health self-evaluation system based on eye movement and electrocardiosignal | |
| CN114881089A (en) | Depression electroencephalogram classification method based on double-branch fusion model | |
| Liang et al. | The classification of flash visual evoked potential based on deep learning | |
| Yu et al. | Epilepsy seeg data classification based on domain adversarial learning |
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 |