CN109222906B - Method for constructing pain state prediction model based on brain electrical signals - Google Patents
Method for constructing pain state prediction model based on brain electrical signals Download PDFInfo
- Publication number
- CN109222906B CN109222906B CN201811066031.1A CN201811066031A CN109222906B CN 109222906 B CN109222906 B CN 109222906B CN 201811066031 A CN201811066031 A CN 201811066031A CN 109222906 B CN109222906 B CN 109222906B
- Authority
- CN
- China
- Prior art keywords
- pain
- domain
- frequency
- brain
- prediction model
- 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.)
- Active
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/48—Other medical applications
- A61B5/4824—Touch or pain perception evaluation
-
- 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
-
- 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]
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
-
- 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
- A61B5/7253—Details of waveform analysis characterised by using transforms
- A61B5/7257—Details of waveform analysis characterised by using transforms using Fourier transforms
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Hospice & Palliative Care (AREA)
- Pain & Pain Management (AREA)
- Psychology (AREA)
- Power Engineering (AREA)
- Mathematical Physics (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a method for constructing a pain state prediction model based on brain electric signals. The method comprises the following steps: 1) preprocessing brain electric signal data; 2) extracting the characteristics of a time domain, a frequency domain and a wavelet domain; 3) screening the characteristics according to the contribution degree to the pain state; 4) constructing a prediction equation by using the single characteristics; 5) all the characteristics are integrated to carry out a multiple regression model to form a final pain state prediction model. The method has the advantages that the brain features are depicted from multiple dimensions and used for predicting the future state of a pain patient, so that the understanding of a pain mechanism and a treatment mechanism can be further deepened, and a foundation is laid for intelligent nerve regulation.
Description
Technical Field
The invention relates to the technical field of physiological signal processing and modeling, in particular to a method for constructing a pain state prediction model based on brain electric signals.
Background
Neuropathic pain is pain caused by injury or disease to the somatosensory nervous system, and this disease is often drug refractory. Neuropathic pain affects the physical and psychological health of 6% -8% of people worldwide for a long time, severely compromising their quality of life. In clinical diagnosis and treatment of the disease, doctors can usually only judge the disease by self symptom description of the patient, but the patient cannot objectively and accurately describe the pain feeling by speaking correctly. If the location of the pain and the type of pain cannot be accurately specified, the true extent of the pain may be exaggerated or concealed. The accurate evaluation of the state and the change degree of the patient has important research value for clinical diagnosis, treatment and rehabilitation guidance, and can provide theoretical and technical basis for the research of the physiological mechanism of the organism, the development of treatment theory and the research and development of new nerve regulation and control technology.
Subjective assessment methods commonly used clinically include pain self-assessment and other pain questionnaires, facial expressions, and the like, and objective detection methods include measurement of some physiological indicators such as electroencephalogram, electromyogram, electrocardiosignal, blood pressure, nuclear magnetic resonance, and the like. These subjective or objective assessment methods provide auxiliary references for diagnosis and treatment of clinical diseases, but all have certain disadvantages, such as subjective assessment methods, which have strong subjective consciousness, and require experienced physicians or patients to provide subjective feelings during assessment; although the objective evaluation method provides accurate physiological state data, the objective evaluation method generally only can provide a simple analysis result of the state of the patient at a certain stage, cannot provide a specific and accurate state or degree magnitude, is mostly single index evaluation, and is not comprehensive and objective.
Experiments prove that the human brain electrical signals can reflect physiological and pathological functions, and a plurality of indexes are closely related to neuropathic pain, so that the human brain electrical signals can be hopefully used for evaluating the treatment effect of the pain. However, because the activity of the brain electrical signal has the characteristic of complexity, for example, a plurality of nerve activities in the brain electrical signal participate in pain sensing and treatment at the same time, and a more complex relation network is formed between the nerve activities, the simultaneous measurement of the multidimensional nerve behaviors of the brain electrical signal can more comprehensively represent the brain functional state.
Disclosure of Invention
In view of the above technical problems, the present invention provides a method for constructing a pain state prediction model based on electrical brain signals. The invention comprehensively characterizes and quantifies the electrical activity of the brain from a multi-dimensional angle, combines subjective evaluation with objective detection means, fuses a plurality of detected biomarkers and establishes a quantitative prediction model of the state or the change degree of a patient, and the model can be used for realizing or accurately judging the pain state.
The brain electrical signals may include scalp electroencephalography (EEG), cortical electroencephalography (ECoG), Local Field Potentials (LFP), and the like.
The multidimensional neural behavior of brain electrical activity is divided into three dimensions: time domain, frequency domain and wavelet domain, and selecting proper analysis method to quantize the signal activity characteristics in each dimension. Multidimensional neural behavior of electrical brain signals may exist simultaneously, and the inclusion of many features in these dimensions may be correlated with the state of the patient. For example, but not limited to, changes in the amplitude of the signal activity in time domain may be correlated to a patient's specific pain level, the amplitude ratio between specific bands belonging to the frequency domain may be correlated to a specific disease, and the strength of cross-frequency coupling between bands belonging to the wavelet domain may indicate a specific brain function state. In one example, simultaneous measurement of multidimensional neural behavior of brain electrical signals to distinguish different disease type states demonstrates application of the method in brain electrical signal research, and shows feasibility of the brain electrical signal multidimensional neural behavior quantification method in brain function research. The technical scheme of the invention is specifically introduced as follows.
A method for constructing a pain state prediction model based on brain electrical signals comprises the following specific steps:
1) preprocessing the brain electric signals to remove signals with poor quality and noise in the signals;
2) extracting features from three dimensions of a time domain, a frequency domain and a wavelet domain to represent the electrical brain activity;
3) screening the electrical brain activity characteristics in the time domain and wavelet domain according to the correlation with the pain state; obtaining key components representing each feature group by using a Principal Component Analysis (PCA) method on a frequency domain according to the contribution rate so as to screen the brain electrical activity features; 4) using the characteristics screened from three dimensions of a time domain, a frequency domain and a wavelet domain as independent variables, using the pain relief degree as a dependent variable, and respectively establishing a state prediction model through regression analysis;
5) and (3) taking the state prediction results in different dimensions as independent variables, taking the clinical subjective evaluation result of the patient as a dependent variable, and establishing an integrated pain state prediction model by utilizing multivariate regression analysis.
Preferably, in the step 1), the preprocessing includes removing 50Hz power frequency interference and baseline drift, and the recording should be performed again when the signal quality is poor.
Preferably, in step 1), the preprocessing further includes performing normalization processing on the noise-removed signal to remove inaccuracy of model evaluation due to individual difference.
Preferably, in step 2), the features extracted from the time domain include an average value, a standard deviation and an information entropy of the signal amplitude; before extracting the features, the signal is normalized by dividing the signal value of each sampling point by the maximum value of the amplitude.
Preferably, in step 2), the frequency domain is characterized by power values of power spectral densities integrated in different frequency bands after fourier transform and ratios of powers between different frequency bands; before extracting the characteristics, the signal is normalized by dividing the power spectral density value at each frequency point by the integral of the power spectral density of the 2-90Hz frequency band.
Preferably, in step 2), the wavelet domain is characterized by the percentage of the existence time of the synchronization states of delta, theta, alpha, low-beta, high-beta, low-gamma and high-gamma frequency bands and the percentage of the existence time of 4 states 00,01,10 and 11 consisting of the binarization codes of the 7 frequency band synchronization levels to the total time.
Preferably, in step 3), features with significance less than 0.05 or 0.01 to the pain state are selected in the time domain and the wavelet domain; the features are selected in the frequency domain according to the 1-3 principal components with the largest contribution rate.
Preferably, in step 5), the current data is included in the pain state prediction model for modifying the model parameters each time the prediction is completed.
Compared with the prior art, the invention has the beneficial effects that:
the invention comprehensively characterizes and quantifies the brain electrical activity from a multi-dimensional angle, combines subjective evaluation and objective detection means, separates a plurality of biomarkers from single brain electrical activity and fuses the biomarkers, realizes accurate judgment of the state, and establishes a quantitative prediction model of the state or the change degree of a patient. The model of the invention can effectively predict the pain state.
Drawings
Fig. 1 is an exemplary flow diagram for constructing a model for predicting pain state (pain relief level of a pain patient) based on brain electrical signals.
Fig. 2 is an exemplary graph of the results of a calculation of the ratio of power at frequency points and power between frequency bands of a recorded electrical signal (i.e., LFP) in the brain of a neuropathic pain patient using power spectral analysis.
FIG. 3 is a graph of the results of correlation coefficients from a correlation analysis of frequency domain characteristics with the degree of pain relief of a patient and illustrates the identified characteristics in relation to the degree of pain in a patient.
FIG. 4 is an exemplary graph of predictive model performance comparison and validation results obtained in pain patient data.
Detailed Description
The present invention is described in further detail below with reference to the attached drawings so that those skilled in the art can implement the invention by referring to the description text.
Example 1
A method for constructing a pain state prediction model based on brain electrical signals, a flow chart of which is shown in fig. 1, comprising:
1) electrical brain signals are recorded for a period of time by scalp electrical brain signals (EEG) recorded by scalp electrodes, by cortical electroencephalogram (ECoG) signals placed under the skull in contact with the cerebral cortex or by Field Potential Signals (LFPs) implanted in deep nuclei associated with pain. The recording time is not less than 180 s. There is also a need to document the results of the assessment of the state of the painful patient by clinical staff. A common assessment scale is the pain analog visual scale (VAS).
2) The recorded electrical brain activity is preprocessed by first truncating a piece of more stable 50s data without singular values from the raw data. Then, a Chebyshev II type filter is used for carrying out 2-90Hz passband filtering, and the filter is used for removing 50Hz power frequency interference. Finally, the signals are normalized, and different dimensional features are extracted differently, for details, see the following.
3) And extracting features from three dimensions of a time domain, a frequency domain and a wavelet domain to characterize the brain activity.
Wherein the time domain extracted features include the mean, standard deviation and information entropy of the signal amplitude. The signal is normalized before feature extraction. The specific method is to divide the signal value of each sampling point by the maximum value of the amplitude.
Wherein the frequency domain is characterized by power values integrated over different frequency bands for the power spectral density after fourier transformation and by the ratio of the power between the different frequency bands. The integral of the power spectral density over a certain frequency interval characterizes the activity level of the signal in that frequency band, and the ratio of the activity levels between different frequency bands also characterizes the brain activity. In order to eliminate individual difference among patient signals, the power spectrum of each patient is normalized by dividing the power spectrum value at each frequency point by the integral of the power spectrum density of the 2-90Hz frequency band; the ratio of power between different frequency bands refers to the ratio between amplitude values of signals of different frequency bands, and in order to find a proper frequency combination, in the embodiment, a method of traversing the frequency combination is adopted, and considering that a commonly adopted rhythm frequency band is generally 4Hz or a multiple thereof, the analysis adopts a step size of 0.5Hz of the 4Hz frequency band to traverse the frequency combination.
The wavelet domain features adopt the method designed in the applicant patent 201610487800.X to perform synchronization state discrimination on signals of 7 frequency bands in total. These bands are delta (3-6Hz), theta (6-9Hz), alpha (9-12Hz), low-beta (12-24Hz), high-beta (24-36Hz), low-gamma (36-60Hz), and high-gamma (60-90Hz), respectively. And combining the 7 frequency band signals pairwise to obtain 20 combined states, wherein each combination has four states which are 00,01,10 and 11 respectively. The percentage of the single frequency over-synchronization state to the total time and the percentage of the four states of the combined state to the total time are respectively calculated.
4) All three-dimensional features were correlated with the results of clinical assessments, which in the example were associated with a 12-month post-operative pain reduction. Brain electrical activity features related to patient state in each dimension were extracted based on significance statistical analysis p < 0.05. The results of the correlation analysis of each feature of the time domain and the wavelet domain with the postoperative pain score and the postoperative remission level are shown in tables 1-3. In the time domain only the mean of the amplitude values correlated with the post-operative 12-month pain level, while the entropy of the amplitude values correlated with the post-operative 12-month pain relief level (table 1). The synchronization time of a single rhythm in the wavelet domain of highbeta alone is significantly correlated with the degree of post-operative 12-month pain relief. There are multiple combined states consisting of two rhythms that correlate with a degree of pain relief for the 12-month post-operative period.
Table 1 shows the results of a correlation analysis of time domain characteristics with the degree of pain and the degree of pain relief at 12 months after surgery in patients with neuropathic pain.
Table 2 shows the correlation analysis results of the percentage of the existing time of the 7-frequency band oversynchronization state of the wavelet domain characteristics and the postoperative 12-month pain degree and pain alleviation degree of the neuropathic pain patients.
Table 3 shows the partial results of significant correlation between the percentage of time that the synchronization status of wavelet domain features in a two-by-two combination of 7 frequency bands is present and the degree of pain relief of neuropathic pain patients at 12 months after surgery.
TABLE 1
TABLE 2
TABLE 3
5) Because the frequency domain has more extraction features, PCA is used for analyzing the feature vectors by respectively comparing the power of the frequency band with the power between frequencies, and proper principal components are selected as key components for representing each feature group according to the contribution rate. Fig. 2 is an exemplary graph of the results of a calculation of the ratio of power at frequency points and power between frequency bands of a recorded electrical signal (i.e., LFP) in the brain of a neuropathic pain patient using power spectral analysis.
FIG. 3 is a graph of the results of correlation coefficients from a correlation analysis of frequency domain characteristics with the degree of pain relief of a patient and illustrates the identified characteristics in relation to the degree of pain in a patient. The power spectral characteristics of the frequencies in the left graph of fig. 3, where the thick solid lines are located, are significantly correlated with the post-operative 12-month pain relief (P <0.01), while the ratio characteristics of the power spectral density between the frequency bands, boxed in the right graph of fig. 3, are significantly correlated with the post-operative 12-month pain relief (P < 0.01). The method using principal component analysis in modeling selects only the first principal component in fig. 2 and 3.
6) And (3) respectively constructing a prediction model by regression analysis by taking the characteristics selected by the three dimensions as independent variables and taking the postoperative pain degree or the postoperative pain relief degree as dependent variables. And (3) taking the result of the prediction model as an independent variable, taking the postoperative pain degree or pain relief degree as a dependent variable, and establishing a pain state prediction model integrating three dimensional characteristics through multivariate linear regression analysis. Fig. 4 shows specific prediction results. The results show that the prediction effect of using the characteristics of the three domains of the integration time domain, the frequency domain and the wavelet domain is better than that of using the characteristics of only a single dimension. The integrated prediction on the postoperative 12-month pain degree can reach 75%, and the integrated prediction on the postoperative 12-month pain relieving degree has the accuracy of 83%.
Claims (6)
1. A method for constructing a pain state prediction model based on brain electric signals is characterized by comprising the following specific steps:
1) preprocessing the brain electric signals to remove signals with poor quality and noise in the signals;
2) extracting features from three dimensions of a time domain, a frequency domain and a wavelet domain to represent the electrical brain activity;
3) screening the electrical brain activity characteristics in the time domain and wavelet domain according to the correlation with the pain state; in the frequency domain
Obtaining key components representing each characteristic group by using a Principal Component Analysis (PCA) method according to the contribution rate so as to screen the brain electrical activity characteristics;
4) using the characteristics screened from three dimensions of time domain, frequency domain and wavelet domain as independent variables, and using the pain relief degree as the independent variable
Respectively establishing state prediction models for the dependent variables through regression analysis;
5) the state prediction results in different dimensions are used as independent variables, and the clinical subjective evaluation results of patients are used as dependent variables, so that the method is favorable for
Establishing an integrated pain state prediction model by using multivariate regression analysis; wherein:
in the step 2), the characteristics extracted in the time domain comprise the average value, the standard deviation and the information entropy of the signal amplitude; before extracting the characteristics, the signal is standardized, and the specific method is to divide the signal value of each sampling point by the maximum value of the amplitude; the wavelet domain extraction is characterized in that the percentage of the existing time of the synchronization states of delta, theta, alpha, low-beta, high-beta, low-gamma and high-gamma frequency bands and the percentage of the appearance time of 4 states 00,01,10 and 11 formed by the binarization coding of the synchronization level of each frequency band in 21 combined states obtained by pairwise combination of the 7 frequency bands account for the total time.
2. The method of claim 1, wherein in step 1), the preprocessing comprises the steps of removing 50Hz power frequency interference and baseline wander.
3. The method of claim 1, wherein in step 1), the preprocessing further comprises the step of normalizing the noise-removed signal.
4. The method according to claim 1, wherein in step 2), the frequency domain is characterized by power values integrated by power spectral density in different frequency bands after Fourier transform and ratios of power between different frequency bands; before extracting the characteristics, the signal is normalized by dividing the power spectral density value at each frequency point by the integral of the power spectral density of the 2-90Hz frequency band.
5. The method according to claim 1, characterized in that in step 3), features with a pain state significance of less than 0.05 or 0.01 are selected in the time domain and wavelet domain; in the frequency domain, the features are selected according to the 1-3 principal components with the largest contribution rate.
6. The method of claim 1, wherein in step 5), the current data is included for modifying the model parameters each time the pain state prediction model completes a prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811066031.1A CN109222906B (en) | 2018-09-13 | 2018-09-13 | Method for constructing pain state prediction model based on brain electrical signals |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811066031.1A CN109222906B (en) | 2018-09-13 | 2018-09-13 | Method for constructing pain state prediction model based on brain electrical signals |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109222906A CN109222906A (en) | 2019-01-18 |
| CN109222906B true CN109222906B (en) | 2021-06-04 |
Family
ID=65057975
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811066031.1A Active CN109222906B (en) | 2018-09-13 | 2018-09-13 | Method for constructing pain state prediction model based on brain electrical signals |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109222906B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111951958A (en) * | 2020-08-21 | 2020-11-17 | 深圳大学 | Pain data evaluation method based on self-coding and related components |
| CN112957014B (en) * | 2021-02-07 | 2023-06-20 | 广州大学 | Pain detection and positioning method and system based on brain waves and neural network |
| CN113197585B (en) * | 2021-04-01 | 2022-02-18 | 燕山大学 | Neuromuscular information interaction model construction and parameter identification optimization method |
| CN113436728B (en) * | 2021-07-05 | 2022-10-28 | 复旦大学附属儿科医院 | Method and equipment for automatically analyzing electroencephalogram of clinical video of neonate |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103405225A (en) * | 2013-08-28 | 2013-11-27 | 深圳先进技术研究院 | Method, apparatus and device for obtaining pain feeling evaluation indexes |
| WO2014018661A1 (en) * | 2012-07-24 | 2014-01-30 | Cerephex Corporation | Method and apparatus for diagnosing and assessing central pain |
| CN106037725A (en) * | 2016-06-28 | 2016-10-26 | 中国科学院苏州生物医学工程技术研究所 | Bio-electrical brain signal analysis method based on wavelet packet transformation |
| CN107184204A (en) * | 2017-04-28 | 2017-09-22 | 北京易飞华通科技开发有限公司 | The extraction of nociceptive component and expression way in a kind of brain wave |
-
2018
- 2018-09-13 CN CN201811066031.1A patent/CN109222906B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014018661A1 (en) * | 2012-07-24 | 2014-01-30 | Cerephex Corporation | Method and apparatus for diagnosing and assessing central pain |
| CN103405225A (en) * | 2013-08-28 | 2013-11-27 | 深圳先进技术研究院 | Method, apparatus and device for obtaining pain feeling evaluation indexes |
| CN106037725A (en) * | 2016-06-28 | 2016-10-26 | 中国科学院苏州生物医学工程技术研究所 | Bio-electrical brain signal analysis method based on wavelet packet transformation |
| CN107184204A (en) * | 2017-04-28 | 2017-09-22 | 北京易飞华通科技开发有限公司 | The extraction of nociceptive component and expression way in a kind of brain wave |
Non-Patent Citations (1)
| Title |
|---|
| "神经病理性疼痛脑深部场电位分析及临床应用";黄永志 等;《万方数据库》;20161231;第14-58页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109222906A (en) | 2019-01-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Khosla et al. | A comparative analysis of signal processing and classification methods for different applications based on EEG signals | |
| CN109222906B (en) | Method for constructing pain state prediction model based on brain electrical signals | |
| Hamad et al. | Feature extraction of epilepsy EEG using discrete wavelet transform | |
| EP2498676B1 (en) | Brain activity as a marker of disease | |
| Henderson et al. | Development and assessment of methods for detecting dementia using the human electroencephalogram | |
| Boubchir et al. | A review of feature extraction for EEG epileptic seizure detection and classification | |
| US20150245800A1 (en) | Method for Detection Of An Abnormal Sleep Pattern In A Person | |
| Gruszczyńska et al. | Epilepsy identification based on EEG signal using RQA method | |
| CN111466876B (en) | Alzheimer's disease auxiliary diagnosis system based on fNIRS and graph neural network | |
| Yadav et al. | Morphology-based automatic seizure detector for intracerebral EEG recordings | |
| Sriraam et al. | Multichannel EEG based inter-ictal seizures detection using Teager energy with backpropagation neural network classifier | |
| Islam et al. | Probability mapping based artifact detection and wavelet denoising based artifact removal from scalp EEG for BCI applications | |
| Fathima et al. | Wavelet based features for classification of normal, ictal and interictal EEG signals | |
| Bibina et al. | Time-frequency methods for diagnosing alzheimer's disease using eeg: A technical review | |
| Radhakrishnan et al. | Investigating EEG Signals of Autistic Individuals Using Detrended Fluctuation Analysis. | |
| Wei et al. | Automatic recognition of chewing noises in epileptic EEG based on period segmentation | |
| Peya et al. | ASD Detection using Higuchi’s Fractal Dimension from EEG | |
| Saidatul et al. | The assessment of developed mental stress elicitation protocol based on heart rate and EEG signals | |
| Gomathi et al. | Smart System to Recognize EEG Signal for Finding Brain Diseases Using K Means Clustering | |
| Bhat | Assessing Performance of Detectors of High Frequency Oscillations in EEG Signals | |
| URAL et al. | AUTOMATED PSYCHIATRIC DATA ANALYSIS from SINGLE CHANNEL EEG with SIGNAL PROCESSING and ARTIFICIAL INTELLIGENCE METHODS | |
| Gruenwald et al. | Unsupervised quantification of high-gamma activity in electrocorticographic signals | |
| Abásolo et al. | Fractal dimension of the EEG in alzheimer's disease | |
| Lacan et al. | P35-T An automated EEG grading system for the early neonatal brain injury severity prediction post-perinatal asphyxia | |
| Babu et al. | Ocular and muscle artifacts suppression from EEG through MEMD and statistics |
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 |