Fig 5 - uploaded by Olga Sushkova
Content may be subject to copyright.
A diagram of AU C values calculated for various frequency bands. In the upper left triangle of the diagram: the abscissa is the lower bound of the frequency band and the ordinate is the upper bound of the frequency band. In the lower right triangle of the diagram: the abscissa is the upper bound of the excluded frequency band and the ordinate is the lower bound of the excluded band. The frequency varied from 2 to 25 Hz with the 0.5 Hz step. The background EEG was analyzed, right hand tremor patients, the C3 electrode. (Color figure online)
Source publication
A method of analysis and visualization of electroencephalograms (EEG) based on wave trains is developed. In this paper, we use the “wave train” term to denote a signal localized in time, frequency, and space. The wave train is a typical pattern in a background EEG and detecting/analyzing such signals gives useful information about the brain activit...
Context in source publication
Context 1
... values MinF req, MaxF req, M inP ower, MaxP ower, MinDurat, M axDurat, M inBandwidth, and M axBandwidth can be implemented to investigate the multidimensional space, but we prefer an accurate consideration of different slices of the space using various 2D and 3D diagrams not to miss any interesting regularities in the space of the wave trains. Fig. 5 reveals three interesting frequency ranges that may be prospective for research. The first range is mu (a blue region, the 10.5 . . . 13.5 Hz frequency band approximately), the second is mu too (a red region, the 6 . . . 9.5 Hz frequency band approxi- mately), and the third range is beta (a dark blue region, 18 . . . 24 Hz frequency ...
Citations
... The idea of AUC diagrams was proposed by the authors for the analysis of wave train electrical activity of the brain in the framework of the problem of diagnosing neurodegenerative diseases [17][18][19][20][21]. In this paper, the method of analyzing wave train electrical activity is not used; however, the principles of constructing and reading AUC diagrams remain the same. ...
... Thus, examples of IAMs in the beta-I frequency range are not given for brevity. Fig. 12 demonstrates an example of IAM of the first component of PCA in the beta-II frequency range (17)(18)(19)(20)(21)(22)(23)(24). Notably, IAM in the beta-II range differs from IAM in the alpha range. ...
The article describes a new type of AUC diagrams intended for the analysis of interhemispheric asymmetry of amplitude-frequency characteristics of electroencephalograms (EEG) of patients with subarachnoid hemorrhage, as well as a new type of head maps named maps of interhemispheric asymmetry of EEG. AUC diagrams are a new statistical tool for identifying regularities in biomedical signals. The idea of AUC diagrams is to visually represent the dependence of the area under the ROC curve (AUC) when comparing data samples from the bounds of the range of given characteristic of this data, for example, frequency or amplitude, etc. The article demonstrates that this principle of data analysis allows us to identify some signs of postoperative complications that may occur in patients undergoing intensive care unit. It is known that the signs of such complications are changes in the amplitude and frequency of EEG oscillations in the neurophysiological frequency ranges delta, theta, alpha, and beta; however, amplitude changes can be caused by other reasons including the state of sleep and exposure to pharmacological drugs. Changes in the amplitude caused by postoperative complications can be revealed by analysis of the interhemispheric asymmetry of the patient's EEG. The developed type of AUC diagrams and interhemispheric EEG asymmetry maps help to automate such EEG analysis. The effectiveness of the developed statistical tools was demonstrated by the analysis of data in two patients with clinically confirmed delayed cerebral ischemia induced by subarachnoid hemorrhage.
... We have developed a visualization method that facilitates the comparison of datasets and the search for the intervals of the wave train parameters [20][21][22][23]. The visualization method includes the following steps: ...
... During the data acquisition, the patients sat in an armchair with arms outstretched forward. The experimental setting is described in more detail in [20,21,24,25]. In the example under consideration, the occipital region of the cerebral cortex is investigated (the O2 EEG channel). ...
... Other types of AUC diagrams are considered in [23]. The experience of data analysis demonstrated that the wave train electrical activity analysis method can be successfully applied to different kinds of biomedical signals, including EEG [20,21,[24][25][26], EMG [22,23,27,28], and accelerometer signals [22,29]. Further development of the method may include the usage of additional parameters of the wave trains as well as a statistical analysis of relationships between the wave trains with different attributes. ...
This paper proposes a new mathematically founded concept
called wave train electrical activity analysis to investigate both local and global features in
biomedical signals simultaneously. Specifically, its application is demonstrated in the investigation of epileptic seizures as well as the differential diagnosis of neurodegenerative diseases, Parkinson’s disease and essential tremor.
... We have developed a visualization method that facilitates the comparison of datasets and the search for the intervals of the wave train parameters [20][21][22][23]. The visualization method includes the following steps: 1. ...
... The intervals are tested using all wave trains detected in the wavelet spectrograms of given datasets. [20,21,24,25]. In the example under consideration, the occipital region of the cerebral cortex is investigated (the O2 EEG channel). ...
... Other types of AUC diagrams are considered in [23]. The experience of data analysis demonstrated that the wave train electrical activity analysis method can be successfully applied to different kinds of biomedical signals, including EEG [20,21,[24][25][26], EMG [22,23,27,28], and accelerometer signals [22,29]. Further development of the method may include the usage of additional parameters of the wave trains as well as a statistical analysis of relationships between the wave trains with different attributes. ...
Classical methods for signal analysis are limited to describe either the global features or the local features. This paper proposes a new mathematically founded concept called wave train electrical activity analysis to investigate both local and global features in biomedical signals simultaneously. First, mathematical means for the investigation of the properties of the wave trains observed in the biomedical signals, histograms of wave train parameters and AUC diagrams, are discussed. Second, several examples of the practical application of the method of the wave train electrical activity analysis are considered. Specifically, its application is demonstrated in the investigation of epileptic seizures as well as the differential diagnosis of neurodegenerative diseases, Parkinson’s disease and essential tremor.
... The method developed for analyzing the wave train electrical activity is a universal method for exploratory data analysis and can be applied to other types of biomedical signals [58][59][60][61][62][63][64][65][66][67][68][69][70]. In particular, we demonstrated that the statistical analysis of some characteristics of wave trains in EEG can identify features of the preclinical stage of PD [58][59][60][61]. ...
... The method developed for analyzing the wave train electrical activity is a universal method for exploratory data analysis and can be applied to other types of biomedical signals [58][59][60][61][62][63][64][65][66][67][68][69][70]. In particular, we demonstrated that the statistical analysis of some characteristics of wave trains in EEG can identify features of the preclinical stage of PD [58][59][60][61]. It was found that the number of wave trains in wavelet spectrograms in the beta frequency range in first-stage PD patients was significantly reduced in comparison with the control subjects [71][72][73]. ...
A statistical method for exploratory data analysis based on 2D and 3D area under curve (AUC) diagrams was developed. The method was designed to analyze electroencephalogram (EEG), electromyogram (EMG), and tremorogram data collected from patients with Parkinson’s disease. The idea of the method of wave train electrical activity analysis is that we consider the biomedical signal as a combination of the wave trains. The wave train is the increase in the power spectral density of the signal localized in time, frequency, and space. We detect the wave trains as the local maxima in the wavelet spectrograms. We do not consider wave trains as a special kind of signal. The wave train analysis method is different from standard signal analysis methods such as Fourier analysis and wavelet analysis in the following way. Existing methods for analyzing EEG, EMG, and tremor signals, such as wavelet analysis, focus on local time–frequency changes in the signal and therefore do not reveal the generalized properties of the signal. Other methods such as standard Fourier analysis ignore the local time–frequency changes in the characteristics of the signal and, consequently, lose a large amount of information that existed in the signal. The method of wave train electrical activity analysis resolves the contradiction between these two approaches because it addresses the generalized characteristics of the biomedical signal based on local time–frequency changes in the signal. We investigate the following wave train parameters: wave train central frequency, wave train maximal power spectral density, wave train duration in periods, and wave train bandwidth. We have developed special graphical diagrams, named AUC diagrams, to determine what wave trains are characteristic of neurodegenerative diseases. In this paper, we consider the following types of AUC diagrams: 2D and 3D diagrams. The technique of working with AUC diagrams is illustrated by examples of analysis of EMG in patients with Parkinson’s disease and healthy volunteers. It is demonstrated that new regularities useful for the high-accuracy diagnosis of Parkinson’s disease can be revealed using the method of analyzing the wave train electrical activity and AUC diagrams.
... There are different methods of signal processing, the main methods are spectral analysis [5] and phase synchronization research. Methods based on the use of wavelets are a variety of spectral analysis [7][8][9][10][11][12][13]. The advantage of wavelet analysis is the ability to take into account the time-frequency dynamics of signals. ...
... The use of ridges allows to select the leading frequency in the signal, as well as to find signal areas with a high signal-tonoise ratio. In [8][9][10][11][12][13], an approach is used based on the method of analysis of wave train electrical activity, a distinctive feature of which is that not the original signals are analyzed, but wave trains of the spectral power density on wavelet spectrograms. Unlike standard wavelet analysis, this method allows to reveal the properties of electrophysiological signals (EMG, tremor, and electroencephalograms) over long time intervals and at the same time, in comparison with Fourier analysis, take into account local time-frequency changes in the characteristics of non-stationary signals. ...
An investigation of surface electromyogram (EMG) of antagonist muscles in patients with Parkinson's disease (PD) and essential tremor (ET) was carried out. A comparison was made between two methods for calculating instantaneous phases of envelopes of EMG signals. The first method is based on calculating the ridges of wavelet spectrograms of envelopes of EMG signals. The second method is based on using the Hilbert transform. Statistically significant difference between the mean values of the phase difference of the EMG signal envelopes was found in the antagonist muscles in PD and ET patients.
... Usually, the frequency range below 4 Hz is not investigated in EMG, since it is considered that it is impossible to find statistically significant differences between groups of patients and healthy subjects in this range. In the EEG, the frequency ranges of theta (4-7 Hz), alpha (8)(9)(10)(11)(12), beta (13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24), and gamma [3,4] are mainly studied. It is rare to find papers where the delta (1-4 Hz) range is studied in EEG [5]. ...
... Earlier, we developed a method for analyzing the wave train electrical activity of the cerebral cortex, based on wavelet analysis and ROC analysis [22,23]. The idea of this method of analysis is in that an electroencephalogram (EEG) is considered as a set of wave trains [24]. In contrast to works on the detection of the electrical activity of one or two specific types, such as alpha spindles [25] and sleep spindles [26][27][28][29][30][31], we analyze any type of the wave train electrical activity in the cerebral cortex over a wide frequency range. ...
... The number of wave trains is compared with the healthy subject data using special AUC-diagrams and Mann-Whitney non-parametric statistical test. A detailed description of the AUC diagrams is given in [23,24,[34][35][36][37][38][39]. ...
An investigation of the 0.5-4 Hz little-studied frequency range electromyograms (EMG) was performed in patients with Parkinson's disease (PD) and essential tremor (ET). In this frequency range, new neurophysiological regularities were revealed that were not previously described in the literature. There are statistically significant differences between groups of patients with PD/ET and a control group of subjects. A new method for analyzing wave train electrical activity of the muscles based on the wavelet analysis and ROC analysis was used. This method enables to study the time-frequency features of EMG signals in patients with PD and ET. The idea of the method is to find local maxima (that correspond to the wave trains) in the wavelet spectrogram and to calculate various characteristics describing these maxima: the leading frequency, the duration in periods, the bandwidth, the number of wave trains per second. The degree of difference of the group of patients from the control group of subjects is analyzed in the space of these parameters. ROC analysis is used for this purpose. The functional dependence of AUC (the area under the ROC curve) on the values of the bounds of the ranges of the parameters under consideration is investigated. This method is aimed at studying changes in the time-frequency characteristics (the shape) of signals including changes that are not related to the power spectral density of the signal. The application of the method allowed revealing new statistical regularities in EMG signals, which previously were not detected using standard spectral methods based on the analysis of the power spectral density of signals.
... Earlier, the authors have developed a method for analyzing the wave train electrical activity of the cerebral cortex, based on wavelet analysis and ROC analysis [19,20]. The idea of this method of EEG analysis is in that the EEG signal is considered as a set of wave trains [21]. Unlike works devoted to the detection of wave train electrical activity of one or two specific types, such as alpha spindles [6] and sleep spindles [11,10,4,9,5,1], we analyze all kinds of wave train electrical activity in the brain cortex in a wide frequency range. ...
In recent years, spindle-shaped electrical activity became interesting for researchers looking for new methods of time-frequency analysis of electromyograms (EMG) and acceleration (ACC) signals. We call signals of this type as wave trains; a wave train (a wave packet) is an electrical signal that is localized in space, frequency, and time. Examples of wave trains in electroencephalograms (EEG) are alpha spindles, beta spindles, and sleep spindles. We analyze all kinds of wave train electrical activity of the muscles in a wide frequency range. We have developed a new method for analyzing wave train electrical activity of muscles based on wavelet analysis and ROC analysis that enables to study the time-frequency features of EMG and ACC in limbs’ tremor in patients with neurodegenerative diseases such as Parkinson’s disease (PD).
... Примерами всплескообразной электрической активности мозга являются α-, β и сонные веретена. Метод анализа всплескообразной электрической активности ЭЭГ был опубликован в ряде математических публикаций [13][14][15]. Метод анализа всплескообразной электрической активности коры мозга основан на вейвлет-анализе и применении ROC-кривых и AUC. В данной работе он был использован для изучения новых частотно-временных параметров ЭЭГ и исследования их изменений у пациентов с БП. ...
... Количество всплесков у здоровых и пациентов на ранней стадии БП анализируется с помощью используемых в математике ROC-кривых, также известных как кривые ошибок [14,15]. ROC-кривая (англ.: receiver operating characteristic, рабочая характеристика приёмника) -график, позволяющий оценить качество бинарной классификации; отображает соотношение долей объектов, классифицированных верно (TPR, англ.: true positive rate -количество истинно-положительных результатов), и ошибочно (FPR, англ.: false positive rate -количество ложноположительных результатов). ...
Aim:
To develop a mathematical method of analysis and visualization of EEG based on the ROC analysis of burst electrical activity in the cerebral cortex.
Material and methods:
Using a new method of analysis of EEG burst activity, the frequency parameters of brain electrical activity have been investigated in patients in the first stage of Parkinson's disease (PD) defined by the Hoehn and Yahr scale. Patients were right-handed, with disease onset in either the right or the left side. The burst term is used in neurophysiology for the description of wave activity in EEG signals. Bursts are reflected in the local peaks of wavelet spectrograms, some of the parameters of which have been analyzed. Electrical activity of the left and right central cortex areas was investigated. The results were compared with those obtained from healthy volunteers.
Results:
In PD patients, burst activity was changed in alpha- and beta bands. Compared to healthy volunteers, it was higher in alpha band 8-9 Hz and lower in upper alpha band 11-13 Hz and beta band 18-24 Hz. With regard to asymmetry of the brain in PD patients, there was the change in burst activity in both brain hemispheres. Diagrams of burst activity showed the difference between the patients with tremor onset in the left hand and tremor onset in the right hand.
Conclusion:
This suggests differences in brain electrical activity changes in patients with left-sided and right-sided disease onset. The initial results of the study demonstrate that the method of analysis and visualization based on the evaluation of certain parameters of EEG bursts is promising for the analysis of EEG features in PD patients.
... Earlier, we have developed a method for analyzing the wave train electrical activity of the cerebral cortex, based on wavelet analysis and ROC analysis [11]. The idea of this method of EEG analysis is in that we consider the EEG signal as a set of wave trains [12]. Unlike works devoted to the detection of electrical activity of one or two specific types, such as alpha spindles [13] and sleep spindles [1], [2], [14]- [17], we analyze any kinds of wave train electrical activity in the brain in a wide frequency range. ...
... The investigation of the specificity of wave train features of PD discovers considerable differences between the PD patients, ET patients, and healthy volunteers. We have observed differences in quantity of wave trains in PD patients and ET patients in the following frequency bands: 5-9.5 Hz (approximately the theta and alpha frequency ranges) and [11][12][13][14][15][16] Hz (approximately the upper alpha and lower beta frequency ranges). The detailed analysis of the results indicates that both PD and ET patients differs in quantity of wave trains from the healthy volunteers, but the difference between ET patients and the healthy volunteers is less. ...
