, , , , , , , 
As the main method of automatic computer analysis of EEG, spectral analysis based on Fourier transform is used, the representation of the native EEG picture as a set of a set of sinusoidal oscillations that differ in frequency and amplitude.
The main output parameters of the spectral analysis are:
- average amplitude;
- average and modal (most frequent) frequencies of EEG rhythms;
- spectral power of EEG rhythms (integral index corresponding to the area under the EEG curve and depending both on the amplitude and on the index of the corresponding rhythm).
Spectral analysis of EEG is usually performed on short (2-4 sec) fragments of recording (epochs of analysis). Averaging the EEG power spectra for several dozen individual epochs with the calculation of the statistical parameter (spectral density) gives an idea of the most characteristic picture of the EEG for a given patient.
By comparing the power spectra (or the spectral density, the EEG coherence index is obtained in different leads, which reflects the similarity of the biopotential oscillations in different regions of the cerebral cortex.) This index has a certain diagnostic value: for example, increased coherence in the α-frequency band (especially when desynchronization EEG) are detected with the active joint participation of the relevant departments of the cerebral cortex in the performed activity.In contrast, the increased coherence in the 5-rhythm band reflects with izhennoe functional state of the brain (e.g., superficial tumors).
Periodometric analysis (period-analysis, or amplitude-interval analysis) is used less often when the periods between the characteristic points of EEG waves (wave vertices or zero-line intersections) and the peak amplitude of the waves (peaks) are measured.
Period-analysis of EEG allows you to determine the average and extreme values of the amplitude of EEG waves, the average periods of waves and their dispersion, accurately (by the sum of all periods of waves of a given frequency range) to measure the EEG rhythm index.
Compared to Fourier analysis, the period-analysis of EEG has greater resistance to interference, since its results are much less dependent on the contribution of single high-amplitude artifacts (for example, interference from patient movements). However, it is used less often for spectral analysis, in particular, because standard criteria for detection thresholds of EEG wave peaks have not been developed.
Other non-linear methods of EEG analysis
Other nonlinear EEG methods are described, based, for example, on the calculation of the probability of the appearance of successive EEG waves belonging to different frequency ranges, or on the determination of temporal relationships between certain characteristic fragments of EEG | EEG patterns (eg, α-rhythm spindles) | in different leads. Although the experimental results showed the informativeness of the results of such types of EEG analysis in regard to the diagnosis of certain functional states of the brain, in practice, these methods are practically not used.
Quantitative electroencephalography makes it possible to determine the localization of foci of pathological activity in epilepsy and various neurological and vascular disorders more accurately than in the case of visual analysis of the EEG, to detect violations of amplitude-frequency characteristics and spatial organization of the EEG, in a number of mental disorders, to quantify the effect of therapy including psychopharmacotherapy ) on the functional state of the brain, as well as to perform automatic diagnosis of certain disorders and / or functional states of a healthy person by comparing individual EEG with databases of normative EEG data (age norm, different types of pathology, etc.). All these advantages allow to significantly reduce the time of preparation of the conclusion based on the results of the EEG survey, increase the probability of detecting abnormalities of the EEG from normal.
The results of quantitative EEG analysis can be given both in digital form (in the form of tables for subsequent statistical analysis) and also as a visual color map, which can be conveniently compared with CT, magnetic resonance imaging (MRI) and positron emission tomography PET), as well as with estimates of local cerebral blood flow and neuropsychological testing data. Thus, it is possible to directly compare the structural and functional disorders of brain activity.
An important step in the development of quantitative EEG was the creation of software for determining the intracerebral location of equivalent dipolar sources of the most high-amplitude EEG components (eg, epileptiform activity). The latest achievement in this area is the development of programs combining MRI and EEG maps of the patient's brain, taking into account the individual shape of the skull and the topography of the brain structures.
When interpreting the results of visual analysis or EEG mapping, it is necessary to take into account age-related (both evolutionary and involutional) changes in the amplitude-frequency parameters and spatial organization of the EEG, as well as changes in the EEG against the background of medication that naturally arise in patients in connection with treatment. For this reason, the EEG record is usually performed before or after temporary discontinuation of treatment.
Electrophysiological study of sleep, or polysomnography - one of the areas of quantitative EEG.
The aim of the method is to objectively assess the duration and quality of night sleep, to identify sleep structure disorders [in particular, the duration and latent period of different phases of sleep, especially the sleep phase with fast eye movements], cardiovascular (cardiac and conduction disorders) and respiratory apnea) disorders during sleep.
Methodology of research
Physiological parameters of sleep (night or daytime):
- EEG in one or two leads (most often C3 or C4);
- data of the electrooculogram;
- data of electromyogram;
- frequency and depth of breathing;
- general motor activity of the patient.
All these indicators are necessary to identify the stages of sleep according to generally accepted standard criteria. Slow-wave sleep is determined by the presence of EEG sleepy spindles and σ-activity, and the phase of sleep with rapid eye movements - desynchronization of the EEG, the appearance of rapid eye movements and a deep decrease in muscle tone.
In addition, an electrocardiogram (ECG) is often recorded. HELL. Skin temperature and oxygenation of the blood (using an ear photo-oximeter). All these indicators allow you to assess vegetative disorders during sleep.
Interpretation of results
Reducing the latency of the sleep phase with rapid eye movements (less than 70 min) and early (at 4-5 h) morning awakening - established biological signs of depressive and manic states. In this regard, polysomyography makes it possible to differentiate depression and depressive pseudodementia in elderly patients. In addition, this method objectively identifies insomnia, narcolepsy, somnambulism, as well as nightmares, panic attacks, apneas and epileptic seizures that occur during sleep.