JPS5927567B2 - EEG waveform correlation display device - Google Patents

Description

Detailed Description of the Invention The present invention provides an electroencephalogram waveform correlation display that quantifies and displays the degree of similarity of electroencephalogram waveforms derived from a plurality of parts of the head in order to know the relationship between each part of the brain. It is related to the device.

Generally, brain waves are recorded as potential changes based on the earlobe, etc., and show wave-like changes from about 0.1 Hz to 100 Hz.

Currently, for clinical purposes, brain waves derived from multiple sites are recorded on paper using Bengal bars, etc., but brain waves change in accordance with the state of brain activity, and their frequency and waveform can be determined from alpha waves, beta waves, etc. It is classified into waves, θ waves, etc.

By visually observing (inspecting) records on paper, doctors recognize changes in the frequency of waves and the relationships between different parts of the brain, and obtain information for diagnosis.

In this case, the low-frequency alpha waves (8-13Hz) and theta waves (
4 to 8 Hz) can be classified and recognized relatively easily, but high-frequency beta waves (13 Hz or higher) that appear when the brain is actively active, such as when thinking, have complex waveforms that require skilled doctors to However, it is difficult to classify the differences in detail.

In addition, various types of automatic brain wave analysis devices have been devised in recent years, and these can be classified into temporal analysis of brain waves and spatial analysis for understanding the distribution of waves in the head.

The main types of temporal analysis are frequency analysis such as auto- and cross-correlation using filters and Fourier analysis, and power spectrum. Spatial analysis is also known as topography or toposcope, and is used to identify specific areas in brain waves. It extracts waves and displays them in correspondence with each part of the brain.

Therefore, with any analysis method, it is possible to obtain data useful for diagnosis to some extent in the relatively low frequency region, but it is extremely difficult to obtain meaningful data in the high frequency β wave region. Have difficulty.

However, in reality, not only when resting or falling asleep,
It is necessary to extract valid data from beta waves when the brain is active, that is, when the brain waves exhibit beta waves.

The present inventors have discovered that the similarity of electroencephalogram waveforms can be effectively used to analyze brain functions.

The cerebrum is divided into two hemispheres, the left and right, each with different information processing functions.The left hemisphere is primarily responsible for logical functions such as language and calculations, while the right hemisphere is responsible for music recognition and spatial cognition. It is said that

Therefore, we derived brain waves from five parts of the head when humans are processing logical information and when processing spatial information, and analyzed the brain waves using waveform similarity analysis. , functional differences were detected using electroencephalograms, and the analysis method was found to be effective in analyzing human psychological information processing.

By deriving brain waves from multiple points in more detail, it is possible to analyze the information processing functions of various parts of the brain in more detail than by left-right differences, and these analysis results can also be used for medical and psychological diagnosis of the brain. I can do it.

From the above-mentioned viewpoint, the present invention focuses on the fact that the mutual activity of each part of the brain can be known by comparing the electroencephalogram waveforms derived from two parts of the brain. The present invention aims to provide a display device that quantifies and displays the degree of similarity between.

Embodiments of the present invention will be described below with reference to the drawings. As shown in FIG. , an amplifier 2 that amplifies the derived brain waves, an arithmetic unit 3 that quantifies the similarity of the brain wave waveforms of a certain time width derived at each of two adjacent derived regions, and the brain waves of those regions. It consists of a display section 4 that displays information corresponding to each part.

The calculation unit 3 calculates the similarity of brain waves between one derived region and each derived region located around it. This method extracts brain waves and quantifies the similarity of the brain wave waveforms themselves.

Various methods can be considered to quantify the above, but
Some of them are illustrated below.

(1) As shown in the following equation, two brain waves X(t), Y(t
) at every instant and integrate it every unit time T.

The above calculation can be performed using either analog or digital methods, but to specifically explain the analog method, brain waves X(t) and Y(t) are input to a multiplier, and the product of the two is calculated.

Then set that value to a specific time constant (e.g. o, about i seconds)
input into an integrator with , and calculate the instantaneous average value.

The value is then used as the output of the calculation section.

(2) A-transform two brain waves at a constant sampling time to create two number sequences for each unit time, and use these number sequences as data to find a correlation coefficient between the two.

Specifically, calculations are performed digitally.

For example, the brain waves X(t) and Y(t) are each subjected to A-D conversion every 4 m5ec to create two number sequences for each unit time.

If the unit time is 1 second, each number sequence will have 250 points.

250 points from X(t) and 2 from Y(t)
A correlation coefficient is determined using a 50-point numerical sequence as data.

The value r of the correlation coefficient is determined by the following formula according to Pearson's method.

Here, X・Y・ is the value of the sampled electroencephalogram, and Y
is the average value of 250 points.

Then, this value of r becomes the output of the calculation section.

(3) Figure 2A shows two electroencephalograms X(t) and Y(t) of a certain time width sampled at a certain sampling time.

As shown in B, at a certain point in time, the polarity is classified into four categories: either Tanachi or negative (@), and either increase (↑) or decrease (↓) compared to the previous point. Then, the two brain waves X(t) and Y(t) are paired to create an association frequency distribution table as shown in C in the figure according to the above-mentioned classification.

Chi-square or the amount of transmitted information is known as a statistical method for calculating the similarity between two brain waves from such a table, and these methods can be used to calculate χ2 or T(x,
Find y).

The method for determining the degree of similarity based on the amount of transmitted information is as follows.

From the association frequency distribution table, HCXI, H(Y), H(X, Y
) is determined by the following formula.

Hoo−1ΣP (Xi) log 2 P (Xi
)H(Y)=-ΣP(Yi) log 2 P(Y
i ) H(X, Y) −1ΣΣP(Xi, Yi)log
The amount of information transmitted between 2P(Xi, Yi)
Y).

This T(X, Y) becomes the output of the calculation section.

The calculations associated with quantification in each of the above methods can be easily performed using a well-known electronic device such as a computer.

The display unit 4 is constructed of a cathode ray tube display device or the like, and displays the output from the calculation unit 3, that is, the amount indicating the degree of similarity, in correspondence with each part of the brain, by means of brightness modulation, contour stripes, color, etc. It is.

As detailed above, according to the electroencephalogram waveform correlation display device according to the present invention, the degree of similarity of electroencephalogram waveforms of a certain time width derived from a plurality of parts of the head is quantified and displayed. Regardless of the high or low frequency, the relationship between each part of the brain can be observed very easily and in a short time.

Moreover, compared to methods that calculate the cross-correlation, cross-spectrum, coherency function, etc. of two brain waves, the program is simpler and the calculations are easier, so the entire device can be made smaller by using a small-capacity calculation section. .

[Brief explanation of drawings]

Fig. 1 is a block diagram of the device according to the present invention, Fig. 2A
-C is an explanatory diagram showing an example of a method of quantifying waveform similarity. 1...Deriver, 2...Amplifier, 3...
...Calculating section, 4...Display section.

Claims (1)

[Claims] 1. A large number of derivators attached to the human head to derive brain waves, an amplifier to amplify the derived brain waves, and two adjacent derivation sites each having a fixed time width of the brain wave waveforms derived at those sites. An electroencephalogram waveform correlation display device comprising: a calculation section that quantifies the degree of similarity; and a display section that displays the quantified degree of similarity in correspondence with each region of the brain.