JPH0779937A - Detection system for arousal degree - Google Patents

Abstract

PURPOSE:To detect with certainly the condition of a driver dozing off simply through the sensing of the brain waves even if the pattern of the frequency distribution of the brain waves varies by counting the number of times when the amplitude of alpha2 wave, alpha1 wave, and theta wave exceeds the standard amplitude value respectively and passing judgement on the driver's arousal degree depending on the sum total of the count numbers of respective components of the brain waves. CONSTITUTION:Brain waves which are sensed via electrodes are sent by wireless from the transmitting antenna 2 of a brain wave sensing unit 1 to the receiving antenna 3 and then input into a frequency analyser 4, alpha2 wave, alpha1 wave, and thetawave are extracted in the frequency analyser 4 and input into a control unit 5 in which the amplitude values of respective brain wave signals are compared in turn with the preset standard values of large amplitude and when the former exceeds the latter, pulse signals are put out. The numbers of the pulse signals for respective brain waves are counted within the specified time period and judgement on the arousal degree is passed on the basis of the sum total values obtained by adding up the count numbers for respective brain waves.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wakefulness determination device for determining a wakefulness from an electroencephalogram to detect a dozing state.

[0002]

2. Description of the Related Art EEGs change their waveforms according to the blocking of light and the activity level of the cerebrum, and are generally classified into five types according to frequency and amplitude as shown in Table 1.

[0003]

[Table 1]

Further, the frequencies of α waves are classified into three levels, α1 waves (7 to 9 Hz) and α2 waves (9 to 11H).
z) and α3 waves (11 to 13 Hz), and it is said that an increase in amplitude and a decrease in frequency indicate a decrease in cerebral activity level. Regarding the relationship between the electroencephalogram and the arousal level, generally, the appearance of α2 waves (9 to 11 Hz) is seen in the early stage of the dozing state, and the amplitude and the appearance amount increase with the decrease in the arousal level. . Further, as the awakening level decreases, the frequency component of the brain wave shifts to the lower side, and when the consciousness disappears, the α wave disappears and the θ wave (4 to 7 Hz) appears.

Therefore, conventionally, as a method of detecting a dozing state from an electroencephalogram, a method of detecting a θ wave disclosed in Japanese Patent Laid-Open No. 50-107897, and Japanese Patent Laid-Open No. 55-86.
There is one that detects an α wave that appears in the early days of dozing, as shown in No. 443. In addition, JP-A-2-151532
Like the "drowsiness warning device" disclosed in the publication,
Some of them judge dozing depending on the magnitude of the frequency component of the slow α (α1).

[0006]

However, in the conventional method of detecting a dozing state by the electroencephalogram, the individual difference in the appearance pattern of the frequency component of the electroencephalogram is not taken into consideration. Here, perform monotonous work, measure EEG data of multiple people from awake state to dozing state,
As a result of investigating the appearance pattern of the electroencephalogram in the state where the degree of arousal is lowered, as shown in FIGS. 6 to 9, four patterns A to D are roughly classified. It should be noted that the state of decrease in arousal level was determined by viewing together with the electro-oculogram (representing blinking speed and number of blinks).

The pattern A in FIG. 6 is a general pattern, and α is set in the initial stage of the lowering of the arousal level (the initial stage of the lowering of the arousal level in the figure).
Two waves (9 to 11 Hz) appear, and the amplitude and the amount of appearance increase as the awakening level decreases, and the high-amplitude α2 wave appears frequently. In the pattern B of FIG. 7, α1 waves appear in addition to the α2 waves having a slightly high amplitude in the early stage of the decrease in arousal level.

In the pattern C of FIG. 8, high-amplitude α2 waves and α1 waves appear similarly in a state where the arousal level is rapidly lowered, but soon the α1 waves become dominant. In the D pattern of FIG. 9, the appearance of the θ wave can be seen in the state where the arousal level is rapidly reduced similarly to the C pattern. Also, α2 wave and α1
The waves are appearing a little later. In this way, there are individual differences in the appearance pattern of the electroencephalogram that accompanies the decrease in alertness,
As in the prior art, in the dozing detection of only the α1 wave or the θ wave, it is difficult to detect the A pattern in which the α2 wave is more dominant. In addition, for example, when only the α2 wave is detected, it is difficult to detect a C pattern in which the α1 wave is predominant or a D pattern in which the θ wave appears before the α2 wave, and even for a B pattern in which the α1 wave appears together with the α2 wave. It's hard to say that we can handle it. Furthermore, there is a problem that the combination of α2 wave and θ wave cannot deal with the dominant C pattern of α1 wave, and the combination of only α1 wave and α2 wave cannot deal with D pattern where θ wave appears. The determination accuracy of the arousal level was not sufficient.

Note that, as described above, the present inventors have found that there are individual differences in the appearance pattern in the determination of only the electroencephalogram. Therefore, as a method of determining the arousal level, the electroencephalogram (α2 wave) is
Conventionally, a method of evaluating arousal level by combining electro-oculogram and facial expression has been used. This is based on the appearance state of the α2 wave, the number of blinks and the speed state based on the electro-oculogram, and the facial expression, respectively, and from the total point, a state of clear awakening to a state of faintness just before falling asleep. The awakening level is determined, and the accuracy of the awakening level determination is improved by combining a plurality of indexes in this way.

However, such a method of determining the arousal level using indexes other than the electroencephalogram is inefficient because it requires a plurality of detecting means. The present invention has been made in view of the above circumstances, and provides a high-accuracy awakening degree determination device that can detect a dozing state at a reliable and appropriate time only by detecting an electroencephalogram even when the appearance pattern of the frequency component of the electroencephalogram is different. The purpose is to do.

[0011]

Therefore, in the arousal level determination device according to the present invention, as shown in FIG. 1, an electroencephalogram detecting means for detecting an electroencephalogram and a frequency analysis of the electroencephalogram detected by the electroencephalogram detecting means. And a frequency analysis means for extracting the α2 wave, the α1 wave, and the θ wave, and the α2 wave extracted by the frequency analysis means.
Comparing means for comparing the respective amplitude values of the wave, the α1 wave and the θ wave with a preset reference amplitude value, and counting the number of times of the α2 wave, the α1 wave and the θ wave which have become the reference value or more within a predetermined time. The measuring means, the calculating means for calculating the total of the respective count values measured by the measuring means, and the judging means for judging the awakening degree based on the total value calculated by the calculating means.

[0012]

In this structure, the electroencephalogram detected by the electroencephalogram detection means is frequency-analyzed by the frequency analysis means to obtain α2 waves and α2 waves.
One wave and theta wave are extracted. Next, each amplitude value of the α2 wave, α1 wave, and θ wave extracted by the comparison means is compared with a preset reference amplitude value, and the number of times of each electroencephalogram having a reference amplitude value or more is counted by the counting means, The total value is calculated by the calculation output. Then, the determination means determines the awakening degree according to the calculated total value.

[0013]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 2 shows the configuration of the first embodiment of the awakening degree determination device according to the present invention, which is an embodiment when applied to an automobile. In FIG. 2, an electroencephalogram detector 1 as an electroencephalogram detection means has, for example, a cap shape in which electrodes are integrally provided,
The driver can wear it on his head. In this case, it is desirable that the electrodes are brought into contact with the driver's crown and occipital region in two places to adopt a bipolar induction system. The electroencephalogram signal detected through the electrodes is received by a reception antenna 3 installed at a location wirelessly separated from a transmission antenna 2 provided integrally with the electroencephalogram detector 1, and input to a frequency analyzer 4 as a frequency analysis means.

In the frequency analyzer 4, as shown in FIG.
α2 wave (9 to 11 Hz), α1 wave (7 to 9 Hz) and θ
Waves (4-7 Hz) are extracted. Each extracted EEG (α
(2 wave, α1 wave, θ wave) signal is sent to the control unit 5
4, the amplitude value of each electroencephalogram signal is sequentially compared with a preset large amplitude reference value L, and a pulse signal is output when the large amplitude reference value L or more, Within a predetermined time, this pulse signal is counted for each frequency band, that is, for each electroencephalogram, the count value for each electroencephalogram is added, and the awakening degree is determined based on the added value. Therefore, the control unit 5 has the functions of comparing means, measuring means, calculating means, and determining means.

Based on the result of the awakening degree judgment in the control unit 5, when the awakening degree is lowered, the aroma device 6 for generating a scent according to the degree or the alarm device 7 such as a buzzer is driven. In this way, the driver is warned of the dozing state and awakened. Next, the awakening degree determination operation of this embodiment will be described with reference to the flowchart of FIG.

First, an electroencephalogram detector 1 worn by the driver detects an electroencephalogram (101) and starts a timer (102). Next, by the frequency analyzer 4, the α2 wave,
The α1 wave and the θ wave are extracted (103). Each extracted α2
The amplitude values of the wave, α1 wave, and θ wave are compared with a large amplitude reference value L as a reference value as shown in FIG.
06, 108), if L or more, the process proceeds to YES, and the pulse signal generated at that time is counted for each frequency component, that is, for each brain wave (α2 wave, α1 wave, θ wave) (105,
107, 109). On the other hand, if the actual amplitude value is smaller than the large amplitude reference value L, the process proceeds to NO and frequency analysis is performed again (103). Here, as the large amplitude reference value L, the average amplitude value of the α2 wave at the time of rest of eyes closed before the start of driving is used, and it is measured and stored in advance before the start of driving.

Next, when counting, it is judged whether or not the timer time has reached a predetermined time (110), and when it has reached, the judgment is YES, and α2 wave, α1 wave,
The total value N of the count values n of the θ wave is calculated (11
1). Then, it is judged whether or not the total value N is less than 5 (112), and if it is less, it is judged as the arousal level A (11).
4). When the total value N is 5 or more, it is determined whether it is 5 or more and less than 10 (113), and when it is less than 10, it is determined as the arousal level B (115), and when the total value N is 10 or more. , Awakening degree C (116). After determining the awakening level for the first predetermined time, clear the timer (11
7) Then, the timer is started again (102) and the same processing is repeated.

Then, for example, when the arousal level is A, it is determined that the arousal state is in effect, and the incense device 6 and the alarm device 7 are not driven. Only the device 6 is driven to present a scent with awakening action such as peppermint. The fragrance alone has a wakefulness effect even in the initial stage of awakening lowering and is a soft stimulus, so it is not bothersome to the driver and is appropriate in the initial state of awakening lowering. When the arousal level is C, it is determined that the arousal level has decreased and the alarm device 7 is driven to notify the driver with an alarm sound, and then the fragrance device 6 presents a scent. As a result, by exhibiting a scent after the arousal level is increased to some extent by a strong stimulus such as an alarm sound, the arousal effect can be maintained even when the arousal level is reduced.

As described above, if the arousal level is determined based on the appearance frequencies of the high-amplitude α2 wave, α1 wave, and θ wave that appear with a decrease in the awakening level, the influence of individual differences in the appearance pattern of the brain wave can be determined. It is possible to detect drowsiness without receiving it, and it is possible to determine the awakening degree with high accuracy. In addition, it is possible to accurately determine the arousal level by using only the electroencephalogram without combining other indices such as the electro-oculogram and the facial expression, and no other detection device is required.

Next, a second embodiment of the awakening degree judging device of the present invention will be explained. The hardware configuration of this embodiment is shown in FIG.
Since it is the same as the first embodiment shown in FIG. 9 and only the software configuration in the control unit is different, only the awakening level determination operation in the control unit will be described here with reference to the flowchart in FIG. In this embodiment, as the frequency is lower and the amplitude is larger, the count value is weighted higher, so that the awakening degree can be determined with higher accuracy than that in the first embodiment. It is a thing.

First, an electroencephalogram is detected (201), a timer is started (202), and α2 wave, α
The extraction of 1 wave and θ wave (203) is the same as in the first embodiment. Next, in the present embodiment, α2 wave, α1 wave, θ
The amplitude value of the wave is compared with the large amplitude reference value L and the small amplitude reference value S, respectively (204, 206, 208, 210,
212, 214), the number of amplitude values greater than the large amplitude reference value and the number of amplitude values greater than the small amplitude reference value are counted for each frequency component (205, 207, 209, 211, 213, 21).
5). Here, the small amplitude reference value S uses the average amplitude value of the α2 wave at the time of rest of the eyes before the start of driving, and is measured and stored in advance before the start of driving. The large amplitude reference value L is set similarly to the first embodiment.

Then, when counting each amplitude value, α
When the amplitude values of the two waves are greater than or equal to the large amplitude reference value L, the count value is set to 1 for one pulse, and when less than the large amplitude reference value L and greater than or equal to the small amplitude reference value S, one pulse And set the count value to 0.5. That is, the pulse when the α2 wave has the small amplitude reference value or more is weighted so that two pulses are counted as one pulse when the α2 wave has the large amplitude reference value or more.

Similarly, when the pulse of the α1 wave has a large amplitude reference value or more, one pulse has a count value of 1.5, and when the pulse has a small amplitude reference value or more, the pulse has a count value of 1 and a large θ wave. The pulse when the amplitude reference value is equal to or larger than 1 pulse has a count value of 2, and the pulse when the amplitude reference value is equal to or larger than the small amplitude reference value has a count value of 1.5, respectively. Next, when counting, it is judged whether or not the timer time has reached a predetermined time (216), and when it has reached, the judgment is YES, and the large amplitude reference value L of the α2 wave, α1 wave, and θ wave. As described above, the total N of all count values equal to or larger than the small amplitude reference value S is calculated (217). Then, based on the total value N, the total value N is N <5, 5 ≦ N <10, or 10 ≦ N.
It is determined whether <15, 15 ≦ N <20, or N ≧ 20 (2
18, 219, 220, 221), and the arousal level is determined in five stages from the arousal level A to the arousal level E (222 to 226).
When the awakening level of the first predetermined time is judged, the timer is cleared (227) and the timer is started again (2
02), the same processing is repeated.

Then, the intensity of the scent from the fragrance device 6, the intensity of the alarm sound of the alarm device 7, etc. are adjusted according to the determination result of the awakening level to awaken the driver without causing any trouble. Try to make it into a state. In this way, considering that the increase in the amplitude of the electroencephalogram and the decrease in the frequency indicate the decrease in the activity level of the cerebrum, the larger the amplitude, the lower the frequency,
By highly weighting the count value, it is possible to determine the awakening degree with higher accuracy.

[0025]

As described above, according to the present invention, the α2 wave, the α1 wave, and the θ wave are extracted from the detected brain waves, and the number of times when the amplitude of each of these brain waves becomes equal to or larger than the reference amplitude value is counted. Since it is configured to determine the arousal level based on the total value, it is possible to reliably determine the arousal level for all persons without being affected by individual differences in the appearance pattern of the electroencephalogram. Highly accurate alertness determination is possible. Further, since only the electroencephalogram is required to be detected, only the electroencephalogram detection means is required as the detection means, and the other detection means for determining the arousal level can be eliminated.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating a configuration of a wakefulness determination device according to the present invention.

FIG. 2 is a configuration diagram showing a first embodiment of a wakefulness determination device according to the present invention.

FIG. 3 is a waveform diagram of an electroencephalogram signal extracted in the first embodiment.

FIG. 4 is a flowchart for explaining the awakening level determination operation of the first embodiment.

FIG. 5 is a flowchart illustrating an awakening degree determining operation of the second embodiment of the awakening degree determining apparatus according to the present invention.

FIG. 6 is a diagram showing an example of an electroencephalogram pattern that appears with a decreased alertness.

FIG. 7 is a diagram showing another example of an electroencephalogram pattern that appears with a decreased alertness.

FIG. 8 is a diagram showing another example of an electroencephalogram pattern that appears with a decreased alertness.

FIG. 9 is a diagram showing another example of an electroencephalogram pattern that appears with a decreased alertness.

[Explanation of symbols]

 1 EEG detector 4 Frequency analyzer 5 Control unit

Claims (1)

[Claims] 1. An electroencephalogram detecting means for detecting an electroencephalogram and a frequency analysis of the electroencephalogram detected by the electroencephalogram detecting means are performed to obtain α
Frequency analysis means for extracting two waves, α1 waves and θ waves, and comparison means for comparing each amplitude value of the α2 waves, α1 waves and θ waves extracted by the frequency analysis means with a preset reference amplitude value Measuring means for counting the respective numbers of α2 waves, α1 waves, and θ waves that have become equal to or greater than a reference value within a predetermined time; and calculating means for calculating the sum of the respective count values measured by the measuring means, An awakening degree determination device comprising: a determination means that determines the awakening degree based on the total value calculated by the calculation means.