[Special Feature: Dementia and Society] Yasue Mitsukura: Detecting Dementia Risk through EEG Measurement

2022/11/07

Japan's elderly population is expected to reach approximately 40 million by 2040, leading the country into an unprecedented super-aging society. Looking at the near future, the population of people with dementia is said to reach approximately 7 million by 2025—remarkably, one in five people aged 65 or older. This era is coming in just three years. Most cases of dementia are Alzheimer's-type dementia, caused by the loss of brain cells and subsequent brain atrophy. It often begins with forgetfulness and progresses slowly and gradually. While Alzheimer's drugs are effective in improving symptoms, they do not have a fundamental curative effect; therefore, even if cognitive functions such as memory improve with medication, the disease gradually progresses. In other words, early detection at the initial stage is vital for the treatment of Alzheimer's disease.

In this context, the government's Comprehensive Strategy for the Promotion of Dementia Measures (commonly known as the New Orange Plan) outlines seven pillars, including (1) promotion of dissemination and awareness to deepen understanding of dementia, and (2) provision of timely and appropriate medical and nursing care according to the condition of dementia. Among these, the strengthening of measures for early-onset dementia is mentioned, emphasizing early detection to "delay the onset of dementia" and "slow the progression even if dementia occurs." We resonate with this initiative and have developed a system that can easily measure Mild Cognitive Impairment (MCI). Our group has clarified the following two points:

・ Measuring the EEGs of people with dementia and Mild Cognitive Impairment (MCI) suggests that each has distinct characteristics compared to healthy individuals.

・ Furthermore, this is possible through simple EEG measurement.

In short, it becomes possible to identify healthy individuals, MCI patients, and dementia patients through simple EEG measurement alone!

Using an electroencephalograph (EEG) that can measure brain waves easily anytime and anywhere, we measured the EEGs of 120 subjects divided into three groups: healthy, MCI, and dementia. We were able to clarify the characteristics of EEG frequencies for each measured group (Figure 1).

To begin with, what you likely hear most often regarding brain waves are things like "high alpha waves = concentrating," "high theta waves = sleepy," or "listening to a certain song produces alpha waves." These terms "alpha" and "theta" are the names of the bands that appear when the signals obtained from the EEG sensors on the head are converted into frequencies.

Normally, our brain waves are simply microvolt waves obtained as signals on the scalp. When these waves are converted into frequencies, human brain waves generally fall within values up to about 45 Hz. Within this range, up to 4 Hz is called delta waves, 4–6 Hz is theta waves, 7 to 13 Hz is alpha waves, around 14 to 23 Hz is beta waves, and anything above that is called gamma waves (definitions of frequency bands may vary depending on the researcher).

These alpha and beta waves are referred to as bands. Naturally, brain waves contain a significant amount of noise when acquired (sometimes over 60% is noise). These bands must be calculated after removing this noise in real-time. As long as noise removal is performed, brain waves can be captured cleanly using any device. However, if noise is not accurately removed, the data will naturally be completely different.

We have obtained a patent for an algorithm that removes noise from EEG data in real-time. Using this, it has become possible to easily measure brain waves in any situation. In the past, measuring brain waves meant entering a special room called a shielded room to prevent noise and measuring while at rest. Now, however, they can be acquired in a normal state, allowing brain waves to be measured naturally and accurately according to the situation (it's not normal to have to go into a shielded room just to take an EEG!).

Furthermore, older EEGs used multiple electrodes—for example, as many as 64 channels—attached to the brain to gather various information. Of course, many electrodes are still attached to the head in some cases today, but it is difficult to tell someone not to feel stressed in that state. Between the feeling of being constricted and the fact that it takes nearly an hour just to put the EEG on, it can be a very stressful situation. We select the locations to attach electrodes based on the objective. Moreover, ideally, we aim to use as few electrodes as possible. By reducing the number of electrodes to suit the purpose, we reduce the burden. Figure 2 shows an EEG measurement system that has reduced the number of electrodes in this way and can perform noise removal in real-time.

Based on the information obtained from this EEG, and by using the characteristics of the EEG bands for dementia, MCI, and healthy individuals introduced earlier, anyone can determine whether they are in a healthy state or a state of dementia or MCI just by acquiring their brain waves. The EEG used is a non-burdensome headband type that takes about 15 seconds to put on and 15 seconds for calibration before measurement begins, allowing for measurement in a total of about 30 seconds. Even if the head moves, noise is instantly removed for stable measurement, making it easy to measure anywhere.

By using these in daily life, it is expected that people will be able to easily measure the possibility of MCI or dementia while staying at home without going to a hospital. Until now, people had to make an effort to find time to go out, undergo MRI or PET scans at a hospital, and then go back to hear the results. However, it will now be possible to know the likelihood of being healthy, having MCI, or having dementia with a simple device. If discovered at the MCI stage, it is possible to delay progression.

Above all, it is necessary to be aware of one's own condition. Previously, this could only be done at a hospital, but since determination is possible through simple measurement, you can know your condition at home. In the future, I believe that even a headband-type EEG will not be necessary, and the determination will be possible with, for example, a ring sensor or a wristband-type sensor. We will not let the saying that one in five people will have dementia in 2025 come true. Early detection can slow its progression. Currently, drug discovery is also progressing by leaps and bounds. There is a possibility that a "cure" for dementia that did not exist yesterday will be created tomorrow. That is why, even if MCI is found, we delay its progression. By doing so, a drug to cure dementia will eventually emerge.

Our efforts are not limited to dementia. We are also developing devices that can easily measure mental illnesses such as depression and sleep states. We are building a system that allows people to understand their condition easily, anytime, anywhere, even at home. Utilizing the power of a comprehensive university, which is the greatest feature of Keio University, we are prepared to confront dementia through medicine-engineering-pharmacy collaboration. Until dementia becomes a curable disease. And so that we can welcome a bright 2040.