Predicting your brain age with portable EEG and machine learning

Key takeaways:

• Brain age is a helpful measurement to assess cognitive performance
• Brain age can be measured by either sleep or meditation EEG data
• Muse headbands, combined with machine learning, can calculate brain age in real-time
  
  

Aging is a difficult topic for many to process. Understanding it demands that we accept our mortality, and studying it can reveal the ways in which we can optimize, slow, or even negate some of the impacts of age on our performance. 

While we can easily see and feel the biological signs of it – like wrinkled skin, gray hairs, or physical decline – the psychological impact of aging is subtle and harder to detect. We know that age can impact cognition, but there’s a pretty wide gap in how that plays out in real life. You can still be a virtuoso musician or a prolific writer well into your old age, but that isn’t always the norm, either. And while we can take a good guess at someone’s physical age based on their appearance, their brain age, that is, their relative cognitive performance, is difficult if not impossible to measure with the naked eye. 

However, the arrival of portable EEG technology, combined with the power of machine learning, has made it possible to realistically (and quickly) calculate brain age. A recent study by Muse, published in Imaging Neuroscience, offers promising insight into how we can assess brain age by measuring changes in sleep and neurological activity patterns. 

Developing the real-world brain age metric

You may have heard of a brain age measurement as the payoff to a game of brain teasers or a series of quizzes. The clinical viability of sudoku notwithstanding, the popularity of such games reveals a growing desire to understand and analyze our cognitive performance – and perhaps even our fears about getting older. 

Recently, by analyzing age-related patterns in the brain’s structure and/or function, researchers are developing AI models that can predict a person’s brain age from brain recordings (MRI, EEG, etc.), and are looking at its deviation from their chronological (actual) age, which they call the brain age index. They have found that a predicted brain age that is lower than a person’s chronological age may indicate premature aging or neurological conditions that cause the affected brains to appear older.

Clinically-sound brain age studies have so far been largely limited to clinical settings due to the complexity of long-term measurement. There is a growing body of literature that suggests M/EEG signals can be used to establish brain age benchmarks at a population level, but everyday use of EEG technology in brain age studies has been limited by the challenges of running daily sessions and affordable consumer access to EEG equipment. However, the advent of wearable research-grade EEG devices like the Muse headband have brought the power of neuroscience out of the lab and into our homes.

The recent study, conducted in partnership with Université Paris-Saclay, set out to establish a baseline for a real-world brain age metric based on measurable values using consumer-grade EEG recordings via the Muse S headband from more than 5,200 individuals aged 18 to 81. Using this data set, the researchers established a machine learning model to predict brain age. Data was collected in two settings, during meditation and during sleep in at-home environments, and then fed into a machine learning algorithm to predict and evaluate brain age, and subsequently evaluated for variability among longitudinal recordings for up to 529 days.  

Sleep and focus can yield brain age

The study found that brain age measurement can be expanded to at-home settings by measuring neurological activity during meditation activities and sleep to “capture known age-related electrophysiological phenomena.” Certain EEG characteristics, namely 𝛼𝑙𝑜𝑤 and 𝛼mid power for meditation, N2  𝛼ℎ𝑖𝑔ℎ power in the spindle range, and N3 𝛿, helped accurately predict brain age.

The study found that brain age predictions made using sleep recordings alone yielded more accurate results than by just using meditation data.Together, they can produce reasonable and comprehensive brain age predictions. The study benefited from the fact that participants could record their data from home, which massively improved the size and quality of the data researchers could process.

Variability in brain-age over time

Another notable finding in the research was that individuals experienced fluctuations in daily brain age measurements. While you might expect a steady decline in cognitive function over a lifetime, it turns out that the participants brains’ appeared a few years ‘younger’ or ‘older’ at different moments in time depending on a variety of factors, like mood, mental state, or even the quality of their sleep the night before. To further explore the causes of brain age variability, the researchers suggest gathering data over an even longer period of time.

Bringing neuroscience labs into your home

As the science around aging grows, brain age tests can give us meaningful goal posts with which to measure and compare cognitive function across time. For researchers, it can help quantify and compare biomarkers among different cohorts, providing additional context to understand neurological and physical disorders and their impact on cognitive function. For individuals, understanding brain age can provide insight into their wellbeing both at a particular moment and over time, helping them make thoughtful, data-driven changes to their lifestyle that can optimize their performance. Additionally, this data can help individuals understand how their unique circumstances are impacting their cognition. 

Because Muse can help both researchers and individuals collect and evaluate brain age from the comfort of home, the possibilities for experimentation and data collection are nearly endless. Muse headbands can free researchers from the logistical challenges of lab-based data collection, empowering them to collect more data from more subjects over longer periods of time. And individuals can access research-grade data about themselves without needing to book appointments, get referrals, or see specialists just to quantify how they’re feeling. Whether they’re biohackers looking to maximize performance, athletes fine-tuning their routines, or even regular people trying to optimize their lifestyles and understand themselves, brain age can provide never-before-seen insights into their brains.

Making sense of your mind with Muse

Whether you’re researching brain age and neuroplasticity, or trying to improve your focus at work, Muse can help you master your mind. Muse's portable EEG technology is research-backed and has been featured in over 200 third-party research studies to date and is trusted by leading institutions like Harvard, NASA and The Mayo Clinic. To find out how Muse can help you level up your research, check out our library of research studies or visit our store to learn more.