Application of Magnetoencephalography (MEG) to Analyze Changes in Whole-Brain Oscillatory Microstate Patterns Across the Lifespan: Cambridge Centre for Aging and Neuroscience Cohort Study

Research Background

With the increasing seriousness of the aging population problem, understanding the neurophysiological changes during the aging process becomes more critical. The aging brain is a major risk factor for many neurodegenerative diseases; however, how whole-brain oscillatory activity affects healthy aging is not fully understood. At the cellular level, the bioelectrochemical properties of neurons enable them to generate electromagnetic fields, and detecting changes in these fields can serve as potential tissue pathological biomarkers. Five typical oscillatory brain signals (delta, theta, alpha, beta, and gamma waves) have been widely studied, but their specific roles in aging await further exploration. Researchers propose that monitoring the regularity and predictability of these oscillatory signals can help identify potential processes of cognitive decline. Specifically, changes in alpha waves exhibit significant characteristics during aging, such as slowing of alpha waves, reduction in alpha power, and diminished alpha reactivity.

Research Source

This paper was co-authored by Yujing Huang, Chenglong Cao, Shenyi Dai, Hu Deng, and other scholars from institutions such as Zhejiang University, the University of Cambridge, and the University of Science and Technology of China. The study was published in the journal Brain Communications on April 29, 2024.

Research Process

Demographic Data

The study was conducted based on the Cambridge Centre for Aging and Neuroscience (Cam-CAN) cohort, comprising 624 participants aged 18 to 88 years. Participants were divided into five age groups: young (18-29 years), early middle age (30-44 years), late middle age (45-59 years), early old age (60-74 years), and old age (75-88 years). All participants underwent resting-state MEG scans.

MEG Data Recording and Preprocessing

MEG data were recorded using a 306-channel Vectorview system, including 204 planar gradiometers and 102 magnetometers. Participants underwent resting-state MEG recordings in a magnetically shielded room, with data recorded for 8 to 40 seconds at a sampling rate of 1 kHz, and preprocessed with a high-pass filter (0.03 Hz). Subsequently, noise removal, motion artifact correction, further filtering, and segmentation processing were carried out through software.

Microstate Pattern Analysis

The study employed an improved k-means clustering algorithm to extract microstate patterns of various oscillatory frequency bands. Specific steps included importing MEG preprocessed data, extracting activity maps at peak time points, running k-means clustering, performing two levels of “global clustering” to obtain global patterns for each age group and frequency band, fitting microstate parameters for each age group, and conducting statistical analyses. Additionally, a series of statistical analyses were conducted to compare the differences in microstate parameters among different age groups.

Main Research Results

Identification of Microstate Patterns

Four typical microstate patterns were identified using machine learning algorithms: left-to-right (MS1), right-to-left (MS2), front-to-back (MS3), and front-central (MS4). These four patterns demonstrated consistent dominance among different age groups and frequency bands.

Multifrequency Oscillatory Attenuation and Alpha Wave Changes

The study found that as age increased, the duration of alpha waves in MS1 and MS2 decreased, while the occurrences of alpha waves increased. Furthermore, changes in theta and beta waves in MS1 may be associated with age-related declines in motor function. The occurrences of alpha waves in MS3 and MS4 also increased, with corresponding changes in beta wave activity, possibly reflecting changes in the top-down salience/attention network.

Research Conclusions

This study both summarizes typical whole-brain oscillatory microstate patterns during aging and provides new insights for predicting healthy aging and potential neuropsychiatric cognitive decline. The microstate patterns identified by machine learning algorithms demonstrate how the brain exhibits functional deficits across different age groups, particularly highlighting the dominant role of alpha waves. The results indicate an increase in global alpha wave activity in the brain during aging, signifying reduced neural efficiency, along with accompanying changes in other frequency bands, such as theta and beta waves.

Research Highlights

1. Identification of Typical Microstate Patterns: Using machine learning algorithms, the study successfully identified four typical whole-brain microstate patterns, which consistently manifested across different age groups and frequency bands.

2. Highlighting the Critical Role of Alpha Waves: The study emphasized significant changes in alpha waves during aging, particularly the reduction in alpha wave duration and the increase in occurrences.

3. Multifrequency Response Attenuation: Apart from alpha waves, changes in theta and beta waves were also detected, which are closely related to the decline in sensory and motor functions.

Research Significance

This study holds significant scientific value for understanding the neural dynamics of the aging brain and provides new biomarkers for predicting healthy aging and early diagnosis of neuropsychiatric cognitive decline. The findings can aid in developing relevant prevention and treatment measures, thereby improving the quality of life for the elderly population.

Future Research Directions

Future research can further integrate resting-state MEG, functional and structural magnetic resonance imaging techniques to obtain multimodal aging biomarkers. Additionally, long-term follow-up observations are needed to better understand the relationship between these microstate patterns and brain atrophy.