Estimating brain age from structural MRI and MEG data: Insights from dimensionality reduction techniques

Abstract

Brain age prediction studies measure the difference between the chronological age of an individual and their predicted age based on neuroimaging data, which has been proposed as an informative measure of disease and cognitive decline. As most previous studies relied exclusively on magnetic resonance imaging (MRI) data, we hereby investigate whether combining structural MRI with functional magnetoencephalography (MEG) information improves age prediction using a large cohort of healthy subjects (N=613, age 18-88) from the Cam-CAN. To this end, we examined the performance of dimensionality reduction and multivariate associative techniques, namely Principal Component Analysis (PCA) and Canonical Correlation Analysis (CCA), to tackle the high dimensionality of neuroimaging data. Using MEG features yielded worse performance when compared to using MRI features, but the combination of both feature sets slightly improved age prediction (mean absolute error of 5.28 yrs). Furthermore, we found that PCA resulted in worse performance, whereas CCA in conjunction with Gaussian process regression models yielded the best prediction performance. Notably, CCA allowed us to visualize the features that significantly contributed to age prediction. We found that MRI features from subcortical structures were more reliable age predictors than cortical features, and that spectral MEG measures were more reliable than connectivity metrics. Our results provide an insight into the underlying processes that are indicative of brain aging, thereby advancing the discovery of valuable biomarkers of neurological syndromes that emerge later during the lifespan.

A reusable benchmark of brain-age prediction from M/EEG resting-state signals

“…First MEG-based brain-age models have allowed to validate MEG-derived brain age against MRIderived brain age. Results from several studies have shown that the MEG-and MRI-derived brain age are statistically related, leading to overlapping correlations between ensuing brain age estimates (Engemann et al 2020;Sabbagh et al 2020;Xifra-Porxas et al 2021) and individual differences in cognition and health. This overlap can be explained by electromagnetic field spread, independently of neuronal activity: As brain structure changes due to aging, cortical activity, even if unchanged, will project differently onto the M/EEG sensor array, making age indirectly decodable (Sabbagh et al 2020).…”

“…This overlap can be explained by electromagnetic field spread, independently of neuronal activity: As brain structure changes due to aging, cortical activity, even if unchanged, will project differently onto the M/EEG sensor array, making age indirectly decodable (Sabbagh et al 2020). Importantly, multiple articles have found that neuronal activity captured by MEG adds specific information not present in MRI-derived brain age (Engemann et al 2020;Xifra-Porxas et al 2021), leading to improved prediction performance and richer neurocognitive characterization (Engemann et al 2020).…”

Repurposing EEG monitoring of general anaesthesia for building biomarkers of brain ageing: An exploratory study

“…The 'power spectrum' model presented uniformly spaced frequencies between (1-30Hz) averaged across all 4 electrodes (16 features), representing the hypothesis of distributed spectral effects. Finally, the 'spatial patterns' model implemented the brain age model from a reference publication 4 which analysed the covariance in the above 5 frequency bands (low [0.1-1.5]Hz, delta [1.5-4]Hz, theta [4][5][6][7][8]Hz, alpha [8][9][10][11][12][13][14][15]Hz and beta [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]Hz), representing the hypothesis of distributed spatio-spectral effects (35 features). The model concatenated the band-specific EEG covariances after vectorization with Riemannian embeddings that accounted for nonlinearities caused by EEG-volume conduction 7 .…”

“…The bulk of the brain age literature is based on MRI 19,24,26,27 . Nevertheless, a recent line of research has demonstrated that brain age can be estimated from EEG and may even contribute additional information over MRI 20,28,29 . It is currently unknown if brain age can be estimated from EEG during GA.…”