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

Abstract

Population-level modeling can define quantitative measures of individual aging by applying machine learning to large volumes of brain images. These measures of brain age, obtained from the general population, helped characterize disease severity in neurological populations, improving estimates of diagnosis or prognosis. Magnetoencephalography (MEG) and Electroencephalography (EEG) have the potential to further generalize this approach towards prevention and public health by enabling assessments of brain health at large scales in socioeconomically diverse environments. However, more research is needed to define methods that can handle the complexity and diversity of M/EEG signals across diverse real-world contexts. To catalyse this effort, here we propose reusable benchmarks of competing machine learning approaches for brain age modeling. We benchmarked popular classical machine learning pipelines and deep learning architectures previously used for pathology decoding or brain age estimation in 4 international M/EEG cohorts from diverse countries and cultural contexts, including recordings from more than 2500 participants. Our benchmarks were built on top of the M/EEG adaptations of the BIDS standard, providing tools that can be applied with minimal modification on any M/EEG dataset provided in the BIDS format. Our results suggest that, regardless of whether classical machine learning or deep learning was used, the highest performance was reached by pipelines and architectures involving spatially aware representations of the M/EEG signals, leading to R^2 scores between 0.60-0.71. Hand-crafted features paired with random forest regression provided robust benchmarks even in situations in which other approaches failed. Taken together, this set of benchmarks, accompanied by open-source software and high-level Python scripts, can serve as a starting point and quantitative reference for future efforts at developing M/EEG-based measures of brain aging. The generality of the approach renders this benchmark reusable for other related objectives such as modeling specific cognitive variables or clinical endpoints.Highlights- We provide systematic reusable benchmarks for brain age from M/EEG signals- The benchmarks were carried out on M/EEG from four countries > 2500 recordings- We compared machine learning pipelines capable of handling the non-linear regression task of relating biomedical outcomes to M/EEG dynamics, based on classical machine learning and deep learning- Next to data-driven methods we benchmarked template-based source localization as a practical tool for generating features less affected by electromagnetic field spread- The benchmarks are built on top of the MNE ecosystem and the braindecode package and can be applied on any M/EEG dataset presented in the BIDS format

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

“…This study explored the feasibility of repurposing EEG-monitoring data from GA for building measures of brain ageing by importing machine learning approaches for brain age prediction originally developed in the laboratory setting. Under propofol anaesthesia, we reached prediction performance comparable to reference studies using research-grade EEG 28,33 . Model comparisons revealed that age-related information was present in spatial activity patterns distributed across the entire power spectrum.…”

“…with high-density MEG and EEG 20,32 . Previous studies during anaesthesia have instead focussed on EEG-signatures closely related to anaesthesia monitoring, with particular emphasis on total power and alpha power 6,36 .…”

“…prior art by directly applying brain-age prediction models for EEG 28 on a larger dataset of EEG during anaesthesia. Under propofol, this approach has led to performance matching recent work with high-density MEG and EEG 20,37 . Previous studies during anaesthesia have instead focussed on EEG-signatures closely related to anaesthesia monitoring, with particular emphasis on total power and alpha power 6,38 .…”

“…PSDs were estimated in 244 frequency bins between 0 and 30 Hz, and Hamming windows w ere averaged using a trimmed mean (cut = 25%) to increase robustness to artefacts. To explore the importance of spatio-spectral patterns, we computed the covariance between all 4 electrodes in 5 frequency bands adapted from reference publications on brain age-prediction 4, 5 .…”

Harmonized-Multinational qEEG Norms (HarMNqEEG)

“…We note that the Riemannian Geometry of Symmetric Positive Definite matrices has been widely used for Brain-Computer Interface applications (Barachant et al, 2013;Yger et al, 2017). Recently, in a seminal series of papers, (Sabbagh et al, 2020;Engemann et al, 2021) showed the superiority of HR DPs to predict brain age with MEEG/EEG. Our work differs from these previous applications in two essential aspects.…”