Musicians’ brains at rest: Multilayer network analysis of MEG data

Kanad Mandke,Prejaas Tewarie,Peyman Adjamian,Martin Schürmann,Jil Meier

DOI: https://doi.org/10.1101/2024.01.02.573886

2024-01-02

Abstract:The ability to proficiently play a musical instrument requires a fine-grained synchronisation between several sensorimotor and cognitive brain regions. Previous studies have demonstrated that the brain undergoes functional changes with intensive musical training, which has also been identified in resting-state data. These studies have investigated changes in brain networks using fMRI or analysing electrophysiological frequency-specific networks in isolation (i.e., neural oscillatory networks). While the analysis of such “mono-layer” networks has proven useful, it fails to capture the complexities of multiple interacting networks. To this end, we applied a multilayer network framework for analysing publicly available data (Open MEG Archive) obtained with magnetoencephalography (MEG). We investigated resting-state differences between participants with musical training (n=31) and those without (n=31). While single-layer Network-Based Statistics analysis did not demonstrate any group differences, our multilayer analysis revealed that musicians show a modular organisation that spans visuomotor and frontotemporal areas, which are known to be involved in the execution of a musical performance. This twofold modular structure was found to be significantly different from non-musicians. The differences between the two groups are primarily seen in the theta (6.5-8Hz), alpha1 (8.5-10Hz) and beta1 (12.5-16Hz) frequency bands. No statistically significant relationships were found between self-reported measures of musical training and network properties, which could be attributed to the heterogeneity of the dataset. However, we demonstrate that the novel analysis method provides additional information that single-layer analysis methods cannot. Overall, the multilayer network method provides a unique opportunity to explore the pan-spectral nature of oscillatory networks, with studies of brain plasticity in musicians as a potential future application of the novel method.

Neuroscience

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to explore the influence of music training on the brain's resting - state network structure through multi - layer network analysis methods. Specifically, the researchers hope to identify whether the brains of musicians show different neural characteristics or behavior - specific neural markers from non - musicians at rest after long - term music training. In addition, the study also pays special attention to the influence of music training on the changes in the functional connections of the brain's motor areas, and these changes have been reported in task - related data.

The main purpose of the study is to use multi - layer network analysis methods to integrate data from multiple frequency bands (theta, alpha1, alpha2, beta1, beta2) and construct a more comprehensive brain network model to capture the interactions between different frequency bands, thus providing more information than single - frequency - band analysis. Through this method, the researchers hope to reveal how music training affects the modular structure of the brain, especially whether there are differences in the connection strength between sensory, motor and cognitive regions. The study hypothesizes that compared with non - musicians, the connections between the sensory, motor and cognitive regions of musicians will be closer, resulting in a lower degree of brain modularity in musicians.