Permutation Entropy for the Characterisation of Brain Activity Recorded with Magnetoencephalograms in Healthy Ageing

Abstract

Abstract:The characterisation of healthy ageing of the brain could help create a fingerprint of normal ageing that might assist in the early diagnosis of neurodegenerative conditions. This study examined changes in resting state magnetoencephalogram (MEG) permutation entropy due to age and gender in a sample of 220 healthy participants (98 males and 122 females, ages ranging between 7 and 84). Entropy was quantified using normalised permutation entropy and modified permutation entropy, with an embedding dimension of 5 and a lag of 1 as the input parameters for both algorithms. Effects of age were observed over the five regions of the brain, i.e., anterior, central, posterior, and left and right lateral, with the anterior and central regions containing the highest permutation entropy. Statistically significant differences due to age were observed in the different brain regions for both genders, with the evolutions described using the fitting of polynomial regressions. Nevertheless, no significant differences between the genders were observed across all ages. These results suggest that the evolution of entropy in the background brain activity, quantified with permutation entropy algorithms, might be considered an alternative illustration of a 'nominal' physiological rhythm.

Complexity Changes in Brain Activity in Healthy Ageing: A Permutation Lempel-Ziv Complexity Study of Magnetoencephalograms

“…Nevertheless, similar to EEG, MEG analysis at the sensor level is affected by volume conduction, although to a much smaller degree [ 12 ]. MEG data have been used to: identify epileptic zones [ 13 ], investigate brain function, evaluate neurofeedback [ 14 ], determine the effects of pathology, investigate effect of cognitive decline [ 15 ], and, to determine the effects of ageing on the brain [ 16 ].…”

“…Once the brain recordings are obtained, a wide range of linear and non-linear and processing techniques, such as causality [ 17 ], entropy [ 16 ], synchronisation [ 18 ], correlation [ 19 ], and complexity [ 4 ], can be applied to them. Complexity is a concept stemming from non-linear analysis methods that could be applied to evaluate changes in brain activity recorded in MEG signals (see [ 20 ] for a detailed description of complexity).…”

Multichannel Time–Frequency Complexity Measures for the Analysis of Age-Related Changes in Neuromagnetic Resting-State Activity

“…However, our results do not depend on the temporal scales, but they reflect the changes in all the frequency bandwidth and are consistent across spatial scales. Increased complexity of MEG signals with age were also reported in [8] and [9]. Evidences of aging-related changes in the Fourier spectrum of EEG with a decrease in power for low frequencies and an increase for high frequencies were also reported in [38], [39], [40].…”

“…Fernández et al found significant increases in Lempel-Ziv complexity measures with age [8]. Shumbayawonda et al confirmed increases in permutation entropy with age [9], and in permutation Lempel-Ziv complexity with age [10]. By contrast, Coquelet et al did not find any significant agerelated changes in both static and dynamic power envelopebased connectivity from resting-state MEG data, suggesting that the electrophysiological connectome is maintained in healthy aging [4].…”

Sex Differences in the Complexity of Healthy Older Adults’ Magnetoencephalograms

“…Moreover, our correlation analysis showed that higher complexity values over highly overlapping regions to the ones mentioned above were associated with larger hippocampal volumes only in females. Different techniques were applied over the years to acquire more information related to the differences between males and females on the activity of the brain [39,[41][42][43][44]. Detection of these sex differences is not a trivial task as male and female activity in MEG signals is very similar, especially for healthy subjects.…”