Relating structural and functional anomalous connectivity in the aging brain via neural mass modeling
Introduction
The brain functions as a complex and highly organized spatiotemporal dynamical system, whose activity is supported by a massive and exquisitely structured network of neurons. The organization of this network extends over multiple spatial scales ranging from the neuronal level to the whole brain. Furthermore, the brain evolves dynamically at temporal scales stretching from milliseconds to years. Growing efforts are being devoted to understanding this astounding level of spatiotemporal complexity. For instance, the functional coordination between cortical areas and large populations of neurons has been assessed in many studies by using noninvasive techniques such as EEG (or its magnetic counterpart, MEG) and functional MRI (Achard et al., 2006, Eguíluz et al., 2005, Fox and Raichle, 2007, Friston et al., 1996, Supekar et al., 2009, Valencia et al., 2008). In contrast, very few studies have approached the relationship between the anatomical and the functional connectivities of the brain (Sporns et al., 2000) in health and disease.
Some information about the structural connectivity of the brain can be obtained in vivo by means of magnetic-resonance-based techniques such as diffusion tensor imaging (DTI) and diffusion weighted imaging (DWI). These methods measure the characteristics of water diffusion through brain tissue (Damoiseaux et al., 2009), and allow us to estimate the strength of the connections between gray matter areas (in the case of DWI) and the location of white matter tracts (in the case of DTI). Studies comparing structural and functional connectivity at an empirical level are beginning to appear (Damoiseaux and Greicius, 2009, Greicius et al., 2009), and the use of mathematical modeling is enabling the prediction of the latter given the former (Honey et al., 2007, Honey et al., 2009). Indeed, the possibility of modeling brain tissue with a specific pattern of structural coupling allows us to determine which specific anatomical connectivity features are responsible for a given correlation at the functional level. This process has the potential of helping us understand the emergence of certain neurological disorders that are characterized by anomalous functional connectivity patterns (Schnitzler and Gross, 2005, Uhlhaas and Singer, 2006). In this article, we perform a systematic investigation of how the different types of structural connectivity defined here affect the functional connectivity profiles observed as the brain ages, either normally or pathologically in the preclinical stages of Alzheimer's disease.
Mild cognitive impairment (MCI) is a condition characterized by a relatively mild, but clinically significant, memory loss that frequently predates the appearance of more severe neurodegenerative disorders such as Alzheimer's disease (AD) (Petersen et al., 1999). The neuropathological characteristics of AD (amyloid plaques, neurofibrillary tangles, synaptic loss, and neural death) are already present at incipient levels in the MCI brain (Guillozet et al., 2003, Markesbery et al., 2006, Price and Morris, 1999). These structural features cannot be identified in vivo to date, and thus a large body of work has been devoted to an alternative approach based on determining changes in functional correlation as an AD biomarker (Celesia et al., 1993, Jeong, 2004, Rossini et al., 2006, Wada et al., 1998.
A first group of studies has addressed changes in the oscillation amplitude of neuronal populations, which can be interpreted as a measure of collective neural behavior at a regional level. EEG studies have shown, for instance that the spectral power of the delta (Moretti et al., 2004) and gamma (van Deursen et al., 2008) frequency bands increase in AD patients with respect to healthy controls. MEG studies have also reported significant power differences in most frequency bands between AD patients and controls (de Haan et al., 2008). Spectral changes, and specially a slowing of the alpha rhythm have been observed in both AD and MCI patients by MEG (Fernández et al., 2006) and EEG (Cantero et al., 2009a). From an inter-regional approach, variations in functional cortico-cortical connectivity in AD/MCI patients versus healthy subjects have been reported in MEG (Gómez et al., 2009, Kurimoto et al., 2008, Stam et al., 2006) and EEG (Kramer et al., 2008). Similar effects were observed in an animal model, in which a selective loss of basal forebrain cholinergic neurons (Grothe et al., 2010), was seen to induce an increase of the phase coupling in the low frequencies (Villa et al., 2000). This suggests that impairment in cholinergic function drastically affects functional coordination in the cerebral cortex. More recently, abnormal patterns of thalamocortical synchronization and complexity have been reported in MCI patients (Cantero et al., 2009a, Cantero et al., 2009b), suggesting that different neural mechanisms underlie the slowing of alpha rhythm present in normal aging and incipient neurodegeneration. Finally, changes in the topological properties of functional brain networks have also been observed in AD patients (Supekar et al., 2008).
Functional studies demand a theoretical framework in which the anatomical changes known to occur in normal and pathological aging can be related to the functional variations observed experimentally. In what follows, we attempt to build such a framework on the basis of a mesoscopic description of the complete brain stemming from a neural mass model. Neural mass models reproduce the average behavior of large populations of neurons by representing their global transfer properties from spike trains to average postsynaptic potentials and vice versa (David and Friston, 2003, Deco et al., 2008). Using this approach and generalizing it to describe the dynamics of the whole brain, we study in a systematic way the influence of different types of structural connectivity (long-range cortico-cortical connectivity, long-range thalamocortical connectivity, and short-range connectivity) on the neural phase synchronization in selected brain areas, as measured by the dominant frequency of signals obtained at different electrodes and the phase synchronization between electrode pairs. Our results allow us to establish a potential scenario of structural connectivity changes that leads to a pattern of functional connectivity variations in agreement with what is observed experimentally as the brain ages normally and pathologically.
Section snippets
Large-scale neural mass modeling
Our main objective was to describe how brain functional connectivity patterns measured by multichannel EEG relate to the structural connectivity properties of the underlying neuronal tissue. Building a model that represents the dynamics at scales ranging from the microscopic neuronal level to the macroscopic brain is currently unfeasible. Thus, we consider an intermediate mesoscopic spatial scale above the size of a cortical column but small enough to allow for a local description of the brain (
Results and discussion
The large-scale neural mass model described above can be used to establish the influence on the brain's functional connectivity of the different types of structural coupling among cortical areas (long-range, short-range, thalamocortical connections). A comparison with experimental results is expected to shed light on the structural changes occurring in the aging brain, and particularly in the presence of subtle connectivity failures in preclinical stages of AD. We begin by presenting the
Conclusions
A common trait of neurological disorders such as mild cognitive impairment and Alzheimer's disease is given by a degradation of the neuronal tissue, that causes among other effects failures in coupling between brain areas at different spatial scales. These structural connectivity changes cannot be easily identified in vivo, and therefore research efforts have concentrated in characterizing the resulting changes in functional connectivity, in order to establish a set of diagnostic tools for the
Acknowledgments
This research has been performed in the framework of the GABA project (Global Approach to Brain Activity, EC contract 043309). Financial support has also been provided by the Ministerio de Ciencia e Innovación of Spain (programa Juan de la Cierva–A.J.P.–and projects SAF 2008-3300, CTS-4604–J.L.C.–and FIS2006-11452–J.G.O.–). We thank Roberto Sotero for kindly providing us with the long-range-connectivity and lead field matrices, and the corresponding voxel distribution per area.
References (53)
- et al.
White-matter vascular lesions correlate with alpha EEG sources in mild cognitive impairment
Neuropsychologia
(2008) - et al.
Increased synchronization and decreased neural complexity underlie thalamocortical oscillatory dynamics in mild cognitive impairment
Neuroimage
(2009) - et al.
Human alpha oscillations in wakefulness, drowsiness period, and rem sleep: different electroencephalographic phenomena within the alpha band
Neurophysiol. Clin.
(2002) - et al.
An electrophysiological study of visual processing in Alzheimer's disease
Electroencephalogr. Clin. Neurophysiol
(1993) - et al.
A neural mass model for MEG/EEG: coupling and neuronal dynamics
Neuroimage
(2003) - et al.
MEG spectral profile in Alzheimer's disease and mild cognitive impairment
Clin. Neurophysiol.
(2006) - et al.
Synchronized oscillations at alpha and theta frequencies in the lateral geniculate nucleus
Neuron
(2004) EEG dynamics in patients with Alzheimer's disease
Clin. Neurophysiol.
(2004)- et al.
EEG functional connectivity and apoe genotype in Alzheimer's disease and controls
Clin. Neurophysiol.
(2008) - et al.
Event-related synchronization of alpha activity in early Alzheimer's disease and mild cognitive impairment: an meg study combining beamformer and group comparison
Neurosci. Lett.
(2008)
Neural mechanisms underlying brain waves: from neural membranes to networks
Electroencephalogr. Clin. Neurophysiol.
The cortical source of the alpha rhythm
Neurosci. Lett.
Organization of thalamic and cortical alpha rhythms: spectra and coherences
Electroencephalogr. Clin. Neurophysiol.
A probabilistic atlas of the human brain: theory and rationale for its development
Neuroimage
Individual analysis of EEG frequency and band power in mild Alzheimer's disease
Clin. Neurophysiol.
Nonlinear multivariate analysis of neurophysiological signals
Prog. Neurobiol.
Conversion from mild cognitive impairment to Alzheimer's disease is predicted by sources and coherence of brain electroencephalography rhythms
Neuroscience
Magnetoencephalographic evaluation of resting-state functional connectivity in Alzheimer's disease
Neuroimage
Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophysiology
Neuron
Corticofugal modulation of functional connectivity within the auditory thalamus of rat, guinea pig and cat revealed by cooling deactivation
J. Neurosci. Methods
Non-linear cortico-cortical interactions modulated by cholinergic afferences from the rat basal forebrain
Biosystems
A resilient, low-frequency, small-world human brain functional network with highly connected association cortical hubs
J. Neurosci.
Functional integrity of thalamocortical circuits differentiates normal aging from mild cognitive impairment
Hum. Brain Mapp.
Greater than the sum of its parts: a review of studies combining structural connectivity and resting-state functional connectivity
Brain Struct. Funct.
White matter tract integrity in aging and Alzheimer's disease
Hum. Brain Mapp.
Resting-state oscillatory brain dynamics in Alzheimer's disease
J. Clin. Neurophysiol.
Cited by (48)
Functional connectivity changes in mild cognitive impairment: A meta-analysis of M/EEG studies
2023, Clinical NeurophysiologyRelationship between regional white matter hyperintensities and alpha oscillations in older adults
2022, Neurobiology of AgingCitation Excerpt :This in turn might explain a positive association between alpha power and WMHs. The hyperactivation of alpha with WMHs could also be ineffective in preserving cognitive performance or even reflect the progression of neurodegenerative alterations (Corriveau-Lecavalier et al., 2019; Pons et al., 2010). Despite a number of reports of age-related alpha power alterations (Babiloni et al., 2006b; Lodder and van Putten, 2011; Vysata et al., 2012), in our study, we replicated other recent studies (Sahoo et al., 2020; Scally et al., 2018) which did not find strong evidence for age-related attenuations of relative alpha power.
When the time is right: Temporal dynamics of brain activity in healthy aging and dementia
2021, Progress in NeurobiologyCitation Excerpt :Conversely, the reported hypersynchrony can also sometimes be ineffective in preserving cognitive performance, or even reflect the progression of neurodegenerative alterations (e.g., de Hemptinne et al., 2015). Pons et al. (2010) observed an increase of resting-state lower alpha (7−9 Hz) fronto-occipital synchrony in healthy older participants relative to younger participants, and in MCI patients relative to healthy controls (see also López et al., 2014b, 2017). Although the association with cognition was not directly assessed, these results importantly suggest that increased long-range connectivity at rest can reflect a failed attempt at compensating modifications of brain networks associated with structural alterations and/or neurodegenerative processes, leading to reduced network efficiency.
Computational models
2018, Molecular-Genetic and Statistical Techniques for Behavioral and Neural ResearchTemporally correlated fluctuations drive epileptiform dynamics
2017, NeuroImageAlzheimer’s Disease: Insights from Large-Scale Brain Dynamics Models
2023, Brain Sciences