21 Brain Morphological and Functional Networks …
337
neural connections, which may pave the way to cross-analysis of these networks. Precisely because they cannot be reduced to spatially close regions, functional modules
contain information about non-structural level of neural organisation, which can only
be investigated via analysis of time series of neural function. An additional dimension
to this investigation are temporal fluctuations in functional networks characterized
by striking differences in network organization, in particular nodal affiliations with
different modules. Understanding whether these time-varying behaviours of functional networks reflects a discrete cytoarchitecture organisation, may encourage a
shift from descriptive correlations to predictive mechanisms. If the later is true, each
cytoarchitecture components and switching modules should exhibit similar spatial
organisation.
21.4 Implications for Neurodegeneration
MRI studies assessing correlated changes in anatomical or functional regional properties have argued that neurodegeneration targets those networks that are highly
correlated in healthy individuals [66], leading to a so called “disconnected network
syndrome” hypothesis [12]. Moreover, recent findings on cross-correlated micro- and
macro-architectures, in particular the size of layer 3 neurons (known to be affected
in Alzheimer’s disease) [72], may inform new approaches in studying neurodegenerative syndromes. Similar approaches have been successful in revealing patterns
of distinct involvement of the two cortical features (thickness and surface area) in
Alzheimer’s disease and behavioral variant FrontoTemporal Dementia (bvFTD) [79].
However, more work is needed for these approaches to be validated in clinical settings. I suggest that for the initial application of these methods, in line with [79],
the connectome can be sampled at the resolution of anatomical landmarks to examine macro-scale organisational units of the cortex and the role of each unit in the
neurodegeneration. By formulating the problem of vulnerability to neurodegeneration as a problem of network topology, one can investigate how different regional
morphological features contribute to this vulnerability. When the nature of this vulnerability is clarified, the cross-scale networks could be studied [72] to evaluate
micro-scale connectivity. Finally, the roles in network vulnerability can be validated
against functional and anatomical networks examined across a range of parcellations
schemes, including random parcellation. Thus, the joint properties of functional and
morphological brain networks may offer better estimates of vulnerability to neurodegenerative syndromes. The examination of these networks across multiple temporal
and spatial scales would represent dynamic network mechanisms underlying not-soeasily differentiated clinical states in these syndromes. These dynamic network interactions and their underpinning morphological properties could inform treatments in
these diseases and may mediate treatments outcome.
337
neural connections, which may pave the way to cross-analysis of these networks. Precisely because they cannot be reduced to spatially close regions, functional modules
contain information about non-structural level of neural organisation, which can only
be investigated via analysis of time series of neural function. An additional dimension
to this investigation are temporal fluctuations in functional networks characterized
by striking differences in network organization, in particular nodal affiliations with
different modules. Understanding whether these time-varying behaviours of functional networks reflects a discrete cytoarchitecture organisation, may encourage a
shift from descriptive correlations to predictive mechanisms. If the later is true, each
cytoarchitecture components and switching modules should exhibit similar spatial
organisation.
21.4 Implications for Neurodegeneration
MRI studies assessing correlated changes in anatomical or functional regional properties have argued that neurodegeneration targets those networks that are highly
correlated in healthy individuals [66], leading to a so called “disconnected network
syndrome” hypothesis [12]. Moreover, recent findings on cross-correlated micro- and
macro-architectures, in particular the size of layer 3 neurons (known to be affected
in Alzheimer’s disease) [72], may inform new approaches in studying neurodegenerative syndromes. Similar approaches have been successful in revealing patterns
of distinct involvement of the two cortical features (thickness and surface area) in
Alzheimer’s disease and behavioral variant FrontoTemporal Dementia (bvFTD) [79].
However, more work is needed for these approaches to be validated in clinical settings. I suggest that for the initial application of these methods, in line with [79],
the connectome can be sampled at the resolution of anatomical landmarks to examine macro-scale organisational units of the cortex and the role of each unit in the
neurodegeneration. By formulating the problem of vulnerability to neurodegeneration as a problem of network topology, one can investigate how different regional
morphological features contribute to this vulnerability. When the nature of this vulnerability is clarified, the cross-scale networks could be studied [72] to evaluate
micro-scale connectivity. Finally, the roles in network vulnerability can be validated
against functional and anatomical networks examined across a range of parcellations
schemes, including random parcellation. Thus, the joint properties of functional and
morphological brain networks may offer better estimates of vulnerability to neurodegenerative syndromes. The examination of these networks across multiple temporal
and spatial scales would represent dynamic network mechanisms underlying not-soeasily differentiated clinical states in these syndromes. These dynamic network interactions and their underpinning morphological properties could inform treatments in
these diseases and may mediate treatments outcome.