Modeling Stroke Recovery with Virtual Brains

What is the primary function of The Virtual Brain (TVB) platform as used in this study?

TVB generates personalized functional neuroimaging data using individual structural connectome data to model large-scale brain dynamics.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What three types of MRI data are used as inputs to construct and fit a personalized model in The Virtual Brain?

The inputs are structural MRI (to create the custom brain surface), diffusion-weighted MRI (to infer anatomical connections), and functional MRI (to modify model parameters).

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Explain the purpose and method of the virtual brain transplant process in this study.

It addresses mechanical deformations in stroke brains by replacing the lesioned cortical region with homologous, healthy tissue from the contralesional hemisphere of the same subject to enable semiautomatized parcellation.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How is the structural connectivity measure capacities calculated in this study?

Capacities represent the maximum rate of information transmission and are calculated using the number of streamlines at the minimum cross-sectional area of an edge.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How is distance defined and calculated in the generated structural connectivity matrices?

Distance is defined as the length of each edge and is calculated by averaging the lengths of all streamlines within that edge.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Why was the Stefanescu-Jirsa 3D (SJ3D) model chosen as the local brain region model in this study?

The SJ3D mean field model does not rely heavily on synaptic delays, making it compatible with the low temporal resolution of BOLD signals.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What biological interaction does the local parameter K21 represent in the SJ3D model?

K21 represents the coupling influence of inhibitory neural populations over excitatory neural populations within a local brain region.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What change in the local parameter K21 was observed in stroke patients, and what does it biologically indicate?

Stroke patients showed a significant decrease in K21, indicating a reduction in local inhibitory control (local disinhibition) within brain regions.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How does the finding of decreased K21 in stroke patients align with cellular studies of stroke in rodents?

It matches animal findings of local disinhibition, specifically a decrease in GABA receptor expression and an increase in NMDA receptor expression in tissue surrounding the lesion.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What change in conduction velocity was observed in stroke patients compared to healthy controls?

Stroke patients exhibited a consistent decrease in mean conduction velocity (46 m/s) compared to healthy controls (62 m/s).

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What change in global coupling was observed in the stroke group compared to healthy controls?

Stroke patients showed a statistically significant increase in global coupling compared to healthy controls.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How did the direction of global parameter changes in stroke patients compare to controls across individuals?

The direction of parameter changes was highly consistent, presenting as a decrease in conduction velocity and an increase in global coupling in all affected patients.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What overall system imbalance does the combination of increased global coupling and decreased local inhibition (K21) suggest in stroke?

It represents an imbalance between global and local brain dynamics, where local, disinhibited dynamics exert a disproportionately stronger influence over global brain coordination.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Which local parameter from the pre-therapy models predicted long-term motor recovery (Fugl-Meyer scores)?

The K12 parameter, which represents the coupling of excitatory over inhibitory populations, showed a significant negative relationship with motor recovery post-therapy and at 1-year maintenance.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What clinical motor measure at 1-year maintenance was positively correlated with global coupling?

The Functional Ability Scale of the Wolf Motor Function Test (WMFT) showed a significant positive relationship with global coupling.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Did stroke characteristics (such as size and location) or patient demographics correlate with the modeled TVB parameters?

No, multiple linear regression showed no significant relationship between TVB parameters and lesion size, location, side of stroke, age, or sex.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Why is the lack of correlation between TVB parameters and baseline anatomical stroke severity/demographics significant?

It highlights the unique predictive potential of modeled biophysical dynamic parameters over traditional anatomical, physical, or demographic metrics.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Why did the authors match the duration of the simulated BOLD signals (4 minutes) to the empirical data?

It was done to minimize the influence of temporal variability over the course of the time series, as functional connectivity metrics are known to be nonstationary over time.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What is the proposed biological basis for the decreased conduction velocity observed in stroke patients?

It is likely caused by structural damage to axons and oligodendrocytes, incomplete remyelination during recovery, and increased nodal/paranodal lengths.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How did the authors constrain the boundaries of the biophysical parameter space during model fitting?

By using heat maps of global variance (mean variance across brain regions) to identify parameter ranges with high global variance flanked by bifurcation points.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What was the primary difference in range between empirical and simulated BOLD signals?

The empirical signals exhibited a much wider range of amplitude values than the narrower range produced by the simulated BOLD signals.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How was the phase similarity between simulated and empirical BOLD signals evaluated in this study?

By calculating the pairwise linear correlation coefficient between the simulated functional connectivity matrix and the empirical functional connectivity matrix.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What is a key conceptual difference between Dynamic Causal Modeling (DCM) and The Virtual Brain (TVB)?

DCM fits parameters over small sub-networks but does not generate whole-brain simulated data, whereas TVB uses biophysical models to actively generate simulated whole-brain time-series data.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What paradigm shift in clinical biomarkers does this paper propose for nonstationary imaging data?

It proposes shifting from complex, nonstationary imaging metrics to the deterministic set of coefficients in a data-constrained mechanistic mathematical model.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How was the spatial parcellation template for the human brain defined in this study?

A 96-region macaque template was mapped to the human MNI template via landmarks and functional activation patterns considered homologous between species.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What mathematical integration method was used to simulate the stochastic network dynamics in TVB?

The stochastic Heun's method with an integration step size of 0.0122 ms was used.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What statistical test was used to compare parameter values between healthy controls and stroke cases?

The Wilcoxon rank sum test corrected for multiple comparisons via Bonferroni correction.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Which local parameters of the SJ3D model were explored, and why?

K12, K21, and K11 were explored because they exert the strongest impact on the excitatory and inhibitory dynamics of the Stefanescu-Jirsa 3D model.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What did the qualitative parameter space explorations reveal about the topology of variance in controls versus stroke subjects?

Controls showed a highly consistent, stable topology and distribution of variance, whereas stroke cases displayed fragmented, scattered, and highly variable patterns.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What is a major limitation of how local parameters were assigned in this study's modeling?

Local parameters were assigned uniformly to all brain regions, failing to capture regional variations or specific abnormalities located around the stroke lesion.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

How were the simulated neural field potentials converted into BOLD signals?

The field potentials were filtered using a hemodynamic response function implemented with a gamma kernel.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

Why does poor stroke recovery correlate with a higher local K12 (excitatory on inhibitory) parameter?

Increased local excitatory influence drives a pathological shift toward local over global dynamics, hindering the whole-brain functional network integration needed for recovery.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What did the analysis of structural connectivity weights show when comparing the stroke group to the control group?

There were no statistically significant differences in mean structural connectivity weights, distribution shape, or skewness between the stroke and control groups.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What preprocessing step was used to correct for motion in the diffusion tensor imaging (DTI) dataset?

Motion correction was performed using the FSL eddy current correction tool.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified

What does the SJ3D model represent at the mesoscopic level, and what is its mathematical lineage?

It represents population dynamics of bursting neurons and evolved from the Hindmarsh-Rose model, which is rooted in Hodgkin-Huxley neuron equations.

Source: Functional Mechanisms of Recovery after Chronic Stroke: Modeling with the Virtual Brain · DOI · quote verified