There are many active research projects accessing and applying shared ADNI data. Use the search above to find specific research focuses on the active ADNI investigations. This information is requested annually as a requirement for data access.
| Principal Investigator | |
| Principal Investigator's Name: | Xenia Kobeleva |
| Institution: | University of Bonn |
| Department: | Neurology |
| Country: | |
| Proposed Analysis: | Alzheimer’s disease (AD) is a clinically and pathophysiologically heterogeneous disease. The neuropathological hallmarks of plaques and tangles might be preceded by varying neurodegenerative processes and influenced by different genetic and environmental factors. Thus disentangling the effect of amyloid and tau pathology on brain function can be challenging. One approach to evaluate effect of amyloid and tau pathology is by linking PET scans with MRI data to evaluate the effect of pathology on brain function (as measured by neural dynamics) through computational modeling. Information derived from functional magnetic resonance imaging can be used a proxy for alterations of brain functions and dynamics of these changes might be more sensitive than just investigating averages. Therefore, the aim of our study is to create and compare different types of whole-brain network models of brain dynamics and explore their value in explaining the effect of amyloid-beta and tau on brain dynamics as well as in predicting disease progression. 1. Modelling the effect of amyloid-beta and tau on neural dynamics in two models (dynamic mean-field model and Hopf model) All neuroimaging data will be preprocessed using the standard preprocessing pipeline in the fmriprep software and visually and quantitatively checked for quality. The MRI images will be divided into 100-400 brain regions and then information about fiber connections between the brain regions from diffusion-weighted images (empirical anatomical connectivity) and correlations of neural activity between the regions from resting-state scans (empirical functional connectivity) will be extracted. Using the information on anatomical constraints (given by empirical anatomical connectivity) functional dynamics between brain regions can be mathematically modelled using two different types of models, the dynamic mean-field (DMF) model and the Hopf model. Whereas in the Hopf model the coupling between brain regions is based on the Hopf bifurcation expressing the transition from noisy dynamics to oscillations 1, the DMF model takes into account excitatory currents as mediated by NMDA and AMPA receptors and inhibitory influences as produced by GABA-A receptors 2. To create a realistic fMRI signal, a simulated fMRI BOLD curve is modelled upon the simulated neural activation. By introducing a global coupling parameter, the model with the simulated functional connectivity will be adapted to the empirical functional connectivity for each group. The model allows one to manipulate the parameters (i.e. brain synchrony) to examine the effects on the whole brain network dynamics. Information of amyloid and tau deposition in the regions and their effect on brain dynamics will be added into the model. This will give us an opportunity to compare different proposed pathophysiological mechanisms of amyloid and tau on neural function and explore to what extent amyloid and tau contribute the observed changes in brain dynamics. 2. Modelling the effect of tau and amyloid beta in a model of effective connectivity Another way to explore brain dynamics is to model the directional relationship between brain areas using effective connectivity. Using a multivariate Ornstein-Uhlenbeck (MOU) dynamic model, it is possible to display these relationships on a whole-brain level 3. In comparison with average functional connectivity, effective connectivity entails more information about the underlying brain dynamics and might therefore constitute a more sensitive biomarker of brain activity changes preceding symptom onset. This model neglects underlying pathophysiological processes but tries to be as close as possible to the observed changes in functional connectivity. Therefore it might be more sensitive than other biomarkers based on functional magnetic resonance imaging such as average functional connectivity. |
| Additional Investigators | |
| Investigator's Name: | Gustavo Deco |
| Proposed Analysis: | see Xenia Kobeleva's proposal |
| Investigator's Name: | Matthieu Gilson |
| Proposed Analysis: | see Xenia Kobeleva's proposal |
| Investigator's Name: | Gustavo Patow |
| Proposed Analysis: | see Xenia Kobeleva's proposal |
| Investigator's Name: | Thomas Schultz |
| Proposed Analysis: | see Xenia Kobeleva's proposal |
| Investigator's Name: | Riccardo Leone |
| Proposed Analysis: | see main summary |
| Investigator's Name: | Steven Geysen |
| Proposed Analysis: | see main proposal |
