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: | Katrine Andersen |
| Institution: | Institute of Clinical Medicine |
| Department: | Department of Nuclear Medicine and PET, Aarhus Uni |
| Country: | |
| Proposed Analysis: | We kindly request to access the available neuroimaging data (MRI, amyloid, tau, FDG) for early-MCI, late-MCI, AD participants, and elderly controls in the ADNI dataset. In addition, we request access to clinical data from these participants. Intended use of data/objective: We have recently proposed the synuclein, origin, and connectome (SOC) model as a theoretical disease model to explain the evolution of pathology in Parkinson’s disease. In short, we propose that the first protein pathology starts in a single place and spreads from there through the connectome. In body-first Lewy body disease, the pathology starts in the enteric nervous system, and spreads via overlapping vagal innervation. This leads to more symmetric brainstem involvement and hence more symmetric nigrostriatal degeneration. In brain-first Lewy body disease, the pathology originates in the olfactory bulb and/or amygdala of one hemisphere. Since the human brain is dominated by ipsilateral connections (outnumbering contralateral connections 100:1), unilateral pathology in brain-first disease will spread mainly to the ipsilateral hemisphere first – leading to asymmetric nigrostriatal degeneration. These predictions are supported by imaging studies showing much more asymmetric dopamine loss in clinical brain-first cases compared to body-first cases [Walker 2004, Knudsen 2021, Cao 2020]. This mechanism is based on the fact, that in the human brain, only 1 % of axons are commissural and for each anatomical region, these axons project mainly to the anatomical homolog of the contralateral hemisphere. We are interested in applying this explanatory framework to the field of AD research. We propose that the protein pathologies in AD also originate in a spatially very restricted location in the individual patient, but that the origin site can differ among patients, I.e. a oribitofrontal site in some, occipital site in other patients, and a parietal/cingulate site in yet other patients, as suggested by recent studies [Collij 2020]. We hypothesize that the lateralized connectome plays a crucial role in the dissemination of AD protein pathologies. Importantly, the contralateral (commisural) connections in the human brain mainly innervate the contralateral homologue, so called homotypic innervation [Bruce MR paper]. Therefore, if AD protein pathologies often start in a single region (unilaterally), we predict that the propagation to the contralateral hemisphere will be to the contralateral homologue for this region. This prediction leads to the hypothesis that the two hemispheres will often show different amounts of protein pathology. Furthermore, that the spatial pattern of pathology in the contralateral hemisphere generally will be a time-shifted mirror image of the originating hemisphere. In short, the contralateral hemisphere will show a less pronounced but spatially very similar pattern to the ipsilateral hemisphere. Thus, we initially want to test asymmetry and spatial patterns of pathology in FDG, amyloid-, and tau PET imaging data from early-MCI, late-MCI, AD participants including different subtypes of AD, and elderly controls. We plan to utilize the recently developed SuStaIn model to identify consistent subtypes of patients on each of the imaging modalities [Young 2018]. We hypothesize that (1) early stages of AD exhibit more asymmetric pathology on PET scans (2) in each patient, one hemisphere is mirror image of the other hemisphere, but with less pronounced pathology, (3) all subtypes of AD will converge on a final common pattern, with fully symmetric pathology throughout all vulnerable regions. We will also attempt to estimate time constants for the progression rate of propagating protein pathologies and FDG hypometabolism in each identified AD subtype. The overall aim of the study is to test whether the SOC model, developed in a Parkinson’s disease context, represents a generalizable principle in other neurodegenerative diseases such as AD. In short, that these diseases commonly start in a single location (analogous to a cancer primary tumor), followed by stereotypical, connectome-dependent propagation of pathology. Brief synopsis of proposed analyses: PET images processing steps will be performed using PMOD™ 4.0 software (PMOD4.0, PMOD technologies, Zurich, Switzerland) and FreeSurfer. Images will be pre-processed, including smoothhing using Gaussian kernel smoothing with an FWHM (full width at half maximum) for reduction of possible noise at voxel level. Following, motion correction and averaging of the frames will be performed as additional steps in the pre-processing protocol. For PET processing a pipeline in Neuro tool will include the following steps: Segmentation of MRI T1, Normalization of MRI T1-waighted segmented images (to move the brain MRI into standard MNI space), rigid matching of PET to MRI, Regions-of-interest (ROI) will be generated. Additional, we will defined fitting reference regions for intensity normalization and generation of SUVR values, and partial volume correction. Finely, we will extract data in nifty format for further voxel-based analyses in SPM12 and for application to the SuStaIn model algorithm in Python, to uncover and identify consistent subtypes of patients on each of the imaging modalities in the ADNI dataset. In addition, we plan to analyse asymmetric/symmetric AD pathology in each subgroup in PMOD using the SuStaIn results. Asymmetry index will be calculated simply as: (R-L)/((R+L)/2). We hope that you will grant us access to the necessary data to test these ideas. We aim to publish the study in a peer-reviewed journal specialized in AD research. |
| Additional Investigators |
