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: | Pierre-Yves Jonin |
Institution: | Inria |
Department: | Empenn |
Country: | |
Proposed Analysis: | Our main goal is to test whether the variability of the hippocampal shape differentially impacts various automatic segmentation algorithms. To this end, we would like to take advantage of the HarP database. Segmentation masks derived from manual tracer experts are available, with corresponding subjects'IDs from ADNI. This will allow us to estimate individual shapes, therefore resulting in a "ground truth average shape", and to map inter-individual shape variability (Kernel ACP analyses) within a common space. However we also need the original T1w images to run automatic segmentation algorithms and thus be able to compare across methods. The aim here will be to contrast shape variability across segmentation methods, then to identify whether inter-algorithms divergences may be driven by specific shape variability of the hippocampus. We recently had the opportunity to use another database including semi-automatic segmentations as ground truth, and the whole processing pipeline resulting in shape extraction is completed. Similarly, we have developed a variability metric that allows the computation of an average shape as well as the visualization of across-individual variability. However, the segmentations in the database are of poor quality, this is why we now apply for accessing to the ADNI images used in the HarP project. |
Additional Investigators | |
Investigator's Name: | Claire Cury |
Proposed Analysis: | Our main goal is to test whether the variability of the hippocampal shape differentially impacts various automatic segmentation algorithms. To this end, we would like to take advantage of the HarP database. Segmentation masks derived from manual tracer experts are available, with corresponding subjects'IDs from ADNI. This will allow us to estimate individual shapes, therefore resulting in a "ground truth average shape", and to map inter-individual shape variability (Kernel ACP analyses) within a common space. However we also need the original T1w images to run automatic segmentation algorithms and thus be able to compare across methods. The aim here will be to contrast shape variability across segmentation methods, then to identify whether inter-algorithms divergences may be driven by specific shape variability of the hippocampus. We recently had the opportunity to use another database including semi-automatic segmentations as ground truth, and the whole processing pipeline resulting in shape extraction is completed. Similarly, we have developed a variability metric that allows the computation of an average shape as well as the visualization of across-individual variability. However, the segmentations in the database are of poor quality, this is why we now apply for accessing to the ADNI images used in the HarP project. |