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: | Ziga Spiclin |
Institution: | University of Ljubljana |
Department: | Faculty of Electrical Engineering |
Country: | |
Proposed Analysis: | Computational analysis of structural MR images using advanced machine learning tools offers an exciting opportunity to extract accurate and robust predictors for early diagnosis and prognosis, for instance, of the effect of ageing and neurodegenerative diseases. The objective of the proposed project is to develop and validate the performance machine learning architectures for three competitive analysis strategies. First strategy is to use convolutional neural networks for semantic segmentation of particular structures of interest and extract from them the radiomics features. Second strategy is to use auto encoder neural networks to compactly represent information in the MR image, which yields implicit, but less intuitive features. Features can then be used for prediction of the outcome measure, for instance, a disease phenotype, stage or any other categorical variable, but also continuous variables such as "brain age", lesion volume and count and similar. Third strategy entails classification/regression of the prediction variables directly from the input image. A large number of datasets is required for a faithful training and validation of machine learning tools. Ultimately, the ADNI datasets will be used among other MR image datasets from multiple sites to determine the best image based prediction strategy. |
Additional Investigators |