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: | Selen Erkan |
Institution: | LMU |
Department: | Data Science |
Country: | |
Proposed Analysis: | I would like to access Adni data to use it in my Master's thesis. I am doing my Master’s thesis at AI in Medicine lab of TUM (Technische Universität München) and I am working under the supervision of Prof. Daniel Rückert, Dr. Magda Paschali and Vasiliki Sideri-Lampretsa. The aim of this Master's thesis is to combine volumetric imaging and non-imaging longitudinal data to accurately analyze individual patients and provide automated decisions regarding diagnosis and disease prognosis. In addition to the diagnostics, the work aims to aid the explainability of the deep learning models. For that purpose we would like to use imaging and tabular data of Adni as our data source. The goal is to combine these imaging and tabular data to link the information they have. This combined information later on will be used in a deep learning architecture for disease prediction and improving the explainability of the model. The network will learn from the combination of different data sources and decisions will be made based on this combined information. The methods used in the project will also aim to provide better explainability. |
Additional Investigators | |
Investigator's Name: | Vasiliki Sideri-Lampretsa |
Proposed Analysis: | The objective of this project is to combine volumetric imaging and non-imaging data to accurately analyze individual patients and provide automated decisions regarding diagnosis and disease prognosis. The approach will integrate multiple biomarkers (clinical, imaging, genetic and biochemical) of each patient in the same deep learning model in an end-to-end fashion. The method can be fully supervised or be used in a self-supervision scheme, in which our model can learn the aligned embedding space of imaging and non-imaging data using a proxy task from the data itself. Afterward, the learned embeddings can be finetuned for downstream tasks, e.g. detection of Alzheimer's disease at the earliest possible stage. |
Investigator's Name: | Magdalini Paschali |
Proposed Analysis: | The objective of this project is to combine volumetric imaging and non-imaging data to accurately analyze individual patients and provide automated decisions regarding diagnosis and disease prognosis. The approach will integrate multiple biomarkers (clinical, imaging, genetic and biochemical) of each patient in the same deep learning model in an end-to-end fashion. The method can be fully supervised or be used in a self-supervision scheme, in which our model can learn the aligned embedding space of imaging and non-imaging data using a proxy task from the data itself. Afterward, the learned embeddings can be finetuned for downstream tasks, e.g. detection of Alzheimer's disease at the earliest possible stage. |