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: | Kaidong Wu |
| Institution: | Eindhoven University of Technology |
| Department: | Mathematics and Computer Science |
| Country: | |
| Proposed Analysis: | I am a master-level student major in computer science. In general, I intend to carry out a research of machine learning on neuroimaging data as well as auxiliary data like clinical data, genetic data and specimen data (some may not covered in this dataset). To be specific, the research would be around a multimodal machine learning framework that fuses different sources of data for predicting Alzheimer's disease or potentially other brain disorders. For example, in the scope of MRI techniques, we have functional and structural MRI in general, in practical scenarios, both sources of MRI information can help clinical decision (and apparently with other clinical records like gender, age etc.). However most machine learning models only use a single modal, mostly functional MRI, for predicting Alzheimer's disease. Admittedly these models declare that they have achieved very promising performance, like accuracy of 99.8%, they are very likely to be overfitted with the dataset. Besides, the tasks are mostly binary classification, meaning they only consider two classes, like health control vs MCI, health control vs AZ, in fact, some researches pointed out that if these classes are merged for training a multi-class model, the performance would be salient reduced, typically having an accuracy around 70%. Therefore, this research would try to fully utilize multimodal data from different sources. |
| Additional Investigators | |
| Investigator's Name: | Yulong Pei |
| Proposed Analysis: | The supervisor of this project, detailed can be seen in Kaidong's proposal. |
