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: | Yoon Seong Choi |
| Institution: | Severance Hospital, Yonsei University College of Medicine |
| Department: | Department of Radiology |
| Country: | |
| Proposed Analysis: | The aim of this study is to develop a model to predict to risks of future AD development in patients with MCI, based on conventional MRI and FDG-PET for imaging modality, and convolutional neural network(CNN), hand-crafted radiomics and machine learning for time-to-event data (such as random survival forest) for analysis tools. Previous publications applied CNN for hippocampal TIWI, and cox-based approach. (Li et al. 2019). However, no study applied comprehensive approach that integrates both conventional MRI. and FDG-PET based on CNN and hand-crafted radiomics. In our study, CNN will be applied to analyze signal intensity of the images from conventional MRI and FDG-PET, and hand-crafted radiomic features will be used to analyze region shapes and volume information, which cannot be directly reflected into CNN. These features are to be combined with other clinical risk factors of future AD development by using random survival forest model, which allows more flexible feature selection based on minimal depth, and is independent from the strict assumptions (such as proportional hazard assumption) that Cox regression-based models relies on. This prognostic model for the future progression to AD from MCI patients will be a 'mother' model, which will be used for 'straight' validation, or transfer learning in the local dataset. |
| Additional Investigators |
