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: | Kai Wang |
| Institution: | Wuhan University |
| Department: | Computer Science |
| Country: | |
| Proposed Analysis: | Brain Functional Connectivity-based Early Diagnosis of Alzheimer Disease: Although Alzheimer s Disease (AD) is irreversible, detecting AD as early (e.g., at its preclinical stage, early Mild Cognitive Impairment (eMCI)) as possible is of great clinical importance for exerting possible interventions to delay its progression. As the associations between the topological organisation of the Brain Functional Connectivity Graph (BFCG) and brain disorder disease have been well established, many recent AD diagnosis studies focus on investigating the early changes of the BFCGs based on graph machine/deep learning. However, the small inter-group variation and the large intra-group discordance of BFCG data, e.g., the differences between BFCGs of normal aging group and eMCI group is subtle yet of different eMCI patients can be sizeable, make it challenging for existing work to distinguish early stages of AD. We propose a novel graph metric learning network, named Cross Attention Metric Network (CAMN), to tackle the early AD diagnosis problem. For a pair of feature maps extracted from the input graphs, CAMN generates an attention mask for each of them to highlight the comparative and critical local features based on a novel cross attention module. A metric learning module is then introduced to obtain the matching scores between the input graphs based on a feature alignment module and a similarity calculating module. In this way, CAMN can reveal the subtle difference between highly similar BFCNs for learning an effective distance metric, while being robust to content misalignment cased by the large intra-group discordance problem. |
| Additional Investigators |
