Peer-reviewed and in-progress research, shown as a living lineage: each paper branches into the extensions and follow-up work it makes possible. Follow a node to read the full write-up.
Multimodal Vision-Language Model for Alzheimer's Disease Classification and VQA
A two-stage, missing-modality-aware framework over T1 MRI, DTI FA maps, and clinical features for Alzheimer's disease classification, extended with retrieval-augmented VQA for case-based clinical reasoning.
2,363
ADNI subjects
0.933
CN vs Dem Bal. Acc.
~70M
Parameters