HDFT-MViT: a progressive core-enhanced mix framework for Alzheimer’s disease classification using MRI images
Dongyan Zhang, Jincan Zhang, Bo Liu, Min Liu, Wenna Chen, Ganqin Du · 2026
Read HDFT-MViT: a progressive core-enhanced mix framework for Alzheimer’s disease classification using MRI images by Dongyan Zhang, Jincan Zhang, Bo Liu, Min Liu, Wenna Chen, Ganqin Du (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/PMC13322790.txt.
Description
Dongyan Zhang and co-authors build a light AI model to classify Alzheimer's disease from brain scans in a 2026 study. Standard networks read local detail well but miss long range links, while transformer models capture those links at heavy computing cost. Their HDFT-MViT model mixes both, using cheap early layers and dynamic filters plus a light transformer deeper in.
On two public ADNI scan sets, accuracy reached 98.85 percent for three class sorting and 98.07 percent for four class sorting, with only 3.46 million parameters in the model. Channel attention helped separate the classes. The balance of accuracy and small size suits clinics with modest computing power.
Catalog record
| CREATOR | Dongyan Zhang, Jincan Zhang, Bo Liu, Min Liu, Wenna Chen, Ganqin Du |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47085 |
| LICENSE | https://creativecommons.org/licenses/by/4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |