AI Science Tracker
EN

Novel multi-task learning for Alzheimer's stage classification using hippocampal MRI segmentation, feature fusion, and nomogram modeling.

Affiliation not stated in the reviewed material · 29 Sep 2025 · medicine

Discovery

Novel multi-task learning for Alzheimer's stage classification using hippocampal MRI segmentation, feature fusion, and nomogram modeling.

Source check: Source checked. Not independent scientific confirmation.

Finding: A retrospective multi-center study of 2,956 people classified four Alzheimer disease stages from hippocampal MRI and fused radiomic, deep and clinical features. MedT produced the best reported external hippocampal segmentation Dice score (92.03%); a fused-feature XGBoost classifier reached external accuracy 92.8% and AUC 0.942. An end-to-end 3D CNN-LSTM had AUC 0.840. These dataset results do not establish prospective clinical benefit.

Technology: Hippocampal segmentation compared U-Net, nnU-Net, Swin-UNet and MedT; LSTM attention learned image features, and five classifiers including XGBoost classified stage. SHAP, a nomogram and decision-curve analysis were also reported.

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Deep learning segmented the hippocampus and produced image features; fused radiomic, learned and clinical features fed stage classifiers, with XGBoost performing best in the reported external evaluation.
Test and result
Retrospective multi-center MRI data from 2,956 patients across four stages; external segmentation Dice was 92.03% for MedT, and external classification accuracy/AUC were 92.8%/0.942 for the fused-feature XGBoost pipeline. The end-to-end 3D CNN-LSTM AUC was 0.840.
Limitation / still unknown
The study is retrospective and evaluates classification against the included stage labels; external cohort performance does not establish prospective diagnostic utility or improved care.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://pmc.ncbi.nlm.nih.gov/articles/PMC12482563/ · https://europepmc.org/article/MED/41024255

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 22 Nov 2026. Review scope: Primary result-bearing material for the stated method, result and evaluation, with bibliographic version information..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse medicine discoveries.

Related catalog entries

Links use documented relationships where available, plus shared catalog context. A shared topic or model family does not establish a scientific dependency.