Scholarlink

Alzheimer’s Disease (AD) progression is heterogeneous; it presents significant prognostic challenges. This work introduces a longitudinal radiomics framework using T1w MRI to identify distinct AD progression profiles. Linear Mixed-Effects Models were employed to robustly estimate trajectories of radiomic features in longitudinal image series and effectively managing irregular follow-up intervals. When clustering these trajectories, we found three progression profiles. Statistical analysis using Chi-squared and Kruskal-Wallis tests confirmed significant differences between the clusters (p < 0.05). Our results suggest that the first cluster identified may be related to a higher progression rate (34.72%), while the second cluster represented a stable disease course (33.27%). The last group showed an intermediate progression rate (31.99%). This approach shows promise for enhancing AD prognosis patient stratification.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Longitudinal Radiomic Analysis Using Mixed-Effects Models for Clustering Alzheimer’s Disease Progression Profiles

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Erick Eduardo López-Ríos,
  • Francisco J. Alvarez-Padilla

摘要

Alzheimer’s Disease (AD) progression is heterogeneous; it presents significant prognostic challenges. This work introduces a longitudinal radiomics framework using T1w MRI to identify distinct AD progression profiles. Linear Mixed-Effects Models were employed to robustly estimate trajectories of radiomic features in longitudinal image series and effectively managing irregular follow-up intervals. When clustering these trajectories, we found three progression profiles. Statistical analysis using Chi-squared and Kruskal-Wallis tests confirmed significant differences between the clusters (p < 0.05). Our results suggest that the first cluster identified may be related to a higher progression rate (34.72%), while the second cluster represented a stable disease course (33.27%). The last group showed an intermediate progression rate (31.99%). This approach shows promise for enhancing AD prognosis patient stratification.