中国神经再生研究(英文版) ›› 2026, Vol. 21 ›› Issue (9): 4241-4242.doi: 10.4103/NRR.NRR-D-25-01816

• 观点:退行性病与再生 • 上一篇    下一篇

功能性磁共振成像与人工智能在阿尔茨海默病的融合应用

  

  • 出版日期:2026-09-15 发布日期:2026-05-12

Integration of functional magnetic resonance imaging and artificial intelligence in Alzheimer’s disease

Liqin Yang*, Yuxin Li, Kuangyu Shi, Axel Rominger, Ruiqing Ni*   

  1. Department of Nuclear Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland (Yang L, Rominger A, Shi K, Ni R)
    Department of Radiology, Huashan Hospital, Fudan University, Shanghai, China (Yang L, Li Y)
    Institute for Regenerative Medicine, University of Zurich, Zurich, Switzerland (Ni R)
    Institute for Biomedical Engineering, ETH Zurich & University of Zurich, Zurich, Switzerland (Ni R)
  • Online:2026-09-15 Published:2026-05-12
  • Contact: Ruiqing Ni, PhD, ruiqing.ni@unibe.ch; Liqin Yang, PhD, liqin.yang@unibe.ch.
  • Supported by:
    This work was supported by China Scholarship Council, No. 202306100073 (to LY) and UniBern Forschungsstiftung, No. 31/2025 (to RN).

摘要: https://orcid.org/0000-0002-0793-2113 (Ruiqing Ni)
https://orcid.org/0000-0003-1823-4756 (Liqin Yang)

Abstract: Alzheimer’s disease (AD), the most common cause of dementia, is a progressive neurodegenerative disease characterized by progressive cognitive decline and memory loss. Mild cognitive impairment (MCI) is the prodromal stage of AD, with a conversion rate of 10%–15% per year and 50% conversion rate longitudinally. The pathological features of AD include the aberrant accumulation of amyloid-β plaques, neurofibrillary tangles formed by hyperphosphorylated tau and synaptic dysfunction (Nussbaumer et al., 2025), all of which have cascading effects on the brain activity of AD patients. Functional magnetic resonance imaging (fMRI), specifically blood oxygenation level–dependent fMRI, is a non-invasive and radiation-free technique for indirectly measuring brain activity. It should be noted that fMRI has a double-edged sword nature. While fMRI is a powerful tool for detecting whole-brain-wide alterations in neural activity, its high sensitivity to brain states and complex analysis methods and indices also result in poor reproducibility across studies and compromise its further application in clinical practice. The results derived from extremely limited unpaired samples (e.g., before and after treatment in the same patients) or a longitudinal study design should be further explored.