2026-03-05 ウースター工科大学 (WPI)
＜関連情報＞
- https://www.wpi.edu/news/predicting-alzheimers-disease-0
- https://www.ibroneuroscience.org/article/S0306-4522(25)01177-7/fulltext
神経解剖学に基づく機械学習による性別と年齢を問わないアルツハイマー病の予測 Neuroanatomical-based machine learning prediction of Alzheimer’s Disease across sex and age
Bhaavin K. Jogeshwar ∙ Senbao Lu ∙ Benjamin C. Nephew
Neuroscience Published:December 14, 2025
DOI:https://doi.org/10.1016/j.neuroscience.2025.12.030
Graphical abstract
Highlights
- FastSurfer enabled rapid volumetric analysis of 815 MRI scans.
- Feature importance rankings guided the interpretation of structural MRI predictors.
- Hippocampus, amygdala, and entorhinal cortex were top-ranked in all subgroup analyses.
- Random Forest classified AD, from MCI, and CN with 92.87% accuracy.
- Sex and age subgroups revealed distinct regional patterns of brain atrophy.
Abstract
Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder characterized by cognitive decline and memory loss. In 2024 it affected approximately 1 in 9 people aged 65 and older in the U.S., 6.9 million individuals. Early detection and accurate AD diagnosis are crucial for improving patient outcomes. Magnetic resonance imaging (MRI) has emerged as a valuable tool for examining brain structure and identifying potential AD biomarkers. This study performs predictive analyses by employing machine learning techniques to identify key brain regions associated with AD using numerical data derived from anatomical MRI scans, going beyond standard statistical methods. Using the Random Forest Algorithm, we achieved 92.87 % accuracy in detecting AD from Mild Cognitive Impairment and Cognitive Normals. Subgroup analyses across nine sex- and age-based cohorts (69–76 years, 77–84 years, and unified 69–84 years) revealed the hippocampus, amygdala, and entorhinal cortex as con– sistent top-rank predictors. These regions showed distinct volume reductions across age and sex groups, reflecting distinct age- and sex-related neuroanatomical patterns. Younger males and females (aged 69–76) exhibited volume decreases in the right hippocampus, suggesting its importance in the early stages of AD. Older males (77–84) showed substantial volume decreases in the left inferior temporal cortex. The left middle temporal cortex showed decreased volume in females, suggesting a potential female-specific influence, while the right entorhinal cortex may have a male-specific impact. These age-specific sex differences could inform clinical research and treatment strategies, aiding in identifying neuroanatomical markers and therapeutic targets for future clinical interventions.
