Hippocampal shape analysis of Alzheimer disease based on machine learning methods

Area of Science:

• Neuroimaging
• Machine Learning
• Computational Anatomy

Background:

• Alzheimer disease (AD) is a progressive neurodegenerative disorder.
• The hippocampus is particularly vulnerable in early AD stages.
• Identifying early AD-related changes is crucial for timely intervention.

Purpose of the Study:

• To evaluate AD-associated regional changes in the hippocampus using machine learning.
• To identify specific hippocampal subregions affected by AD.
• To assess the efficacy of machine learning in detecting these changes.

Main Methods:

• High-resolution MRI scans from 19 AD patients and 20 healthy controls.
• Computational anatomic mapping to characterize regional hippocampal changes.
• Support vector machine with feature selection and cross-validation for identifying shape differences.

Main Results:

• Significant hippocampal deformations observed in the CA1 region (bilateral) and left subiculum in AD patients.
• Additional changes noted in the left CA2-4 subregions of AD patients.
• Left hippocampus showed greater surface variations than the right in AD patients; classification accuracy exceeded 80%.

Conclusions:

• Machine learning methods can detect subtle, complex hippocampal deformation patterns in AD.
• These findings demonstrate the potential of machine learning for early AD detection via neuroimaging analysis.
• Spatially complex deformation patterns are characteristic of AD in the hippocampus.