Hippocampal atrophy based Alzheimer's disease diagnosis via machine learning methods

Area of Science:

• Neurology
• Medical Imaging
• Machine Learning

Background:

• Alzheimer's disease (AD) is the leading cause of dementia, posing a growing public health challenge due to an aging population.
• Early diagnosis and treatment development are critical for managing AD.
• Neuroimaging analysis offers potential for early AD detection.

Purpose of the Study:

• To investigate the potential of hippocampal volume reduction as an early indicator for Alzheimer's disease diagnosis.
• To develop a machine learning model for distinguishing between Alzheimer's Disease (AD), Mild Cognitive Impairment (MCI), and Cognitive Normal (CN) using neuroimaging data.

Main Methods:

• T1-weighted magnetic resonance images (MRIs) from 159 AD patients, 217 MCI patients, and 109 cognitively healthy individuals (CN) were analyzed.
• Hippocampal volumes were calculated using semi-automatic segmentation software (ITK-SNAP).
• Machine learning techniques were applied to diagnose AD, MCI, and CN based on age, gender, and hippocampal volume data.

Main Results:

• Volumetric reduction in the hippocampus was identified as a significant indicator for Alzheimer's disease.
• The developed machine learning approach successfully differentiated between AD, MCI, and CN groups.
• This method demonstrated improved performance in computer-aided diagnosis compared to studies distinguishing only AD from CN.

Conclusions:

• Brain MRI analysis, particularly focusing on hippocampal volume, is a viable method for the early and accurate diagnosis of Alzheimer's disease and Mild Cognitive Impairment.
• The proposed machine learning model enhances the capability of computer-aided diagnosis systems for neurodegenerative diseases.
• Further research into hippocampal volume changes can significantly advance Alzheimer's disease diagnostics.