Predicting conversion from MCI to AD using resting-state fMRI, graph theoretical approach and SVM.

Area of Science:

• Neuroimaging
• Machine Learning
• Neurology

Background:

• Investigated the progression of mild cognitive impairment (MCI) to Alzheimer's disease (AD).
• Differentiated between MCI converters (MCI-C) and MCI non-converters (MCI-NC) using resting-state functional MRI (rs-fMRI).

Purpose of the Study:

• To develop a method for predicting AD progression in MCI patients.
• To utilize rs-fMRI data for early diagnosis of Alzheimer's disease.

Main Methods:

• Employed graph theory and machine learning, specifically a support vector machine (SVM).
• Utilized a novel feature selection algorithm to identify optimal features from local and global graph measures derived from rs-fMRI.
• Included 18 MCI-C and 62 MCI-NC participants.

Main Results:

• Achieved high classification accuracy (91.4%), sensitivity (83.24%), specificity (90.1%), and AUC (0.95) in distinguishing MCI-C from MCI-NC.
• Identified specific brain regions affected in Alzheimer's disease progression.

Conclusions:

• The proposed approach shows potential for early AD diagnosis.
• rs-fMRI data can predict MCI to AD conversion by revealing underlying affected brain regions.