Early MCI-to-AD Conversion Prediction Using Future Value Forecasting of Multimodal Features.

Area of Science:

• Neuroscience
• Medical Imaging
• Biostatistics

Background:

• Early identification of Alzheimer's disease (AD) in individuals with mild cognitive impairment (MCI) is crucial for timely intervention.
• Longitudinal tracking of biomarkers is essential for understanding disease progression.

Purpose of the Study:

• To develop a decision support technique for identifying individuals with MCI who will progress to AD.
• To predict future disease marker values using longitudinal data.

Main Methods:

• Utilized multivariate MRI-derived biomarkers and neuropsychological measures.
• Designed a weighted gradient offset-based technique for forecasting future marker values based on at least two previous visits.
• Employed a support vector machine classifier with complete predictor trajectories as features.
• Evaluated unimodal and multimodal models using 5-fold cross-validation on the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort.

Main Results:

• Achieved high classification AUC of 91.2% for 6-month-ahead and 95.7% for 1-year-ahead AD prediction using multimodal markers.
• Demonstrated the efficacy of the proposed technique in identifying MCI-to-AD progressors.

Conclusions:

• The developed technique shows high accuracy in predicting MCI-to-AD conversion.
• Multimodal MRI and neuropsychological markers are effective predictors for future AD.
• Further analysis compared the efficacy of MRI markers versus neuropsychological measures (NM) for MCI-to-AD conversion prediction.