Statistical Method for Identification of Alzheimer Disease With Multimodal Predictive Markers Mild Cognitive...

Area of Science:

• Neuroscience and Neurology
• Medical Imaging
• Biomarker Discovery

Background:

• Predicting progression from mild cognitive impairment (MCI) to Alzheimer's disease (AD) is critical for timely intervention.
• Accurate diagnostic markers are needed to improve clinical decision-making in neurodegenerative disease.
• Distinguishing stable MCI (sMCI) from progressive MCI (pMCI) is a key challenge in AD research.

Purpose of the Study:

• To explore multimodal predictive markers for differentiating sMCI from pMCI.
• To assess the discriminative power of various biomarkers, including neuroimaging and cognitive data.
• To enhance early AD diagnosis and personalized treatment strategies.

Main Methods:

• Analysis of data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort (487 sMCI, 348 pMCI).
• Inclusion of multimodal data: demographics, positron emission tomography (PET), genotyping, magnetic resonance imaging (MRI), and neurocognitive tests.
• Application of rigorous data preprocessing, feature selection, and statistical analysis using Area Under the Curve (AUC) and Wilcoxon test.

Main Results:

• Florbetaben (FBB) PET imaging demonstrated strong predictive potential with an AUC of 0.84.
• Neurocognitive tests, including ADAS13, mPACC (mPACCtrailsB, mPACCdigit), LDELTOTAL, and ADASQ4, showed high discriminatory power (AUC 0.82–0.83).
• Combined neuroimaging and cognitive assessments significantly differentiated between sMCI and pMCI.

Conclusions:

• Multimodal assessments, particularly PET imaging and neurocognitive tests, are crucial for distinguishing sMCI from pMCI.
• These findings support the development of early AD diagnostic strategies.
• The results facilitate personalized intervention planning for individuals with MCI.