A simple nomogram prediction model to identify relatively young patients with mild cognitive impairment who may progress to Alzheimer's disease

Area of Science:

• Neurology
• Neuroimaging
• Geriatrics

Background:

• Mild cognitive impairment (MCI) is a precursor to Alzheimer's disease (AD).
• Early identification of MCI patients at high risk for AD progression is crucial for timely intervention.
• Predictive models can aid clinicians in stratifying risk within younger MCI populations.

Purpose of the Study:

• To develop a clinical predictive model for identifying MCI patients (≤65 years) likely to progress to AD.
• To integrate easily accessible clinical features with structural neuroimaging data for enhanced prediction.
• To validate the model's performance in a distinct patient cohort.

Main Methods:

• Utilized data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) database.
• Selected MCI patients ≤65 years with 6-60 months follow-up.
• Extracted demographic, neuropsychological, and structural MRI data.
• Performed voxel-based morphometry (VBM) and surface-based morphometry (SBM) analyses.
• Constructed predictive models combining clinical and imaging features.

Main Results:

• The developed nomogram achieved a cross-validated AUC of 0.872 (training set) and 0.867 (verification set).
• Calibration curves demonstrated good model fit.
• An online forecasting tool based on the model is available.

Conclusions:

• The predictive model demonstrates strong performance and ease of use.
• This tool can assist clinicians in screening younger MCI patients at risk of progressing to AD.
• Facilitates earlier detection and potential therapeutic strategies for at-risk individuals.