Hippocampal MRI Radiomics and Multimodal Modeling to Predict Biomarker Status Across the Alzheimer’s Disease Continuum
Jose Maria Benítez-Salazar1, Ivan Murrieta-Alvarez2, Alicia Garmendia-Rebolledo3, Juan Carlos Pérez-García1
1Universidad Popular Autonoma del Estado de Puebla, 2Baylor College of Medicine, 3Unidad Médica de Alta Especialidad. Hospital de Especialidades de Puebla. Centro Médico Nacional General de División “Manuel Ávila Camacho”
Objective:
To identify hippocampal radiomic patterns from structural MRI predicting Alzheimer’s biomarker status and compare the performance of radiomic, clinical, APOE4, and combined models.
Background:
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder marked by cognitive decline and abnormal accumulation of beta-amyloid (Aβ) and tau proteins. Radiomics enables automated extraction of quantitative imaging features, capturing subtle textural patterns beyond visual assessment. This approach shows growing potential to improve diagnosis and prognosis in AD.
Design/Methods:
Data were obtained from ADNI, including clinical data, cerebrospinal fluid (CSF), PET biomarkers, and T1-weighted MRI scans. Images were preprocessed, hippocampal regions automatically segmented, and radiomic features extracted. Group differences between biomarker-positive and -negative were analyzed with Benjamini–Hochberg correction. Key features were selected using the minimum Redundancy Maximum Relevance (mRMR). Models were built with LASSO regression and Random Forest, and feature importance was evaluated with DALEX for interpretability.
Results:
The study included 238 participants: 26 cognitively normal controls, 178 with mild cognitive impairment (MCI), and 34 with AD. SmallDependenceLowGrayLevelEmphasis and Skewness were the most important radiomic features across biomarkers. Combined models integrating radiomics, clinical data, and APOE4 achieved the best performance, with AUCs of 0.870 for Aβ42 and 0.844 for amyloid-PET. Removing APOE4 reduced the Aβ42 model’s AUC from 0.813 to 0.753. APOE4 had minimal impact on tau-PET or FDG-PET, where ADAS13 remained the strongest clinical predictor (AUC ≈ 0.77). Radiomic-only models also performed well for Aβ42 and tau-PET (AUC > 0.74). DALEX-based analysis identified SmallDependenceLowGrayLevelEmphasis, APOE4, and ADAS13 as the variables with the highest contribution to model performance, with mean dropout AUCs of 0.25, 0.22, and 0.30.
Conclusions:
Hippocampal MRI radiomics is a promising noninvasive tool to estimate Alzheimer’s biomarker status. Combining radiomic and clinical data enhances diagnostic accuracy, supporting its use for early detection where advanced imaging and genetic testing are limited.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.