Abstract Details

Title Machine Learning Tools to Predict Subclinical Small-vessel Brain Lesions Using AHA Life’s Essential 8: Evidence from the ELSA-Brasil Longitudinal Study

Topic Global Health and Neuroepidemiology

Presentation(s) S4 - Hot Topics in Global Health and Neuroepidemiology (1:48 PM-2:00 PM)

Poster/Presentation
Number 005

Objective

To compare ML algorithms for predicting CSVD lesions on 3T MRI from LE8 adherence levels and temporal changes in the ELSA-Brasil study, and to quantify feature contributions via explainable AI.

Background Subclinical cerebral small-vessel disease (CSVD), consisting of enlarged perivascular spaces (EPS), white-matter hyperintensities (WMH), lacunes (LAC), and microhemorrhages (MH) on brain MRI, is linked to cognitive decline, stroke, and mortality. Longitudinal cardiovascular health, as captured by the AHA Life’s Essential 8 (LE8), may be anticipated using combined modern machine learning (ML) methods.

Design/Methods ELSA-Brasil adult participants were followed over a 15-year period (n=233; mean age 66.1±9.2; 59% women). The outcomes were binary EPS, WMH, LAC, and MH. Predictors included categorical LE8 scores (1–3; higher=better), LE8 trajectory from Wave 1 (2008–2010) to Wave 3 (2017–2019), and baseline demographics (age, income, education, race, marital status). Logistic Regression, Random Forest, XGBoost, LightGBM, CatBoost, and TabPFN were trained using a 70/30 train–test split, with repeated 5-fold cross-validation for hyperparameter tuning. Test AUC-ROC quantified performance; SHAP values assessed feature importance.

Results The prevalence of EPS was 64.8%, WMH 28.3%, LAC 11.6%, and MH 12.1%. For EPS, TabPFN performed best (AUC=0.764). Logistic Regression showed moderate discrimination for LAC and MH (AUC=0.749 and 0.739), while CatBoost was modest for WMH (AUC=0.692). SHAP consistently highlighted age, Wave-3 blood lipids, sleep improvement, and Wave-1 blood glucose among top contributors, alongside selected LE8 trajectory features.

Conclusions ML models leveraging LE8-based cardiovascular health meaningfully predict subclinical CSVD in ELSA-Brasil, with the strongest discrimination for EPS and reasonable performance for WMH. TabPFN emerges as a promising transformer-based alternative for classical machine learning algorithms. Larger, richly phenotyped cohorts and external validation are warranted to refine predictive performance and support translation into clinical and public health settings.

Authors/Disclosures
Marianna Leite PRESENTER	Miss Leite has nothing to disclose.
Carine Savalli, PhD	Prof. Savalli has nothing to disclose.
Arão B. Oliveira, PhD	Prof. Oliveira has received research support from FAPESP. Prof. Oliveira has a non-compensated relationship as a Director with ABRACES that is relevant to AAN interests or activities.
Carlos L. Prazeres, Sr., Diagnostic Imaging Specialist	Mr. Prazeres has nothing to disclose.
Paulo Lotufo	Dr. Lotufo has nothing to disclose.
Isabela M. Bensenor, PhD	Dr. Bensenor has nothing to disclose.
Claudia C. Leite, Jr., MD, PhD	The institution of Prof. Leite has received research support from Finep. Prof. Leite has received publishing royalties from a publication relating to health care.
Maria C. Otaduy, PhD	The institution of Prof. Otaduy has received research support from NIH.
Itamar S. Santos, MD, PhD	Prof. Santos has received research support from CNPq.
Adriana B. Conforto, MD, PhD (HC/FMUSP and HIAE)	Dr. Conforto has received personal compensation in the range of $500-$4,999 for serving as a Consultant for Boehringer Ingelheim. Dr. Conforto has received personal compensation in the range of $0-$499 for serving as a Consultant for Eurofarma. Dr. Conforto has received personal compensation in the range of $500-$4,999 for serving on a Speakers Bureau for HCX. The institution of Dr. Conforto has received research support from CNPq. Dr. Conforto has received research support from CNPq. Dr. Conforto has received publishing royalties from a publication relating to health care. Dr. Conforto has received personal compensation in the range of $5,000-$9,999 for serving as a Travel grant recipient with FAPESP.
Alessandra C. Goulart	Prof. Goulart has nothing to disclose.
Alexandre Chiavegatto, PhD	Prof. Chiavegatto has nothing to disclose.

��ɫ��

��ɫ��