A simplified MRI-based nomogram for risk stratification of cognitive impairment in patients at high vascular risk

To develop a visual nomogram integrating white matter hyperintensity (WMH) and brain atrophy from routine MRI for risk stratification of cognitive impairment in high vascular-risk patients. This retrospective study included 143 patients (93 impaired, 50 normal). Cognitive impairment was defined by MMSE (< 27). WMH was assessed via Fazekas scale on T2-FLAIR, and brain atrophy via Global Cortical Atrophy scale on T1-weighted images. A multivariable logistic model was built and evaluated by AUC, calibration (Hosmer-Lemeshow), 5-fold cross-validation, and decision curve analysis. Brain WMH (aOR = 1.35, 95% CI 1.01–1.79, P = 0.043) and atrophy (aOR = 1.62, 95% CI 1.07–2.43, P = 0.022) were independent predictors. The combined model achieved an AUC of 0.722 (95% CI 0.638–0.806), with a cross-validated testing AUC of 0.699 (95% CI 0.611–0.786). Calibration was excellent ( P = 0.104). Decision curve analysis demonstrated positive net benefit. This practical, visually-based nomogram provides a readily applicable tool for translating routine MRI findings into individualized risk assessment at the bedside, facilitating early clinical decision-making. Not applicable.