Clinical and high-resolution magnetic resonance imaging-based prediction of ischemic stroke in cervical artery
Xuanxiao Zhang1, Chunmei Liu2, Shuo Yin1
1Department of Radiology, The First Hospital of Jilin University, Changchun, China.
Insights
Cervical artery dissection (CeAD) stroke risk can be predicted using a new patient-level model. Key predictors include white blood cell count, intraluminal thrombus, male sex, and alcohol consumption for better risk stratification.
Area of Science:
- Neurology
- Vascular Imaging
- Stroke Medicine
Background:
- Cervical artery dissection (CeAD) is a significant cause of ischemic stroke.
- Early risk stratification for CeAD remains a clinical challenge.
Purpose of the Study:
- Identify clinical and high-resolution vessel wall magnetic resonance imaging (HRMRI) features associated with ischemic stroke in CeAD patients.
- Develop a patient-level predictive model for short-term ischemic stroke risk in CeAD.
Main Methods:
- Retrospective analysis of 129 CeAD patients (148 vessels).
- Utilized Least Absolute Shrinkage and Selection Operator (LASSO) regression and mixed-effects logistic regression for variable selection and association analysis.
- Constructed a nomogram integrating vascular imaging features and clinical variables for patient-level risk prediction.
Main Results:
- Vessel-level predictors of ischemic events: elevated white blood cell (WBC) count, intraluminal thrombus, severe stenosis/occlusion, and alcohol consumption.
- Patient-level predictors of ischemic stroke: WBC count, intraluminal thrombus, male sex, and alcohol consumption.
- The developed nomogram demonstrated good discriminative ability (AUC 0.837) and calibration.
Conclusions:
- A patient-level predictive model effectively identifies short-term ischemic stroke risk in CeAD.
- This model can aid in early risk stratification and personalized clinical decision-making for CeAD patients.
Background:
Cervical artery dissection (CeAD) is an important cause of ischemic stroke, yet early risk stratification remains challenging. This study aimed to identify clinical and high-resolution vessel wall magnetic resonance imaging features associated with ischemic stroke and to develop a patient-level model for short-term risk prediction.
Methods:
A total of 129 patients with CeAD (148 dissected vessels) were retrospectively included. Baseline clinical data and HRMRI features were collected. At the vessel level, least absolute shrinkage and selection operator (LASSO) regression was used for variable selection, followed by a mixed-effects logistic regression model to identify factors associated with ischemic events. At the patient level, representative vascular imaging features were integrated with clinical variables. LASSO regression and multivariable logistic regression were applied to construct a nomogram for risk prediction. Model performance was evaluated using receiver operating characteristic curves, calibration curves, and decision curve analysis.
Results:
At the vessel level, white blood cell (WBC) count, intraluminal thrombus, severe stenosis or occlusion, and alcohol consumption were independently associated with ischemic events. At the patient level, multivariable analysis showed that WBC count, intraluminal thrombus, male sex, and alcohol consumption were independent predictors of ischemic stroke. The nomogram exhibited good discriminative ability, with an optimism-corrected area under the curve of 0.837 (95% CI: 0.810-0.852), along with satisfactory calibration and clinical net benefit.
Conclusion:
A patient-level model shows good performance in predicting ischemic stroke risk in patients with CeAD and may assist in early risk stratification and individualized clinical decision-making.
Related Concept Videos
01:21Imaging Studies for Cardiovascular System IV: CMRI
01:25Imaging Studies for Cardiovascular System VI: Calcium -Scoring CT