Clinical Report: Utilizing Preoperative Multimodal CT for Thrombectomy Candidates
Overview
This study evaluates preoperative multimodal CT models to predict outcomes in patients with acute anterior circulation occlusive stroke undergoing mechanical thrombectomy. The Clinical-non-Perfusion (C-NP) model demonstrated the highest predictive accuracy for favorable clinical outcomes post-recanalization.
Background
Acute ischemic stroke, particularly due to large vessel occlusion, is a leading cause of mortality and disability. Mechanical thrombectomy is the most effective treatment, yet many patients experience unfavorable outcomes despite successful recanalization. Identifying candidates likely to benefit from this intervention is crucial for optimizing treatment strategies and improving patient prognoses.
58.78% of patients achieved favorable clinical outcomes (mRS 0–3).
Significant predictors included age, preoperative blood glucose, NIHSS score, ASPECTS, collateral score, infarct core volume, and hypoperfusion volume.
The C-NP model had the highest AUC of 0.861, indicating strong predictive capability.
5-fold cross-validation yielded an average AUC of 0.828 for the optimal model.
Timely assessment of preoperative imaging can guide clinical decision-making and improve patient outcomes.
Clinical Implications
Healthcare providers can utilize the C-NP model to quickly assess the likelihood of favorable outcomes after mechanical thrombectomy. This model aids in making informed decisions regarding patient management and surgical options, ultimately enhancing patient care in acute stroke settings.
Conclusion
The study underscores the importance of preoperative multimodal CT in predicting outcomes for patients undergoing mechanical thrombectomy. The C-NP model offers a practical tool for clinicians to evaluate prognosis effectively.
