Lianlian Zhang, Haiyan Cao, Lizhu Miao, Xinyuan Zhang, Guofu Shi
Carotid duplex ultrasound radiomics significantly enhances the ability to predict adverse outcomes in AIS, offering a precision-based approach that supports personalized management strategies. It could improve clinical outcomes by allowing targeted interventions based on specific radiomic profiles.
BACKGROUND: Acute ischemic stroke (AIS) is a leading cause of mortality and morbidity worldwide. This study explores the potential of using carotid duplex ultrasound (CDU) radiomics to predict outcomes in AIS patients, enhancing prognostic assessments through advanced imaging technologies.
METHODS: We enrolled 105 AIS patients at the Stroke Center of the First Affiliated Hospital of Soochow University. CDU images were obtained and processed using artificial intelligence to extract 1,477 radiomic features. Key features were identified and used to develop a predictive model for patient outcomes 3 months post-stroke.
RESULTS: The developed predictive model demonstrated high accuracy and clinical utility, confirmed by decision curve analysis. Inter-observer reliability was excellent, with a Cohen's kappa coefficient of 0.93, indicating consistent assessments of plaque vulnerability across different observers.
CONCLUSION: Carotid duplex ultrasound radiomics significantly enhances the ability to predict adverse outcomes in AIS, offering a precision-based approach that supports personalized management strategies. It could improve clinical outcomes by allowing targeted interventions based on specific radiomic profiles.