Abstract
BACKGROUND: Identifying the etiology of acute ischemic stroke (AIS) is critical for secondary prevention and treatment choice in stroke patients. This study aimed to investigate the dual-energy computed tomography (DECT) quantitative thrombus parameters associated with cardioembolic (CE) stroke and develop a nomogram that combines DECT and clinical data to identify CE stroke. METHODS: We retrospectively reviewed all consecutive patients from January 2020 to July 2022 with anterior circulation stroke and proximal intracranial occlusions. Patients were divided into CE stroke and non-cardioembolic (NCE) stroke groups according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria. Univariable and multivariable logistic analyses were conducted, and a nomogram was developed by combining clinical and DECT variables. This nomogram was subsequently validated using an independent internal cohort of patients. RESULTS: A total of 96 patients were analyzed, of which 43 (45%) were diagnosed with CE stroke. The multivariable analysis identified the following factors as being independently associated with CE stroke: normalized iodine concentration (NIC) (per 10(-2) unit increase) (odds ratio (OR) = 1.598, 95% CI: 1.277-1.998; p < 0.001), gender (OR = 0.113, 95% CI: 0.028-0.446; p = 0.002), hypertension (OR = 0.204, 95% CI: 0.054-0.770; p = 0.019), and baseline National Institutes of Health Stroke Scale (NIHSS) (OR = 1.168, 95% CI: 1.053-1.296; p = 0.003). The matching nomogram displayed an area under the curve (AUC) of 0.929 in the study sample (n = 96) and 0.899 in the validation cohort (n = 29). CONCLUSIONS: A nomogram that combines clinical and DECT variables can display good diagnostic performance for CE stroke.