Related Experiment Video
Updated: Apr 5, 2026
Brain Infarct Segmentation and Registration on MRI or CT for Lesion-symptom Mapping
Published on: September 25, 2019
MR image features predicting hemorrhagic transformation in acute cerebral infarction: a multimodal study
Chunming Liu1, Zhengchao Dong2,3, Liang Xu1
1Department of Radiology, Tianjin Union Medicine Centre, 190 Jieyuan Road, Hongqiao District, Tianjin, 300121, People's Republic of China.
Introduction:
The aims of this study were to observe magnetic resonance imaging (MRI) features and the frequency of hemorrhagic transformation (HT) in patients with acute cerebral infarction and to identify the risk factors of HT.
Methods:
We first performed multimodal MRI (anatomical, diffusion weighted, and susceptibility weighted) scans on 87 patients with acute cerebral infarction within 24 hours after symptom onset and documented the image findings. We then performed follow-up examinations 3 days to 2 weeks after the onset or whenever the conditions of the patients worsened within 3 days. We utilized univariate statistics to identify the correlations between HT and image features and used multivariate logistical regression to correct for confounding factors to determine relevant independent image features of HT.
Results:
HT was observed in 17 out of total 87 patients (19.5 %). The infarct size (p = 0.021), cerebral microbleeds (CMBs) (p = 0.004), relative apparent diffusion (rADC) (p = 0.023), and venous anomalies (p = 0.000) were significantly related with HT in the univariate statistics. Multivariate analysis demonstrated that CMBs (odd ratio (OR) = 0.082; 95 % confidence interval (CI) = 0.011-0.597; p = 0.014), rADC (OR = 0.000; 95 % CI = 0.000-0.692; p = 0.041), and venous anomalies (OR = 0.066; 95 % CI = 0.011-0.403; p = 0.003) were independent risk factors for HT.
Conclusions:
The frequency of HT is 19.5 % in this study. CMBs, rADC, and venous anomalies are independent risk factors for HT of acute cerebral infarction.
