Objective  To build a prediction model based on the characteristics of computed tomography angiography (CTA) and evaluate the value of this model in predicting hematoma enlargement risk in patients with cerebral hemorrhage.

  Methods  A total of 88 patients with cerebral hemorrhage were selected as research objects, and all of them were conducted with conventional treatment for cerebral hemorrhage. Demographic data, clinical characteristics and laboratory examination data of patients were collected at hospital admission. A total of 88 patients with intracerebral hemorrhage were divided into expanded group and non-expanded group according to the hematoma enlargement during hospitalization. Cox regression equation was used to analyze the predictors of hematoma enlargement and help to build a prediction model in patients with intracerebral hemorrhage, and prediction efficiency of this model was analyzed as well.

  Results  Among the 88 patients with intracerebral hemorrhage, 29 cases with hematoma enlargement were selected as enlargement group, and 59 cases without hematoma enlargement were selected as non-enlargement group. Univariate analysis showed that there were significant differences in bleeding volume under the first-time CT, hematoma density, hematoma shape, prothrombin time (PT), the National Institutes of Health Stroke Scale (NIHSS) and score of CTA spot sign between the two groups (P < 0.05). Cox regression analysis showed that bleeding volume under the first-time CT (HR=1.384), hematoma density (HR=3.405), PT (HR=1.510) and score of CTA spot sign (HR=4.883) were the predictors of hematoma enlargement in patients with intracerebral hemorrhage. The consistency index of nomogram model in predicting hematoma enlargement of patients with intracerebral hemorrhage was 0.850, and the discriminative power was good. The verification set was internally verified by Bootstrap self-sampling method, and the correction curve showed that the absolute error of the prediction possibility of the nomogram model was 0.042, which revealed the consistency was good.

  Conclusion  Bleeding volume under the first-time CT, hematoma density, PT and score of CTA spot sign are the independent predictors for hematoma enlargement in patients with intracerebral hemorrhage, and the prediction model based on this conclusion can provide guidance for clinicians to identify high-risk patients with hematoma enlargement of intracerebral hemorrhage.