Abstract: A total of 30 cases of cerebral hemorrhage patients with enlarged hematoma were selected as hematoma enlargement group, and 30 patients of cerebral hemorrhage without hematoma enlargement were selected as non-hematoma enlargement group. All patients were examined by CT. Logistic regression equation was used to analyze the related influencing factors of hematoma enlargement. The receiver operating characteristic curve (ROC) was used to analyze the efficacy of cerebral hemorrhage signs in predicting hematoma enlargement and softening, as well as hematoma volume and high-low density difference. The results showed that the volume of newly diagnosed hematoma and the difference between high and low density were significantly related to the increase of hematoma (P<0.05). The confounding sign was combined with lobular sign, black hole sign, irregular hematoma, the volume of the newly diagnosed hematoma and the high-low density difference to predict the hematoma enlargement and softening. The AUC values were higher than those predicted by any single parameter. The results of this study can confirm that CT features are closely related to hematoma enlargement and softening, which including the confounding sign, lobular sign, black hole sign, irregular hematoma, newly diagnosed hematoma volume and high-low density difference. The combined detection has high predictive value, which is convenient to guide the clinical treatment plan.

Key words: CT, cerebral hemorrhage, hematoma volume, softening foci, density difference, signs of cerebral hemorrhage