The outcome prediction using non-contrast computed tomography scan and computed tomography angiography scan in acute hemorrhagic stroke

Background Intracerebral haemorrhage (ICH) outcome depends on hematoma volume, location, and expansion. Objectives To assess the validity of neuroimaging signs for predicting the prognosis of patients with acute ICH in our population. Patients and Methods A prospective cohort study was performed on 90 patients with acute ICH admitted to Shar Hospital from March to October 2019. Inclusion criteria were ages of ≥18 years and spontaneous ICH, and exclusion criteria were trauma, brain tumour, and secondary ICH. Demographic features were recorded. Blackhole, swirl, island, Blend and spot signs, ICH location and volume, and ICH score were assessed by non-contrast computed tomography (CT) scan and CT angiography. Glasgow coma scale (GCS) and modified Rankin scale were used to assessing patients’ outcomes Results Except for the ages of patients (p-values=0.01), other demographic characteristics had no significant associations with the expansion of hematoma and outcome. Modified Rankin Scale, GCS, and hematoma location and volume had statistically significant associations with hematoma expansion and outcome. Further, strong sensitivity of black hole (90.9%) and spot (92.8%) signs, strong specificity of Blend (92.6%) and spot signs (97.1%), substantial positive predictive value for spot sign (92.8%), substantial negative predictive value was for all signs. In addition, substantial accuracy of spot sign (95.8%), were found. Also, significant associations for all the signs, except Blend, with hematoma expansion were found. Conclusion It is better to use neuroimaging signs, at least the signs found on non-contrast CT scans, all together in clinical practice.