Frontiers in Neurology (Mar 2023)
The signs of computer tomography combined with artificial intelligence can indicate the correlation between status of consciousness and primary brainstem hemorrhage of patients
Abstract
BackgroundFor patients of primary brainstem hemorrhage (PBH), it is crucial to find a method that can quickly and accurately predict the correlation between status of consciousness and PBH.ObjectiveTo analyze the value of computer tomography (CT) signs in combination with artificial intelligence (AI) technique in predicting the correlation between status of consciousness and PBH.MethodsA total of 120 patients with PBH were enrolled from August 2011 to March 2021 according to the criteria. Patients were divided into three groups [consciousness, minimally conscious state (MCS) and coma] based on the status of consciousness. Then, first, Mann–Whitney U test and Spearman rank correlation test were used on the factors: gender, age, stages of intracerebral hemorrhage, CT signs with AI or radiology physicians, hemorrhage involving the midbrain or ventricular system. We collected hemorrhage volumes and mean CT values with AI. Second, those significant factors were screened out by the Mann–Whitney U test and those highly or moderately correlated by Spearman’s rank correlation test, and a further ordinal multinomial logistic regression analysis was performed to find independent predictors of the status of consciousness. At last, receiver operating characteristic (ROC) curves were drawn to calculate the hemorrhage volume for predictively assessing the status of consciousness.ResultsPreliminary meaningful variables include hemorrhage involving the midbrain or ventricular system, hemorrhage volume, grade of hematoma shape and density, and CT value from Mann–Whitney U test and Spearman rank correlation test. It is further shown by ordinal multinomial logistic regression analysis that hemorrhage volume and hemorrhage involving the ventricular system are two major predictors of the status of consciousness. It showed from ROC that the hemorrhage volumes of <3.040 mL, 3.040 ~ 6.225 mL and >6.225 mL correspond to consciousness, MCS or coma, respectively. If the hemorrhage volume is the same, hemorrhage involving the ventricular system should be correlated with more severe disorders of consciousness (DOC).ConclusionCT signs combined with AI can predict the correlation between status of consciousness and PBH. Hemorrhage volume and hemorrhage involving the ventricular system are two independent factors, with hemorrhage volume in particular reaching quantitative predictions.
Keywords
