| Objective: To investigate the relationship between retinal nerve fiber layer (RNFL) thickness and cerebral blood flow dynamics and intracranial pressure (ICP) in patients with different traumatic brain injury (TBI). Methods: A total of 80 TBI patients admitted to our hospital from January 2021 to January 2023 were selected as the research objects, including 32 females and 48 males. According to the GCS score, all patients were divided into mild TBI group (n=26), moderate TBI group (n=34) and severe TBI group (n=20). All patients were divided into non-elevated group (n=25) and elevated group (n=55) according to whether the ICP was elevated or not. The cerebral hemodynamic parameters of patients with different degrees of TBI were collected and compared. The cerebral hemodynamic parameters of patients with different degrees of TBI were compared. Multiple linear regression was used to analyze the correlation between RNFL and cerebral hemodynamics. The clinical data of patients with different ICP types were analyzed and compared. Logistic regression was used to analyze the association between RNFL and elevated ICP. LOWESS method was used to analyze the curve correlation between RNFL and intracranial pressure of different degrees of TBI. ROC curve was used to analyze the predictive kinetic energy of RNFL and cerebral blood flow parameters for ICP increase. Results: There are significant differences in cerebral blood flow parameters in patients with different degrees of TBI. Multiple linear regression analysis showed that RNFL was independently positively correlated with Vs, Vm and PI, and RNFL was independently negatively correlated with RI. Logistic regression analysis showed that low RNFL was a protective factor for increased ICP, medium RNFL and high RNFL were independent risk factors for increased ICP. The results of LOWESS method analysis showed that there was a nonlinear relationship between RNFL and intracranial pressure in patients with mild TBI, moderate TBI, and severe TBI (P<0.05), and the nonlinear relationship was negative. ROC analysis results showed that both RNFL and cerebral blood flow dynamic indicators had a certain predictive effect on increased ICP, and the combined prediction efficiency of the five indicators was the best, with AUC=0.912. In univariate prediction, RNFL had the best performance, with an AUC of 0.859. Conclusions: In TBI patients, RNFL is independently positively correlated with Vs, Vm, and PI, and RNFL is independently negatively correlated with RI. RNFL thickness and cerebral blood flow dynamic parameters have guiding significance for ICP in TBI patients. |