Abstract
Background: Studies suggest that the detection of traumatic axonal injuries (TAI) using magnetic resonance imaging (MRI) in traumatic brain injury (TBI) may benefit outcome prediction. However, the association between existing TAI grading systems and outcomes following TBI remains uncertain. Therefore, we documented the presence of TAI in several brain regions, using different MRI sequences, and examined their association to patient outcomes. Furthermore, we created a novel MRI-based TAI grading system, with the goal of improving outcome prediction in TBI. We subsequently assessed the performance of several TAI grading systems with regard to outcome prediction.Methods: This was a retrospective study including TBI patients admitted to the intensive care unit between 2005 and 2019. MRI was performed at a median of 7 days post-trauma. We used a genetic algorithm to identify TAI that distinguish favourable from unfavourable outcomes, laying the groundwork for the novel Stockholm MRI grading system. We assessed the discriminatory performance (AUC) and goodness-of-fit (Nagelkerke’s pseudo-R2) of the Stockholm MRI grading system and the TAI grading systems of Adams et al, Firsching et al. and Abu Hamdeh et al., using both univariate and multivariate logistic regression. The dichotomised Glasgow outcome scale was considered the primary outcome.Results: We assessed the MRI scans of 351 TBI patients. TAI that were found to be associated with unfavourable outcomes include TAI in the midbrain tegmentum and TAI in the posterior limb of the internal capsule. The Stockholm MRI grading system exhibited the highest AUC (0.72 vs. 0.68–0.69) and Nagelkerke’s pseudo-R2 (0.21 vs. 0.14–0.15) values of all the studied TAI grading systems. These differences in model performance were, however, not statistically significant (De long’s test p>0.05). Furthermore, all included TAI grading systems improved outcome prediction relative to established outcome predictors of TBI, such as age and computed tomography-based scoring systems (likelihood-ratio test, p < 0.001).Conclusions: Our findings suggest that the detection of TAI using MRI is a valuable addition to prognostication in TBI patients. Additionally, we present the Stockholm MRI grading system, which could potentially improve outcome prediction in TBI. Although, an external validation is required to confirm these results.
Perfusional Deficit and the Dynamics of Cerebral Edemas in Experimental Traumatic Brain Injury Using Perfusion and Diffusion-Weighted Magnetic Resonance Imaging