European Journal of Biomedical and Pharmaceutical Sciences

MRI VERSUS CT-BASED TRAUMATIC BRAIN INJURY CLASSIFICATION AND ITS ASSOCIATION WITH CLINICAL OUTCOMES: A RETROSPECTIVE COHORT STUDY

*Ibrahim Ghaleb Hassan Okour, Dema Fahed Abu Joudeh, Audai Ahmad Shyyab,
Khaldoun Ahmad Al Hadidi, Saja Mohammad Kofahi

ABSTRACT

Background: Traumatic brain injury (TBI) remains a major cause of morbidity and mortality worldwide. Computed tomography (CT) is the primary imaging modality for initial TBI assessment and is commonly classified using the Marshall scoring system. Magnetic resonance imaging (MRI), however, has greater sensitivity for detecting hemorrhagic and diffuse axonal injuries that may not be fully characterized on CT. While both imaging modalities are frequently performed in moderate-to-severe TBI, limited regional data exist comparing their classification systems in relation to meaningful clinical outcomes. Objective: To evaluate the association between CT-based Marshall classification and MRI-detected hemorrhagic lesions with major clinical outcomes in patients with moderate-to-severe TBI. Methods: This retrospective cohort study was conducted at King Hussein Medical Center, Royal Medical Services, Jordan. Adult patients with moderate-to-severe TBI who underwent both CT and MRI during the same hospital admission were included. Demographic, clinical, and radiological data were collected from electronic medical records and radiology reports. CT findings were classified according to the Marshall classification system, while MRI findings were assessed for the presence of hemorrhagic lesions. The main outcomes were ICU admission, neurosurgical intervention, and in-hospital mortality. Statistical analysis included chi-square or Fisher’s exact tests, Mann–Whitney U tests, and multivariable logistic regression. Result: A total of 700 patients were included. MRI-detected hemorrhagic lesions were significantly associated with severe TBI, higher Marshall classification, ICU admission, neurosurgical intervention, in-hospital mortality, and longer hospital stay. In-hospital mortality occurred in 103 patients. Non-survivors were older, had lower admission Glasgow Coma Scale scores, higher rates of hypotension, more advanced Marshall classifications, and more frequent MRI hemorrhagic lesions. In multivariable analysis, older age, male sex, lower admission Glasgow Coma Scale, hypotension, and Marshall Class V–VI were independent predictors of in-hospital mortality. Marshall Class VI independently predicted neurosurgical intervention. However, MRI-detected hemorrhagic lesions were not independently associated with mortality, ICU admission, or neurosurgical intervention after adjustment. Conclusion: MRI-detected hemorrhagic lesions were associated with greater TBI severity and worse clinical outcomes, but they did not independently predict outcomes after adjustment for clinical and CT based variables. CT based Marshall classification remained a reliable prognostic tool for early outcome prediction in moderate to severe TBI, while MRI may serve as a complementary modality for detecting additional traumatic brain injury burden.

Keywords: Traumatic brain injury; Computed tomography; Magnetic resonance imaging; Marshall score; Clinical outcomes.