Computed tomography (CT) scanning of the head is the principal diagnostic tool to demonstrate brain damage in TBI [1]. Since its introduction in the 1970s, it has revolutionized the management of TBI and has doubtlessly saved many lives. The primary use of cranial CT scanning is to detect life-threatening traumatic intracranial abnormalities that require immediate neurosurgical intervention or admission to an intensive care unit for careful monitoring of neurologic status. CT is sensitive and specific in detecting skull fractures, intracranial hemorrhages (subdural hematomas, epidural hematomas, traumatic subarachnoid or intraventricular hemorrhages, and parenchymal contusions) (see Table 2.1, Figure 2.1). CT is also able to detect local or diffuse brain edema, which can be identified as areas of hypodensity or indirectly by findings such as effacement of cortical sulci, disappearance of the normal gray–white matter demarcation, midline shift, or effacement of basal cisterns (see Figure 2.1). Approximately 10% of patients with severe TBI require a craniectomy based on the findings from an initial CT scan. According to the Brain Trauma Foundation guidelines, these findings include extra-axial hematomas larger than 30 mL in size or associated with greater than 5 mm of midline shift and parenchymal hematomas in a noneloquent cortex greater than 20 mL in size [1]. Patients in whom the original CT scan shows small- or moderate-sized parenchymal hematomas, traumatic subarachnoid hemorrhage, or extra-axial hemorrhages (subdural or epidural hematomas) are admitted to the hospital and usually rescanned within 24 h or sooner if there is a deterioration of neurologic status, as clinically significant expansion of intracranial hematomas is common [1].

In mild-to-moderate brain injury, CT is useful in identifying traumatic lesions that may affect clinical management such as small hematomas that may subsequently expand or traumatic subarachnoid or intraventricular hemorrhages that may result in posttraumatic hydrocephalus [2–4, 7]. However, CT findings relate poorly to long-term outcome in mild TBI (mTBI). In mTBI, the abnormalities identified by acute CT are not associated with long-term functional outcome [8]. Approximately 10% of patients who sustain mTBI with no significant abnormalities on the acute CT have significant problems returning to work [9]. An explanation for this inability to predict outcome may be the insensitivity of CT to detect the diffuse microstructural white matter damage that is characteristic of diffuse axonal injury (DAI) or its failure to identify deficits in cerebral perfusion or cerebrovascular reactivity [9].

This chapter will introduce a systematic approach in the reading of CT, to assist the clinician in recognizing parameters that adversely affect outcome, ascertaining optimal treatment, making clinical decisions, and estimating prognosis. In addition, validated scales for rating CT scans like the Trauma Coma Data Bank (TCDB) classification or Rotterdam CT score, which are helpful in clinical research as well as for stratifying participants by injury severity, will be introduced [2–4]. Although positron emission tomography and technetium-99m-hexa-methyl propylene amine oxime single-photon emission computed tomography (SPECT) may sometimes show abnormalities in the acute and chronic stages when CT or magnetic resonance imaging (MRI) and neurological examination do not show damage, we will consider these and other imaging modalities like magnetic resonance spectroscopy beyond the scope of this book.

How to read the CT?

Because immediate and accurate recognition of intracranial emergencies on head CT is vital to initiate proper treatment, a formalized systematic approach to interpretation of head CTs is useful. In 1998, Perron et al. proposed a mnemonic “Blood Can Be Very Bad” [10]. The use of this mnemonic resulted in a significant improvement in the accuracy rate of CT interpretation by emergency residents. B stands for blood to remind examiners to search for intracranial hematomas, C stands for the appearance of the four key cisterns, B stands for brain to check sulcal patterns and evidence of midline shift, V prompts reviewing the four ventricles, and finally, B is a reminder to assess the bones of the cranium (see Figures 2.1–2.4). As a general guide, use of the mnemonic can be useful to systematically approach the CT. Nevertheless, other specific issues must be taken into account in relation to traumatic lesions.

Blood: Lesion volume

A wide variety of threshold values for hematoma volume, ranging from 5 to 150 mL, have been proposed to ascertain the relationship between hematoma size and outcome [11, 13–15]. Internationally acknowledged injury severity models use different cutoff values to classify trauma victims with multiple (systemic) injuries or isolated head injuries (see Figure 2.3). The Abbreviated Injury Scale (AIS) and the Injury Severity Score both use threshold values of 15, 30, and 50 mL to estimate decreasing probabilities of survival after TBI (Table 2.1) [11, 16]. Reevaluation of 13 962 patients in a large European trauma registry between 2001 and 2008 with a GCS less than 15 at presentation or any head injury with AIS severity code 3 and above demonstrated that large compared to small epidural, subdural, and intraparenchymal hematomas are associated with a substantially higher probability of hospital mortality (Table 2.1) [17]. Although the use of threshold values can support clinical decisions, using lesion volume as continuous variable in prognostic models has been recommended recently [15].

Brain: Shift

A linear relationship between midline shift, measured at the level of the septum pellucidum, and outcome exists in patients with severe TBI [5, 18] (see Figure 2.3). Cutoffs ranging from 3.0 to 20.0 mm have been tested for their association with outcome. Prognosis is worse in the presence of shift above various cutoff values (reviewed in reference [15]). The TCDB CT classification uses a shift cutoff value of 5 mm (Table 2.1) [3].

Basal cisterns

One of the most prominent CT characteristics and a powerful predictor of outcome is the status of the basal cisterns. The term perimesencephalic cisterns is frequently used to denote the basal cisterns (see Figure 2.1). However, no clear definition exists to anatomically delineate the basal cisterns [11]. Complete and partially obliterated basal cisterns are a sign of raised intracranial pressure [12] (see Figure 2.2). Basal cisterns’ compression results in a two- to fourfold increase in mortality or unfavorable outcome compared to normal visibility of the basal cisterns [12–14]. Regarding the aspect of the distinct compartments of the basal cisterns, systematic evaluation of the predictive value of the individual cisterns has been performed by our group, demonstrating that absence (complete obliteration) or compression of both the ambient cisterns and the fourth ventricle is independently related to unfavorable outcome and death [20]. The assessment of the ambient cisterns and the fourth ventricle has satisfactory inter- and intrarater reliabilities (kappa coefficients: 0.80–0.95).

Ventricles

The predictive value of compression of the cerebral ventricles (lateral, third, fourth) in TBI has not been as extensively studied as effacement of basal or perimesencephalic cisterns. Obliteration of the third ventricle is associated with raised ICP and adverse outcome after severe TBI [21, 22] and an independent prognostic variable in a multivariate predictive model of TBI [23]. Small or asymmetrical lateral ventricles are probably not related to outcome [13]. Determining the presence or absence of the status of ventricles on CT by visual inspection is susceptible to interobserver variation. Inter- and intrarater variability for the lateral and third ventricles have not been studied. We found a satisfactory inter- and intrarater reliability of the fourth ventricle. The interrater kappa coefficients of the status of the fourth ventricle were 0.93 (normal vs. abnormal) and 0.95 (normal, compressed, or absent) [20]. The intrarater kappa coefficients of the fourth ventricle score (normal vs. abnormal) were 0.89 and for the three-point scale (normal, compressed, or absent) 0.80. Obliteration of the fourth ventricle and foramen magnum indicates a bad prognosis from the presence of infratentorial brain swelling, a cerebellar mass lesion, or an end stage of progressive vertical brain herniation.

Bone

X-skull radiography and echo were the primary diagnostic tools in the detection of intracranial abnormalities before the introduction of the CT in 1972. X-skull is in particular sensitive to detect fractures. Fractures were then noted as indicators of an increased chance for the presence of intracranial hematoma. The diagnostic value of plain skull radiography as an indirect indicator of the presence of hematoma decreased since the introduction of the CT. Compared with X-skull radiography, CT is also excellent at demonstrating depressed or comminuted skull fractures. Only fractures parallel to the plane of the scan slice such as those involving the vertex of the skull are sometimes missed. CT is also sensitive in visualizing facial fractures and skull base fractures. The importance of detecting skull fractures in TBI is related to (i) an increased risk of cerebrospinal fluid (CSF) leakage and CSF fistula formation after a skull base or temporal bone fracture or open fracture [23], (ii) increased odds for developing meningitis in the presence of a skull base fracture, and (iii) a growing skull fracture in children. Of note, to evaluate the presence of a fracture (skull, facial, skull base), a different window setting is used from the settings to highlight soft tissue or blood or brain parenchyma (see Figure 2.4).

Trauma Coma Data Bank classification

In 1991, a CT scan classification of head injury including four categories of diffuse injury and two surgical mass lesion categories was developed by the TCDB study [3]. This CT classification was the first official system to categorize intracranial injury after moderate and severe TBI in a systematic way and to discriminate between diffuse injury and mass lesions. Four categories indicate diffuse injury of increasing severity, whereas two categories indicate the presence of a mass lesion [3]. The TCDB CT classification categorizes TBI patients into six groups. The primary distinction is made between diffuse injury (categories I–IV) and the presence of mass lesions (evacuated or nonevacuated) (Table 2.2).

A strong association exists between outcome and the initial CT scan diagnosis according to the TCDB CT classification. The TCDB classification (also known as the Marshall classification, after its first author) is widely used in research trials and prognostic studies of moderate and severe TBI to characterize and categorize patients.

The TCDB classification is a composite score of certain individual aspects of the CT evaluation: the mesencephalic cisterns, the degree of midline shift in millimeters, and the volume of intracerebral or extracerebral hematoma [3]. Recent prognostic models based on large international patient cohorts confirmed that the type of intracranial lesion and intracranial hematoma size, effacement of cisterns, and shift of the cerebral midline structures are strong outcome predictors [6, 19, 23]. Controversy remains about how best to combine these parameters into a prognostic scheme, and whether to use threshold values or continuous variables.

While the predictive value of the TCDB CT classification has been confirmed in several studies, there are several limitations that have led to efforts to develop new classification schemes. The TCDB classification does not take into account the presence of traumatic subarachnoid hemorrhage, which is found in 23–63% of patients with severe TBI [24]. Proposals for adding subarachnoid hemorrhage in CT classification systems to enhance the predictive power on outcome appeared in the 1990s [24, 25]. Moreover, the TCDB classification does not recognize the difference between subdural and epidural hematomas. This forms a significant limitation given the more favorable prognosis associated with epidural hematomas. Finally, the TCDB classification can be improved by combining categories V and VI that assess the volume of any mass lesion [26].

The Rotterdam CT score

The Rotterdam CT score demonstrated that discrimination can be further improved by adding intraventricular and traumatic subarachnoid hemorrhage and by a more detailed differentiation of mass lesions and basal cisterns and taking into account the presence of epidural hematoma (Table 2.3) [5]. Several prognostic models have now demonstrated superior prediction of the Rotterdam CT score over the TCDB classification (REF).

MRI

MRI uses a combination of static and dynamic magnetic fields in conjunction with radiofrequency pulses to generate a signal from water protons within tissues. For example, T1- or T2-weighted images measure differences in the longitudinal recovery or transverse decay of excited protons, which permits characterization of subtle differences between normal and injured tissues. In addition, numerous more specialized MR techniques, including sequences that are highly sensitive for microhemorrhages (e.g., susceptibility-weighted imaging [SWI]), techniques that may depict the microstructure of the brain and produce a map of the axon cell processes (diffusion tensor imaging [DTI]), and methods to probe brain physiology directly (e.g., functional MRI [fMRI]), have been introduced during the last 20 years enabling the detection of subtle but functionally important consequences of TBI.

Conventional MRI

Conventional MRI sequences include T1- and T2-weighted, T2* gradient-recalled echo (T2*-GRE), fluid-attenuated inversion recovery (FLAIR), and diffusion-weighted imaging (DWI) (Table 2.4).

T1-weighted sequences are susceptible to the presence of blood or fat in brain tissue depicted by high signal intensities. However, T1 offers minimal diagnostic specificity as abnormal attenuation on T1 sequences is unspecific and may indicate multiple etiologies such as hematomas, parenchymal lesions, or vascular tumors. T2*-weighted gradient echo MR imaging detects brain hemorrhage better than T1 images, but high signal intensities from the CSF in these images complicate the identification of DAI/TAI lesions. FLAIR MRI imaging detects cerebral edema and allows for easier identification of injured tissue, as it nullifies CSF signal and greatly increases the contrast between normal and abnormal brain matter [27, 28]. In addition, FLAIR-weighted MRI can be used to quantify the degree of white matter damage after trauma [29, 30, 38].

Conventional MRI is more sensitive than CT for detecting TBI-related lesions. In particular, it is much better than CT in detecting nonhemorrhagic contusions, edema, punctuate hemorrhages, and DAI [31]. In addition, MRI provides higher spatial and contrast resolution than CT and higher sensitivity to detect indirect brain damage from systemic injuries including hypoxic–ischemic injury and fat embolism [28, 31–34]. In a prospective study directly comparing CT with conventional MRI in level I trauma centers, 27% of mTBI patients with normal admission head CT had trauma-related abnormalities on MRIs done approximately 2 weeks after injury [101]. The most common MRI abnormalities were subcortical microhemorrhages, indicative of traumatic axonal injury, noted on T2*-weighted gradient echo sequences.

MRI is primarily used in the subacute and chronic phases of TBI, when patient safety issues related to transport and prolonged time in the imaging suite play a less important role. In moderate and severe TBI, MRI is often done to explain a neurologic status worse than predicted by a normal or near-normal CT scan, raising suspicion of a more serious injury. Such information can be useful prognostically as well as for guiding rehabilitation strategies. In mTBI with persistent postconcussive syndrome and cognitive deficits, MRI can be useful for counseling and forensic indications.

For clinical usefulness, the ease and reliability of traumatic lesion detection are important. Lesion number, volume, and lesion location may all affect clinical status and outcome. Brainstem injury is a predictor of mortality and poor functional outcome [35, 36]. Lesions at the temporal and frontal lobes may explain memory and executive functioning disorders [37]. The number of white matter hyperintensities by FLAIR MRI is significantly associated with long-term outcome after TBI [30, 38]. In 18 patients with severe TBI, the number of hyperintensities seen on FLAIR was positively associated with functional outcome [30]. In 37 patients with severe TBI, FLAIR lesion volumes in the corpus callosum (CC) correlated with disability and cognition in the first months following injury. FLAIR lesion volume in the frontal lobes correlated significantly with clinical scores at 1 year [38].

Related posts:

ICU care: surgical and medical management—systemic treatment Out-of-hospital management in traumatic brain injury Epidemiology of traumatic brain injury Emergency department evaluation of mild traumatic brain injury Follow-up and community integration of mild traumatic brain injury ICU care: surgical and medical management—neurological monitoring and treatment

Stay updated, free articles. Join our Telegram channel

Join