Eur Radiol (2002) 12:1237–1252 DOI 10.1007/s00330-002-1355-9

N. Bešenski

Published online: 19 April 2002 © Springer-Verlag 2002 Categorical Course ECR 2003

N. Bešenski (✉) Croatian Institute for Brain Research, Šalata 12, 10000 Zagreb, Croatia e-mail: [email protected] Tel.: +385-1-4596922 Fax: +385-1-4596942/945

EMERGENCY RADIOLOGY

Traumatic injuries: imaging of head injuries

Abstract Due to the forces of acceleration, linear translation, as well as rotational and angular acceleration, the brain undergoes deformation and distortion depending on the site of impact of traumatizing force direction, severity of the traumatizing force, and tissue resistance of the brain. Linear translation of accereration in a closed-head injury can run along the shorter diameter of the skull in latero-lateral direction causing mostly extra-axial lesions (subdural hematoma,epidural hematoma, subarachnoidal hemorrhage) or quite pronounced coup and countercoup contusions. Contusions are considerably less frequently present in medial or paramedial centroaxial blows (fronto-occipital or occipito-frontal). The centroaxial blows produce a different pattern of lesions mostly in the deep structures, causing in some cases a special category of the brain injury, the diffuse axonal injury (DAI). The brain stem can also be damaged, but it is damaged more often in patients who have suffered centroaxial traumatic force direction. Computed tomography and MRI are the most common techniques in patients who have suffered brain injury. Computed tomography is currently the first imaging technique to be used after head injury, in those settings where CT is available. Using CT, scalp, bone, extra-axial hematomas, and parenchymal injury can be demonstrated. Computed tomogra-

phy is rapid and easily performed also in monitored patients. It is the most relevant imaging procedure for surgical lesions. Computed tomography is a suitable method to follow the dynamics of lesion development giving an insight into the corresponding pathological development of the brain injury. Magnetic resonance imaging is more sensitive for all posttraumatic lesions except skull fractures and subarachnoidal hemorrhage, but scanning time is longer, and the problem with the monitoring of patients outside the MRI field is present. If CT does not demonstrate pathology as can adequately be explained to account for clinical state, MRI is warranted. Follow-up is best done with MRI as it is more sensitive to parenchymal changes. In routine MR protocol gradient-recalledecho sequences should be included at any other time after a traumatic event since they are very sensitive in detection of hemosiderin as well as former hematoma without hemosiderin. The MR signal intensity varies depending on sequences and time scanning after trauma. Keywords CT · MR · Head injury · Linear translation · Diffuse axonal injury

1238

Introduction Head injury (HI) and its consequences cause a big public health problem. The fact that the frequency of injuries is the highest in the age groups of 20 and 50 years has great sociological and economic importance. The incidence of persons injured in traffic accidents higher than in falls reflects an “epidemic” of trauma as a consequence of increased traffic [1, 2].

Imaging of head injuries Imaging of head trauma in the emergency setting is performed to detect potentially treatable lesions before a secondary neurologic damage occurs. Plain X-ray, CT, and MR are the most common techniques in patients who have suffered brain injury. Although skull fractures can be detected by a different plain X-ray examination, they are obsolete. Scout CT scan and thin slices with bone window are reliable in detecting depressed skull fractures. Twenty-five percent of cases with fatal injuries do not demonstrate a skull fracture, although the incidence of intracranial hematomas in patients who have skull fractures is much higher than in those who do not [3]. Skull fracture do not correspond with the severity of brain injury. Skull fractures are most common in persons injured by falls [4]. Clinical considerations for patients suffering a brain injury are included into and judged by the Glasgow Coma Scale. The indications for CT examination are severely impaired consciousness, focal neurological signs, and penetrating head injury. Clinically, this is a high-risk group. A moderate-risk group are the injured with minor disturbances of consciousness, progressive headaches, fracture of the skull base, and those with multiple injuries. The low-risk group are patients with mild or moderate posttraumatic headaches without a loss of consciousness. The focal neurological signs are sometimes masked by a serious general state of the patient (coma, shock). Generally, the density of traumatic brain lesions demonstrated by CT can be hyperdense, hypodense, or mixed (combined). Pathoanatomically hyperdense lesions on CT correspond to bleeding, whereas hypodense lesions correspond to edema or destruction of axons and necrosis; therefore, hypodense lesions are also important. Around hyperdense lesions hypodense areas can be frequently found depending on the time of scanning in relation to the traumatic event concerned. On the control CT we can follow the evolution of hyperdense lesions into hypodense lesions (resorption of hemorrhage), but also the evolution of hypodense into hyperdense areas [5, 6, 7, 8]; therefore, CT is a suitable method to follow the dynamics of lesion development giving an insight into the

corresponding pathological development of brain injury [6]. Undoubtedly, MR imaging can substantially enlarge the scope of visualization of the lesions, especially at the peracute stages. The availability of MR imaging data obtained in comatose patients after head injury is scarce, because MR imaging is somewhat cumbersome to perform in patients requiring ventilation and because, in the first hours after injury, its relevance is clearly inferior to CT. Because MR is inferior to CT in detecting acute hemorrhages, it is of little practical value in the acute phase. In addition, special equipment is required to obtain MR images in a patient who is respirator dependent; therefore, there is a paucity of MR imaging data obtained in comatose headinjured patients during the acute phase while the patient is dependent on the ventilator. But when MR is performed at any other time after a traumatic event in routine MR protocol, gradient-recalled-echo (GRE) sequences should be included since they are very sensitive in detection of hemosiderin as well as previous hematoma without hemosiderin [9]. The MR signal intensity varies depending on sequences and time scanning after trauma.

Biomechanical considerations of brain injury To interpret the imaging findings in a brain injury properly, the biomechanical and pathological conditions during the traumatic event need to be known. Due to the forces of acceleration, linear translation, as well as rotational and angular acceleration, the brain undergoes deformation and distortion depending on the site of impact of traumatizing force direction, severity of the traumatizing force, and tissue resistance of the brain. At the moment of trauma lacerations of parenchymal, supportive, and vascular brain structures occur in the epicenter of the primary irreversible damage. These epicenters are consequences of the effects produced by the maximal forces combined with a minimal capacity of the resistance of the brain tissue. It is important to keep in mind that not all types of the brain cells are equally vulnerable. The most vulnerable cells are axons and the most resistant structures are blood vessels. These epicentric lesions are surrounded by larger areas of less severely but still irreversibly damaged tissue which is hardly visible by naked-eye inspection but is demonstrated by microscopic examination giving the impression of diffuse damage, known as a diffuse axonal injury (DAI), a special category of brain injury (Fig. 1). Histopathological studies show that the extent of axonal injury always exceeds one that is visualized macroscopically. All imaging modalities underestimate the true extent of DAI. The second type of lesions which occur at the moment of a traumatic event are primarily reversible

1239

Fig. 1 Diffuse axonal injury from the pathological point of view

Fig. 2 Epicentric lesion with traumatic penumbra in diffuse axonal injury (DAI). a Epicentric lesion. b Irreversible damaged tissue. c Penumbra

lesions also related to the epicenters and are distributed at the peripheral epicentric zones. They are termed traumatic penumbra such as an ischemic penumbra. This change represents a locus minoris resistantiae to various favorable or unfavorable secondary factors during the early or later posttraumatic period causing visible lesions to become invisible, and vice versa. Reversible lesions can turn into the irreversible lesions (Fig. 2). That is the reason why delayed scans may demonstrate lesions not

apparent on initial scans (Fig. 3) [10]. Primary CT findings performed immediately after injury are usually negative in 10–30% of cases. On the repeated scanning 75% of this number were positive [11]. These two types of lesions represent a complex which clearly defines the beginning of a traumatic cerebral disease. According to the site of impact and traumatizing force direction the lesions can be expected in some peculiar

1240

Fig. 3 a Small hemorrhagic lesion within the fornix on b the initial CT axial scan. c, d On the control CT 24 h later a new hemorrhagic lesion appeared in the right thalamus

Fig. 4 “The inner cavitation effect” with out-stretching of the corpus callosum. The longer diameter of the head is shortened and the shorter laterolateral direction is widened due to cerebrospinal fluid (arrows)

sites in the damaged brain. Generally, a linear translation of acceleration in a closed head injury can run along the shorter diameter of the skull in latero-lateral direction. In such cases an extra-axial lesion [subdural hematoma (SDH), epidural hematoma (EDH), subarachnoid hemorrhage (SAH)] can occur. Latero–lateral (L-L) courses of a traumatizing force usually lead to pronounced coup and countercoup contusions. Contusions are considerably less frequently present in a fronto-occipital (FO) or

an occipito-frontal (OF) direction of the traumatizing force. The brain stem can also be damaged, but it is damaged more often if the traumatic force acts along the longer diameter of the head. The course of a traumatizing force can run in FO or OF medial or oblique, left or right direction, named as a centroaxial course of the linear translation of an acceleration–deceleration type of trauma. Frontal blows are the most frequent type of a head injury in traffic accidents, whereas the occipito-frontal

1241

Fig. 5 The locations of the brain lesions produced by centroaxial course of linear translation of acceleration–deceleration type of traumas on different coronal schematic brain sections

Fig. 6 Patient, 18 years of age, suffering DAI. An old hemorrhagic lesion within the right frontal subcortical white matter demonstrated by gradientrecalled-echo (GRE) sequences (thin arrow) and T2-weighted sequences (thick arrow). Note that lesion is better seen on the GRE sequences

direction of the traumatizing forces appear most frequently in falls. The centroaxial blows produce a different pattern of lesion located mostly in the deep structures. A biomechanical explanation of the deep, centroaxial lesions has been offered by several authors [5, 6, 12, 13, 14]. They put forth the hypothesis of L–L expansion of the ventricles’ “inner cavitation effect” (Fig. 4). If a traumatic force acts along the longer diameter of the head, FO or OF, medial or paramedial, the longer diameter of the head will be shortened and the shorter L–L diameter will be widened due to cerebrospinal fluid. That causes the out-stretching of the corpus callosum (CC) and other midline structures (septum pellucidum, fornix, tella chorioidea of the third ventricle) with a displacement of the lateral ventricular wall resulting in a lesion within these structures. These structures are often affected. Lesions consist of hemorrhages and necrotic lacerations of various degrees. In such cases DAI occurs as well as brain stem lesions (Fig. 5). A parasagittal complex (PSC) consists of peculiar and significant lesions localized in the subcortical white matter of the parasagittal areas of the brain from the frontal

to the occipital region. The PSC is an associated and almost constant phenomenon of the pattern of deep intraaxial lesions. These lesions are a result of the stretching and tearing of long perforating veins which are attached to the sagittal sinus and suffer through a ‘‘gliding’’ of the brain in the rotating anteroposterior movements of the head in frontal blows (Fig. 6). The PSC occurs much less frequently in cases with occipital blows. The periventricular area belongs to the most frequently affected parts in deep intra-axial lesions. The biomechanical explanation of a periventricular complex (PVC) may be sought in the same mechanism which produces a lesion of CC (inner cavitation effect). These lesions are small, appearing in an acute phase, usually hemorrhagic and multiple, with a predilection in the white matter of the lateral ventricular corners. They are often found in patients with long survival. Demyelination and glial sclerosis were first described in these lesions by Strich 1956 [15]. Such a small lesion may play an important role in the development of a persistent vegetative state (Fig. 7). Lesions in the basal ganglia and the thalamus can be considered as a part of the centro-axial pattern, but they

1242

may also be the result of other mechanisms. A downward shifting of the brain due to large EDH, SDH, or a large countercoup complex, as well as lateral blows, may cause thalamic lesions, too. In the cases of centroaxial blows, especially upper frontal and vertex blows, the brain can be shifted toward tentorial hiatus causing lesions of the hippocampus and

Fig. 7 Axial non-enhanced CT scan shows a small acute hemorrhagic lesion in the left periventricular region. Hemorrhagic cortical parietal contusion on the right surrounded by narrow hypodense area corresponding to edema. Chronic subdural hematoma on the right. The outcome of the patient was persistent vegetative state Fig. 8 Slightly abnormal increase of signal on T1-weighted spin echo secondary to methemoglobin in the right parahipoccampal gyrus (thick arrow) and a small lesion within the midbrain and dorsal part of the pons (thin arrow)

the parahippocampal region (Fig. 8). These structures are damaged by L–L force direction as well as in centroaxial blows due to a direct physical conflict of the hippocampus with the rigid edge of the tentorium. The cortical areas suffer cuts, laceration, and disruptions followed by hemorrhage, necrosis, or edema. During the herniation, circulation is impaired and infarctions occur [5, 6]. Lesions of the hippocampal complex are difficult to visualize on CT scans [7]. Pathoanatomically, a brain stem lesion is a frequent, if not regular, component of the deep centroaxial pattern of lesions accompanied by diffuse hemispheric damage consisting of hemorrhage, laceration, contusion, and infarction [6, 7, 8, 17]. Lesions of the brain stem can be primary or secondary. Primary lesions can be divided into three categories: hemorrhage; necrosis; and lesions of axons including swelling, or fragmentation, with a formation of retraction bulbs. Primary brain stem lesions occur due to a downward shift of the brain or rotational force at the moment of the impact. Many of those patients die very soon after the accident. A secondary brain stem injury is caused by a delayed downward displacement due to edema and an increased intracranial pressure. It includes infarction, hemorrhages, or compression of the brain stems as a result of the adjacent or systemic pathology. A secondary brain stem lesion that occurs as a result of downward herniation or hypoxia–ischemia usually involves the ventral or ventrolateral aspect of the brain stem in contrast to a primary brain stem lesions which are most common in the dorsolateral aspect of the brain stem. A characteristic secondary midline brain stem lesion is Duret hemorrhage. It is believed to result from the stretching or tearing of penetrating arteries as the brain stem is caudally displaced. The brain stem infarction is another type of secondary brain stem injury which typically occurs in the central tegmentum of the pons and midbrain.

1243

Classification of brain injury Generally, a traumatic brain injury (TBI) can be divided into two groups: open-head injury and closed-head injury. By definition, in a closed-head injury dura is intact in comparison with an open trauma when dura is torn. Using imaging techniques the manifestations of a head trauma can be divided into primary and secondary lesions. During a traumatic event the underlying brain can suffer focal or diffuse lesions. Primary lesions are those that occur at the time of trauma as a direct result of the traumatic force. Primary lesions are scalp laceration/hematoma, skull fractures, the extra-axial (extra-cerebral) hemorrhage (epidural hematoma, subdural hematoma, subarachnoid hemorrhage, intraventricular hemorrhage), and intra-axial lesions (cortical contusion, diffuse axonal injury, and brain stem injury). Since scalp laceration/hematoma are easily detectable by naked-eye inspection, no special attention is paid to them in this article. Skull fracture can be detected by CT or X-ray. Secondary lesions occur as a consequence of primary lesions usually as a result of mass effect or vascular compromise. Secondary lesions represent diffuse cerebral edema and swelling, cerebral herniation, traumatic ischemia and infarction, and hypoxic injury. This division is clinically important because secondary lesions are often preventable, whereas primary lesions are a consequence of direct mechanical changes and are not preventable. Using imaging techniques it is possible to make fine distinctions between the two, as well as define an open vs closed brain injury.

Open-head injury/missile head injury The energy impacted into the head by penetrating missiles is primarily dependent on their velocity and mass, whereas the missile track within the brain is also related to the design and configuration of the projectile, the firing distance, and weapon orientation [17, 18]. When a highvelocity bullet enters the body and ploughs through the tissue, the material in its path will thoroughly disintegrate. Computed tomography is indispensable in missile head wounds, because it is faster than any other method and provides more detailed information on the extent of brain injury and the presence of retained bone and/or shell fragments (Fig. 9). This information is of the utmost importance for decisions on treatment. Imaging analysis in missile head injury should include the following steps: assess missile path; determine the extent of wound; bone fragments and in-driven shell fragments; and follow-up evolution of missile head injury. Since many such patients have major vascular lesions, cerebral angiography should also be considered.

Fig. 9 Transventricular missile track filled with blood and indriven bone fragments

Extra-axial hemorrhage There are three types of extracerebral hemorrhage: epidural hematoma (EDH); subdural hematoma (SDH); and subarachnoidal hemorrhage (SAH). In this group intraventricular hemorrhage (IVH) can also be included, considering ventricles as extra-axial, meaning extraparenchymal, structures. Epidural hematoma is usually arterial in origin and often results from a skull fracture that disrupts the middle meningeal artery. Because EDH exists in the potential space between the dura and the inner table of the skull, such hematoma can cross dural attachment but not cranial sutures (Fig. 10). On CT acute EDH appears as well-defined biconvex hyperdense extra-axial collection. Occasionally, within an acute EDH heterogeneous foci appear containing irregular areas of lower attenuation. This finding indicates active extravasation of fresh unclothed blood (swirl sign) and warrants immediate surgical attention (Fig. 11). Subdural hematoma is usually venous in origin, resulting from the stretching or tearing of cortical veins. Most of such hematoma are supratentorial, located along the convexity, frequently seen along the falx and tentorium from the anterior to the posterior falx, typically crescent shaped. Unlike epidural hematoma, subdural hematoma can cross sutural margins but not dural attach-

1244

Fig. 10 Epidural hematoma (EDH) and subdural hematoma (SDH) relative to the meningeal anatomy. (Adapted from Gean AD [32])

Fig. 11 Axial non-enhanced CT shows a large acute right parietal EDH. Note that the small low-density area (swirl sign) represents active bleeding with unretracted liquid clot

Fig. 12 A chronic bilateral SDH with re-bleeding and fluid–fluid levels

ments. In an acute stage a traumatic SDH shows the same imaging findings on CT and MR as any other acute non-traumatic hemorrhage. Diffuse swelling of the underlying hemisphere is common with SDH. Because of the increase in tissue fluid, edema causes decreased attenuation on CT images with a loss of gray matter–white matter differentiation. Occasionally, rebleeding occurs during the evolution of an SDH causing a heterogeneous

appearance from the mixture of fresh blood and partially liquefied hematoma. A sediment level or “hematocritic effect” may be seen either from rebleeding or in patients with clotting disorders (Fig. 12). Chronic SDH has low attenuation values, very similar to cerebrospinal fluid on CT and on all MR sequences. The MR appearance of SDH and hygromas are variable depending on the evolution of hemorrhage like in any other hemorrhage. During

1245

Fig. 13 Gradient-recalled-echo T2* sequences and T2-weighted imaging demonstrate diffuse hemosiderosis. The black outline of the subarachnoidal spaces is noted

the transition in appearance from acute to chronic SDH, an isodense phase occurs, usually between several days and 3 weeks after the acute event. Such isodense SDH can be overlooked on CT scans; therefore, attention must be paid to the following: gray–white matter interface is medially displaced, the border between the extra-axial collection and the underlying brain is barely discernible, and failure of surface sulci to reach the inner calvarial table is present. Contrast enhancement can help identify nonacute, isodense SDH by demonstrating an enhancing capsule or displaced cortical vessels. The brain stem and cerebellum are usually spared and may appear relatively hyperdense to cerebral hemispheres (white cerebellum sign). Focal areas of edema are frequently associated with cerebral contusions and may contribute significantly to mass effect. Subarachnoid hemorrhage results from disruption of small subarachnoidal vessels or direct extension into the subarachnoidal space by a contusion or hematoma (Fig. 13). Intraventricular hemorrhage results from a rotationally induced tearing of subependymal veins on the surface of ventricles or by direct extension of parenchymal hematoma into the ventricular system. On imaging SAH and IVH exhibit the same findings as in non-traumatic cases. During an MR examination GRE sequences are highly recommended with the aim of detecting hemosiderin even a few years after the trauma happened.

Cortical contusion Cortical contusions are bruises and lacerations of the brain which are covered by dura. These are wedgeshaped lesions with their base on the cortical surface and their tip pointed toward the center of the brain. Countercoup lesions are usually a little bit larger than coup lesions. Lesions are multiple or solitary and present in the cortical or subcortical region. During the first week after a traumatic event the characteristic CT pattern of mixed areas of hypo- and hyperdensity (“salt and pepper”) becomes more apparent. Cortical contusions are underestimated by CT and are best depicted by MR imaging [19]. Cortical contusions can lead to regional ischemia caused by extensive release of excitotoxic amino acids leading to increased cytotoxic brain edema and raised intracranial pressure [20]. There are some predilections for a certain portion of the brain where contusions occur, e.g., bases of the frontal lobes and the tips, and bases and lateral surfaces of the temporal lobes due to brain gliding upon the uneven (rough) surface of the skull base. Four to 6 months after trauma, this lesion becomes cystic (encephalomacia), which means hypodense on CT (Figs. 14, 15). Diffuse axonal injury (shearing injury)

Intra-axial lesions With cortical contusions, a diffuse axonal injury (DAI, or” shearing” injury) has been identified as the most important cause of significant morbidity in patients with a traumatic brain injury.

Diffuse axonal injury is one of the most common types of primary neuronal injury in a patient with severe head trauma. It is characterized by a widespread disruption of axons that occurs at the time of an acceleration or deceleration injury. A DAI lesion tends to be small, only a few millimeters in diameter, and multiple: 15–20 lesions

1246

Fig. 14 Extensive hemorrhagic cortical contusions of mixed density within the tip of the right temporal lobe and basal part of the frontal lobes due to the “gliding trauma”

Fig. 15 T2-weighted imaging shows very small cortical contusion along the superior temporal gyral surface surrounded by high-signal edema (long thin arrow). The right thalamic lesion (small thin arrow) and a small lesion within the splenium of corpus callosum (thick arrow) due to shearing injury is also seen

Fig. 16 Lesions of corpus callosum (corpus, splenium) demonstrated by a T2-weighted imaging(arrow) and b, c GRE sequences show the course of the traumatic force within splenium (b, arrow) and corpus callosum (c, arrow). Note that lesions are better seen on the GRE sequences

can be found in severely injured patients. This type of lesion occurs in a specific location of the traumatic brain [21, 22, 23, 24]. This peculiar pattern of lesions caused mostly by a centroaxial traumatic force are located within the corpus callosum complex (septum pellucidum, fornix, tela chorioidea), gray matter–white matter junction in parasagittal areas, and deep periventricular white matter (not superficial), especially in the frontal area at the corner of verticles, basal ganglia, internal capsule, hippocampal and parahippocampal region, brain stem, and cerebellum (Figs. 16, 17). Therefore, these areas should be examined very thoroughly, seeking for a minimal lesion. Regarding the location of DAI, three different stages have been established [22]. Magnetic resonance imag-

ing is more sensitive than CT in its detection especially in nonhemorrhagic lesions [21, 22, 23, 24]. The MR appearance of DAI depends on the presence of hemorrhage and the age of the lesions [10, 21, 22]. Brain stem injury During CT or MR examination special attention should be paid to an analysis of the brain stem. The most common form of primary brain stem injury is DAI which affects the rostral dorsolateral aspect of the midbrain and upper pons. Commonly lesions are in the periaqueductal region in comparison with a secondary lesion which usu-

1247

Fig. 17 Multifocal low signal on GRE T2* sequences caused by shearing injury correspond to a small foci of former hemorrhage within the periventricular white matter (arrow), basal ganglia, and b deep parasagittal subcortical hemispheric white matter (arrows)

tions of the posterior fossa. Magnetic resonance imaging is absolutely the method of choice in brain stem injury analyses and is also valuable in predicting the outcome [16, 25, 26].

Secondary lesions in brain injury

Fig. 18 Primary brain stem hematoma in the periaqueductal region on the left accompanied by SDH temporoparietal on the left and SAH. Perimesencephalic cysterns are compressed, brain stem is edematous due to edema, and contralateral temporal horn is enlarged

ally appears in the ventral part of the brain stem (Fig. 18). According to Hashimoto et al. only 8.8% of brain stem injury can be detected by CT [25]. Firsching et al. has detected brain stem lesion using MR imaging in 64% of patients who suffered severe brain injury [16]. On cranial CT scanning brain stem lesions are difficult to detect because of bone artifacts within the lower por-

The secondary effects of a craniocerebral trauma are sometimes of greater importance than direct manifestations such as focal hematoma, contusion, or DAI. Secondary alterations represent diffuse cerebral edema, hypoxia, infarction, necrosis, secondary hemorrhage, and cerebral herniation (Fig. 19). Most secondary injuries are caused by an increased intracranial pressure or cerebral herniations. They occur in the early posttraumatic period. Using CT and MR those lesions can be easily distinguished from the late sequelae which are the final result of brain injury such as hydrocephalus, pneumocephalus, cerebrospinal fluid leak, and encephalomalacia. In routine MR protocol GRE sequences should be absolutely included to exclude a former hemorrhage, since they are very sensitive with regard to the detection of different blood products. The MR signal intensity varies depending on sequences and time scanning after trauma. From the group of early secondary lesions we describe only chosen topics that might be a threat to the patient’s life. Diffuse cerebral edema Diffuse brain swelling is a common manifestation of head trauma. It may occur either because of an increase of cerebral blood volume or an increase of tissue fluid

1248

content. Hyperemia refers to an increase of blood volume, whereas edema refers to an increase in tissue fluid; both lead to generalized mass effect with effacement of sulci, suprasellar and quadrigeminal plate cistern, and compression of the ventricular system. Homogeneous attenuation of brain parenchyma is present on CT scans with a loss of gray matter/white matter interface. Cerebellum may appear relatively hyperdense in comparison with edematous cerebral hemispheres (white

cerebellum sign). “Pseudo-subarachnoidal hemorrhage” is seen in cases of severe, diffuse cerebral edema when the brain becomes very low in attenuation and dura and circulating blood in the cranial vasculature appears unusually hyperdense on CT (Fig. 20). Cerebral swelling from hyperemia is most commonly seen in children and adolescents due to loss of normal cerebral autoregulation. The main problem of closed head injury is brain swelling as a result of all perifocal tissue damages. In children, a brain swelling starts developing as early as 20–30 min after a head injury and advances very fast; therefore, the observation of consciousness in children is necessary even in cases with a mild head injury [27, 28]. Brain herniation

Fig. 19 On axial non-enhanced CT infarction of internal carotid artery is seen. Within the ischemic brain parenchyma multiple secondary hemorrhagic lesions are discerned. Extensive subfalcine herniation of the brain parenchyma with mass effect. Acute lamelar subdural hematoma on the right Fig. 20 Pseudo-subarachnoidal hemorrhage in severe diffuse brain edema shown by CT

Cerebral herniations are caused by a mechanical displacement of the brain, cerebrospinal fluid, and blood vessels from one cranial compartment to another. Using CT or MR different types of brain herniation are recognized as follows: subfalcine herniation; transtentorial (descending and rarely ascending due to upward herniation of the cerebellum); transphenoidal (descending and ascending); and tonsilar when cerebellar tonsils are forced through the foramen magnum. All types of herniation are a serious sign of cerebral injury accompanied by displacement of blood vessels and nerves. In cases involving subfalcine herniation the cingulate gyres is displaced across the midline under the falx and anterior cerebral artery is displaced resulting in secondary ischemia and infarction. A downward displacement of the temporal lobes and brain stem through tentorial incisura is termed descendent transtentorial herniation (Fig. 21). Descending transtentorial herniation is a very serious consequence of brain injury. On imaging it can

1249

ly studied [4, 5, 6, 12, 13, 14]. Some works have emphasized the value of CT data in the study of the biomechanical condition of the brain injury [7, 11]. Since brain damage in a closed head injury of acceleration– deceleration type depends on the site of impact and on the course of traumatizing forces, we attempted to reconstruct it by CT in the entire traumatized brain as well as in the corpus callosum. Along with that we studied CT in a group of patients in a persistent vegetative state as a consequence of brain injury. We were also interested in missile head injury and its evolution assessed by CT. The pathoanatomical studies and the known biomechanical condition in acceleration–deceleration brain injury were the guiding ideas for us to investigate the limits and possibilities of CT in patients who had suffered closed- and open-head injury, respectively. Some of our results have already been published and are briefly presented here, whereas some of them represent our new investigation results.

Materials and methods Fig. 21 Axial CT shows transtentorial herniation. Subdural hematoma fronto-temporal on the right with mass effect. Ventricle system is compressed and shifted. The contralateral temporal horn is enlarged. Basal cisterns are compressed. The brain stem and brain parenchyma are hypodense due to the edema

be recognized by a dilatation of the contralateral ventricular system, compression of the basal cisterns, and midline shifts of the brain parenchyma. Mortality rate in such cases is high.

Late sequelae of trauma Late sequelae of trauma can also be detected by imaging. Mostly they are hydrocephalus, pneumocephalus, ischemia–infarction, cerebrospinal fluid leak, leptomeningeal cyst, encephalomalacia, and atrophy. Focal encephalomalacia consists of tissue loss with surrounding gliosis and is a frequent manifestation of remote head injury. The appearance of encephalomalacia is not specific for posttraumatic injury but its locations are characteristic: the anteroinferior part of frontal and temporal lobes (predilections for cortical contusions).

Purpose of the study The course of a traumatizing force and its effect on brain damage in a closed head injury were systematical-

Very shortly presented are the results of 45 standard CT examinations of patients with closed acceleration–deceleration head injury with the aim to reconstruct the site of impact and the course of traumatizing force [29]. The site of impact and the course of traumatizing force were reconstructed and graphically presented. A comparison between the computerized graphic presentation of the site of impact and direction of the traumatic force and the location of lesions revealed a high correlation between them. Our results show that CT is very useful for a reconstruction of the site of impact and of the course of the traumatizing force in acceleration head injury. Data obtained by this procedure may have far-reaching prognostic and forensic implications. Since the corpus callosum is an especially interesting formation in the complex of closed-head injury, we studied patterns of lesions of CC in inner cerebral trauma visualized by CT [30]. In this study we analyzed the topographic variations of the lesions of CC and tried to correlate them with the known biomechanical conditions of the traumatic event. Our results showed significant congruence between the course of linear translation of acceleration and the sites of the lesions for brain and for CC. In a group of 285 patients admitted to the Special Hospital for Medical Rehabilitation Krapinske Toplice, Croatia, during 1998 and 1999 who had suffered brain injury, we selected 37 patients in a clinically established persistent vegetative state as a consequence of severe brain injury. They all underwent conservative treatment for several months and showed no recovery. The study group consisted of 33 men (mean age 38.8 years) and 4 women (mean age 38.5 years). In all selected patients retrospective analysis of CT scans was carried out. All patients in the study group suffered severe brain injury with Glasgow Coma Scale Score less than 8 at the admission to the different emergency departments at the moment of respective traumatic events. They were followed-up clinically and by CT in the course of 1 year. The initial CT scan was performed in each patient within 24 h after trauma, and the second between 24 h and 14 days after traumatic event, mostly between 24 and 72 h. From two to seven CT examinations were done in all patients in the period between 14 days and 12 months after trauma. Ten patients were polytraumatized. The mechanism of trauma is demonstrated in Table 1. Twenty six patients from the study group were involved in traffic accidents. The results of CT findings are sum-

1250

Table 1 Mechanism of injury

Fall Traffic Motorcycle Pedestrian Axe injury Missile injury No. of patients

4 26 2 1 1 3 37

Table 2 Initial CT findings (1–14 days after injury). SAH subarachnoid hemorrhage; EDH epidural hematoma; SDH subdural hematoma; IVH intraventricular hemorrhage; ICH intracerebral hemorrhage SAH EDH SDH IVH ICH Edema

4 (10.81) 2 (5.41) 6 (16.22) 3 (8.11) 10 (27.03) 5 (13.51)

Contusion Pneumocephalus White matter Basal ganglia Ischemia Brain stem

8 (21.62) 1 (2.73) 6 (16.22) 15 (40.54) 5 (13.51) 5 (13.51)

Numbers in parentheses are percentages

Table 3 Follow-up CT findings (14 days to 12 months) Numbers in parentheses are percentages

Atrophy 16 (43.24) Hydrocephalus 19 (51.35) Hygroma 4 (10.81) Encephalomalacia 15 (40.54)

marized in Tables 2 and 3. On the two initial CT scans lesions in basal ganglia (40.54%), ICH (27.03%), and cortical contusions (21.62%) were the dominant changes. Lesions in basal ganglia were less than 10 mm in diameter, and were hypodense or hyperdense, or of mixed quality. Cerebral contusions showed similar changes. White matter changes were found in the periventricular areas. They were mostly hyperdense. Changes in the brain stem were present in 5 patients demonstrated by CT through the two initial CT examinations. On later control CT scans hydrocephalus, atrophy, and encephalomalacia were the most common changes, as is shown in Figs. 22 and 23, and in Table 3. Our results in the study group in a persistent vegetative state as a consequence of closedhead injury showed that the outcome in patients with basal ganglia, periventricular white matter changes, and brain stem damage is poor. Although the lesions in basal ganglia and white matter were small, they had to be examined very carefully. Summarizing all of our own results in patients with closedhead injury, we emphasize that lesions in different sites on CT scans due to the known biomechanism of the closed acceleration–deceleration brain injury should be interpreted dynamically and not separately, with special attention given to minimal brain lesions. Minimal brain lesions may complete the mosaic for a reconstruction of biomechanical condition in each case, which may be important from both clinical and forensic standpoints. All our results showed that using CT the mechanism of closedbrain injury can be demonstrated in a living traumatized patient without or before pathological section is performed. We also studied 154 patients who had suffered open, missilehead injury [31]. In 54% of cases CT was performed 1–24 h after injury, and in 27% follow-up CT was obtained showing evolution of hemorrhage, edema, cerebritis, abscess, secondary vascular lesions, necrosis, encephalomalacia, and hydrocephalus. The most dynamic changes occurred 7–14 days after injury.

Fig. 22 Patient, 26 years of age, with persistent vegetative state CT scan 2 months after injury shows an extensive lesion within the frontal periventricular white matter and at the gray matter/ white matter junction (long arrow). A hypodense lesion in the right putamen (thin arrow). Mild dilatation of the ventricles

Fig. 23 Patient, 33 years of age, who suffered DAI with persistent vegetative state. Non-enhanced CT shows marked dilatation of the ventricles with diffuse brain atrophy

1251

Conclusion Computed tomography is currently the first imaging technique to be used after head injury, in those settings where CT is available. Using CT scalp, bone, extra-axial hematomas, and parenchymal injury can be demonstrated. Computed tomography is rapid and easily performed, also in monitored patients. It is the most relevant imaging procedure for surgical lesions. Magnetic resonance imaging is more sensitive for all posttraumatic lesions except skull fractures and subarachnoidal hemorrhage, but scanning time is longer, and the problem with the monitoring of patients outside the MRI field is present.

If CT does not demonstrate pathology, as can adequately be explained to account for clinical state, MRI is warranted. Follow-up is best done with MRI as it is more sensitive to parenchymal changes. In conclusion, CT remains the first-line examination to detect immediately life-threatening lesions. Magnetic resonance imaging is the examination of choice for full assessment of brain lesions. The knowledge of the mechanism of the closed acceleration–deceleration type of brain injury is very important and can help in detecting even minimal traumatic brain lesions with the use of CT and MR which might have influence on the outcome of the patient.

References 1. Frankowski RF, Annergers JF, Whitman S (1985) Epidemiological and descriptive studies. The descriptive epidemiology of head trauma in the United States. In: Becker DP, Polishock J (eds) Central nervous system trauma status report. National Institute of Neurological and Communicative Disorders and Stroke. National Institutes of Health, Bethesda, Maryland, pp 33–51 2. Pavic´ D, Kralj K, Bešenski N, Potocˇki K, Zˇ arkovic K, Jadro-Šantel D, Stiglmayer N (1994) Some epidemiological features of fatal head injury. Neurol Croat 43:3–7 3. Macpherson BCM, Macpherson P, Jennett B (1990) CT incidence of intracranial contusion and hematoma in relation to the presence, site and type of skull fracture. Clin Radiol 42:321–326 4. Leestma JE, Grcˇ evic´ N (1988) Impact injuries to the brain and head. In: Leestma JE, Kirckpatrick JB (eds) Forensic neuropathology. Raven Press, New York, pp 409–416 5. Grcˇ evic´ N (1984) Investigations of neuropathology and pathogenetical mechanisms of closed brain injury of the acceleration–deceleration type. Final report of project no. 02–05–1. National Institutes of Health, Bethesda, Maryland 6. Grcˇ evic´ N (1988) The concept of inner cerebral trauma. Scand J Rehabil Med Suppl 17:1–7 7. Jadro-Šantel D, Grcˇ evic´ N, Bešenski N, Kalousek M (1989) The inner cerebral trauma: the correlative pathology with CT scans. In: Aichner F, Gerstenbrandt F, Grcevic N (eds) Neuroimaging II. Fischer, Stuttgart, pp 263–273 8. Bešenski N, Jadro-Šantel D (1992) CT detection of minimal brain lesions in closed cerebral trauma. Neurol Croat 41:33–42

9. Wardlaw JM, Statham PFX (2000) How often is haemosiderin not visible on routine MR following traumatic intracerebral haemorrhage? Neuroradiology 42:81–84 10. Gentry LR (1991) Primary neuronal injuries. Neuroimaging Clin North Am 1:411–432 11. Tatalovic-Osterman LJ, Jadro-Šantel D, Bešenski N (1991) Diagnostic possibilities of the closed head injuries of acceleration type using computed tomography. Neurol Croat 40:231– 245 12. Lindenberg R, Freytag E (1957) Morphology of cortical contusions. AMA Arc Pathol 63:23–42 13. Lindenberg R, Freytag E (1960) The mechanism of cerebral contusions. AMA Arc Pathol 69:440–469 14. Sellier K, Unterharnscheidt F (1963) Mechanik und Pathomorphologie der Hirnschaden nach stumpfer Gewalteinwirkung auf den Schädel. Hefte Unfallheilkunde 76:1–1 15. Strich SJ (1956) Diffuse degeneration of the cerebral white matter in severe dementia following head injury. J Neurosurg Psychiatry 19:163–185 16. Firsching R, Woischneck D, Diedrich M, Klein S, Ruckert S, Wittig H, Dohring W (1998) Early magnetic resonance imaging of brain stem lesions after severe head injury. J Neurosurg 89:707–712 17. Siccardi D, Cavaliere R, Pau A, Lubinu F, Turtas S, Viale GL (1991) Penetrating craniocerebral missile head injuries in civilians: a retrospective analysis of 314 cases. Surg Neurol 35:455– 460 18. Kirckpatrick JB (1988) Gunshots and other penetrating wounds of the central nervous system. In: Leestma JE, Kirckpatrick JB (eds) Forensic neuropathology. Raven Press, New York, pp 276–299

19. Braun M, Cordoliani YS, Dosch JC (2000) Head and brain injuries. Place of imaging. Ann Fr Anesth Reanim 4:296–298 20. Hoelper BM, Reinert MM, Zauner A, Doppenberg E, Bullock R (2000) rCBF in hemorrhagic, non-hemorrhagic and mixed contusions after severe head injury and its effect on perilesional cerebral blood flow. Acta Neurochir 76:21– 25 21. Paterakis K, Apostolas H, Karantanas, Komnos A, Volikas Z (2000) Outcome of patients with diffuse axonal injury: the significance and prognostic value of MRI in the acute phase. J Trauma 49:1071–1075 22. Gentry LR (1994) Imaging of closed head injury. Radiology 297:1–17 23. Gentry LR, Godersky JC, Thompson BH (1988) MR imaging of head trauma: review of the distribution and radiolopathologic features of traumatic lesions. Am J Neuroradiol 9:101–110 24. Parizel PM, Ozsarlak O, Van Goethem JW et al. (1998) Imaging findings in diffuse axonal injury after closed head trauma. Eur Radiol 8:965–969 25. Hashimoto T, Nakamura N, Richard KE et al. (1993) Primary brain stem lesions caused by closed head injury. Neurosurg Rev 16:291–298 26. Shibita Y, Matsumura A, Meguro K, Narushima K (2000) Differentiation of mechanism and prognosis of traumatic brain stem lesions detected by magnetic resonance imaging in the acute stage. Clin Neurol Neurosurg 102:124–128 27. Zimmeman RA, Bilanuik LT, Buce D, Dolinskas C, Obrist W, Kuhl D (1978) Computed tomography of pediatric head trauma: acute general brain swelling. Radiology 126:403–408

1252

28. Ross PS, Ross AM (1989) Should patients with normal cranial CT scans following minor head trauma be hospitalized for observations? Pediatr Emerg Care 5:216–217

29. Bešenski N, Broz R, Jadro-Šantel D, Pavic D, Mikulic D (1996) The course of traumatising force in acceleration head injury: CT evidence. Neuroradiology 38:36–41 30. Bešenski N, Jadro-Šantel D, Grcˇ evic´ N (1992) Patterns of lesions of corpus callosum in inner cerebral trauma visualized by computed tomography. Neuroradiology 34:126–130

31. Bešenski N, Jadro-Šantel D, JelavicKojic F, Pavic D, Mikulic D, Glavina K, Maškovic J (1995) CT analysis of missile head injury. Neuroradiology 37:207–211 32. Gean AD (1994) Imaging of head trauma. Raven Press, New York, pp 76–78