If the imaging facility is informed in advance, the scanner can be kept free for the patient with acute stroke. Emergent life support should be continued during the patient’s imaging test if necessary. Direct transportation of the patient to the scanner will save time. Neurological examination can be performed on the scanner. While a detailed neurologic examination is not needed before the brain imaging is done, localizing the stroke to the posterior fossa or the cerebral hemisphere will assist in optimizing the imaging techniques selected.

CT should be first performed without contrast with a rapid scan time to reduce motion artifact. A correct head position is crucial to avoid obliquity of the sections. If necessary due to motion artifact, specific section cuts can be repeated. The high density of bone often causes an artifact, which may impair visualization of the lower part of the brain stem in the posterior fossa. Single slice scanning should be preferred to continuous spiral scanning. Thin (e.g. 0.625 mm) transaxial sections with 2.5 mm and 5 mm reconstructions can minimize artifacts and enhances sensitivity for thrombus detection [5]. Image windows should be adjusted so that gray and white matter can be easily distinguished and subtle hypodensities are detected. A window width of 70–80 HU at a center of 35–40 HU is recommended. A close communication between the stroke physician and the radiologists or technician performing the scan will enhance the information gained from CT. The image should be optimized for spatial and contrast resolution in the region of interest. Additional bone window views can be performed if head trauma is a possibility, in order to search for skull fractures, subdural air or blood, or effusions in nasal sinuses and middle ear. A contrast-enhanced scan might be obtained after, if there is suspicion for neoplasm or localized infection and no MRI available.

In case of suspected major cerebral artery occlusion that may require intra-arterial intervention, CT angiography (CTA) should be prepared and performed immediately after non-enhanced CT. Immediate interpretation of the non-enhanced scan may allow the start of thrombolytic infusion in parallel to CTA.

Advanced CT technology provides the opportunity of CTA and CT perfusion imaging with one contrast injection only [6, 7]. A conventional CTA is performed after bolus injection of 130 ml non-ionic contrast (injection rate: 4–5 ml/s) with a spiral scan of the brain base and 3-dimensional reconstruction of the Circle of Willis on a workstation. Perfusion imaging with CT requires repeated imaging of one or more sections to measure the contrast uptake and clearance curve in each voxel. Parameter images are then calculated for cerebral blood volume (CBV), mean transit times (MTT), cerebral blood flow (CBF), or time intervals to the peaks of contrast enhancement—time to peak (TTP) (Fig. 12.2) [8]. Multidetector CT technique now allows whole brain perfusion imaging.

Pathological findings on non-enhanced CT (NCT) in acute stroke patients may be thrombo-embolic occlusion of large vessels, focal brain tissue swelling caused by vasodilatation, ischemic edema, intracranial hemorrhage, focal brain tissue inflammation, or tumor like lesions. Computed tomography identifies these findings by detecting changes in x-ray attenuation of normal brain structures, a shift or replacement of brain structures by pathological substrates, or pathological contrast enhancement. A pathological increase in x-ray attenuation is called “hyperdensity”, a pathological decrease “hypodensity”. These terms are somewhat confusing because they do not define a fixed degree of x-ray attenuation. They are commonly used to characterize the attenuation of a structure in comparison to other tissue, e.g. in saying that a parenchymal hematoma is identified by its hyperdensity if compared with gray matter. These terms are best used, however, to characterize a change in x-ray attenuation by comparing the “density” of an affected structure to its normal “density”. The symmetry of the brain structures in transaxial planes facilitates this comparison. e.g., the putamen is best evaluated by comparing its attenuation to that of the contralateral putamen and to that of the head of the caudate nucleus, because the caudate nucleus and the putamen are portions of the same anatomical structure, the striatum (Fig. 12.3). It is obvious, that low technical quality of the scan (motion artifacts, wrong window width or wrong level) and in particular any obliquity of the scan impair the recognition of real changes in x-ray attenuation (Fig. 12.4).

In acute stroke, blood may be present in one or more of cranial compartments: brain parenchyma, ventricles, subarachnoid space, and subdural or epidural space. Clinically, an acute parenchymal hemorrhage cannot be distinguished reliably from ischemic stroke. After acute hemorrhage, blood appears as a hyperattenuated, often space-occupying mass if clotted (Fig. 12.6). The degree of hyperdensity depends on the amount of blood, its hematocrit, whether it is clotted or not, and whether the blood is intermixed with CSF or brain tissue. Hemorrhages related to coagulopathies or treatment with anticoagulants or thrombolytics are often inhomogeneous with fluid levels (Fig. 12.7). Sensitivity of CT for the detection of parenchymal hemorrhage is nearly 100 %, but small hemorrhages into the brain parenchyma or subarachnoid space can be overlooked (Fig. 12.8) in particular if the patient has a low hematocrit, e.g., in hematologic disorders. The investigators of the European Cooperative Acute Stroke Studies (ECASS, ECASS II) missed two small parenchymal hemorrhages and one subarachnoid hemorrhage (SAH) in 1,420 patients (0.1 %) [29, 30]. The detection of blood within infarcted, hypodense brain tissue is problematic, but of importance. The hemorrhagic transformation of ischemic brain tissue is normal; the extent of hemorrhage varies under the influence of reperfusion, thrombolytics, and anticoagulants. Hemorrhagic transformation may appear on CT like normal gray matter surrounded by ischemic edema—slightly hyperattenuating spots scattered within the infarct—or it may present as a dense hematoma within the ischemic edema, with and without space occupying effect. It is unlikely that mild degrees of hemorrhagic transformation affect the neurological status of the patient. The term “symptomatic hemorrhage” for each sign of blood within ischemic infarctions is, therefore, highly problematic and may have contributed to the overestimation of the risk from these types of bleedings after thrombolytic therapy [31]. It was shown that only intracerebral hematomas with space occupying effect are associated with clinical deterioration and poor outcome [32, 33]. The ECASS investigators have used categories for four different types of hemorrhagic transformation of ischemic brain tissue: hemorrhagic infarction, grade 1 and 2 (HI1 and HI2), and parenchymal hematoma, grade 1 and 2 (PH1 and PH2) [34]. The location of hematoma may provide clues about its underlying etiology (Table 12.2). Acute hemorrhages usually show hyperattenuation without surrounding edema. If marked edema is present under those circumstances, underlying neoplasm should be suspected. Multiple hemorrhagic lesions should suggest metastatic disease, coagulopathy, cavernomas, or cerebral amyloid angiopathy.

Stay updated, free articles. Join our Telegram channel

Join