CT of the head in stroke

Author: Mikael Häggström

Planning

Choice of modality

• CT of the head in stroke is generally the first investigation because of high availability. It detects about 60% of infarcts in the first 3 to 6 hours, and almost always thereafter.
• MRI of the head in stroke is generally less available, but preferred if there are no MRI contraindications, and your clinic has established procedures for its performance in acute stroke (if you do not know, it is reasonable to assume that there is not, and do CT).

Still not sure? See: CT vs MRI in stroke

How soon

According to UK guidelines, imaging should be performed immediately for people with suspected acute stroke if any of the following apply:

• Indications for thrombolysis or early anticoagulation treatment
• Current anticoagulant treatment
• Any known bleeding tendency
• Depressed level of consciousness of Glasgow coma scale (GCS) of less than 13
• Unexplained progressive or fluctuating symptoms
• Papilledema, neck stiffness or fever
• Severe headache at onset of stroke symptoms.

"Immediately" is defined as 'ideally the next slot and definitely within 1 hour.

In case of acute stroke without indications for immediate imaging, it should be performed within a maximum of 24 hours after onset of symptoms.

Configuration

Initially without IV contrast.

Non-contrast CT

NCCT is performed as a first line in neuroimaging as it is usually considered gold standard in ruling out hemorrhage despite MRI being equivalent to CT in this regard due to the ease of availability, rapid results, and safety in both stable and unstable patients . Advancements in technology enabling thinner slices (sub millimeter) and enhanced tissue differentiation allows for a role of NCCT in arterial occlusion and early infarction with a variable detection rate of about 65% in cases imaged within 6 hours. A description of these signs is below.

Hyperdense artery/MCA sign

This is a result of a thrombus or embolus (usually in the MCA) resulting in an increased density of the blocked vessel. The prevalence ranges from about 27% in all types of strokes to 41% with MCA infarct with a specificity of 100%, but sensitivity of only 30% . This sign disappears with resolution of the thrombus in a few days. False positives are due to calcification of the walls or a high hematocrit (Figure 1).

Hypo-attenuating brain tissue

Ischemia causes cytotoxic edema; an increase in brain water by 1% results in a CT attenuation decrease of 2.5 HU (Hounsfield Units) . The specificity of ischemic edema on NCCT for brain infarcts is 85% and sensitivity was 64%, with lack of early CT findings resulting in better 90 day clinical outcomes and vice versa . Also seen is cortical sulcal effacement (Figure 2).

• CT of dense media sign.png

  Figure 1. Hyperdense Artery Sign. The “Dense MCA or hyperdense vessel/artery sign” is one of the early signs of acute middle cerebral artery (MCA) territory infarction. This corresponds to a hyper-dense cord like MCA seen in the regionof the Sylvian fissure on non-contrast CT which represents thrombotic material in main stem of the MCA.

• CT of cerebral infarction.png

  Figure 2. Left middle cerebral artery (MCA) infarction. Axial nonenhanced computer tomography shows foci of hypoattenuation in the left parenchyma (arrows) and sulcal effacement in the left MCA territory, consistent with infarction.

Obscuration of the lentiform nucleus

Also called blurred basal ganglia; it is one of the most common and earliest seen sign of infarction (MCA) due to terminal blood supply pattern. The loss of the grey-white matter interface and CT hypodensity results in the ‘obscuration’ of the lentiform nucleus.

Insular ribbon sign

Another early and indicative sign of infarction (MCA) refers to hypodensity and swelling of the insular cortex results in loss of the insular ribbon (Figure 3).

File:CT of insular ribbon sign.png

Figure 3: Insular ribbon signrefers to loss of the normal grey-white matter differentiation in the insular cortex. It is one of the earliest imaging signs of middle cerebral artery territory infarction. The yellow arrow shows a normal insular cortex (ribbon visble). However upon infarction leading to cytotoxic edema, this ribbon is lost (red arrow).

Hemorrhagic infarct

There is usually a sharp contrast between blood (high attenuating- seen as brighter white areas) and CSF (low attenuating-dark areas) (Figure 4 A & B).

• CT of subarachnoid hemorrhage.png

  Figure 4A: Axial nonenhanced CTshows “bright” or hyper attenuating dense subarachnoid hemorrhage throughout the perimesencephalic cistern (arrow), along the tentorium (double arrows), and from there to the 4th (double arrowheads) and 3rd (arrowhead) ventricles.

• CT of basal ganglionic hemorrhage.png

  Figure 4B: Nonenhanced axial CT demonstrates a large right basal ganglionic hypertensive bleed (*) complicated by mass effect, midline shift (subfalcine herniation)to the left (arrows). The blood is seen as hyper-attenuating or bright.

By altering standard viewing parameters, the sensitivity and specificity of stroke detection can be increased.

Quantification of ischemic involvement - ASPECTS

File:Cerebral regions by ASPECTS.png

Alberta Stroke Program Early Ct Score (ASPECTS). Schematic diagram showing various areas used for scoring on the ASPECTS. Level of internal capsule & insula: M1 - cortical area anterior to Sylvian fissure; M2 - cortical area just posterior to Sylvian fissure; M3 - more posterior cortical area in MCA distribution at same level. At cut near top of the lateral ventricles: M4 - anterior third of cortex; M5 - middle third of cortex; M6 - posterior third of cortex.

Alberta Stroke Program Early CT score (ASPECTS) is a 10-point quantitative topographic CT scan score developed in 2001 offering a reproducible grading system to score early ischemia in anterior circulation strokes to better direct treatment and reduce the variability of observations . Using two standard axial CT slices; one at the level of the thalamus and basal ganglia, and one just rostral to the basal ganglia, the MCA territory is divided into 10 regions,each accounting for one point in the total score, for each involved area, a point is subtracted. This score correlated inversely with the NIHSS (National Institutes of Health Stroke Score) with clinicians agreeing it superior and more systematic compared to the conventional 1/3 MCA rule to exclude thrombolytic treatment . The ASPECTS method is not without its limitations; such as difficulty in scoring due to age related periventricular white matter changes or streak artifacts in the base of the skull or tilt and motion artifacts. (ASPECTS diagram).

Computed tomography angiography (CTA)

CTA is a minimally invasive study with an optimally timed rapid injection of iodinated contrast through a peripheral IV (intravenous) line to cause vascular opacification, obtaining thin section CT images and using software to stitch the images allowing for a 3-dimensional image of cerebral and neck vasculature (from the aortic arch to the circle of Willis). This allows for identification of stenosis and occlusions; assisting therapeutic decisions such as IV or intraarterial TPA, mechanical clot retrieval or in cases of carotid dissection, against such a therapy. CTA also identifies vascular abnormalities such as arterio-venous malformations and aneurysms. CTA demonstrated occlusion does correlate with the NIH Stroke Score and outcome of TPA (Figure 4C, 5A, B, C, D).

• CT angiography of a vascular malformation with intraventricular hemorrhage.png

  Figure 5A: Sagital reformatted images from CT angiogram identify an enhancing vascular malformation (arrow), which was the etiology of the intraventricular hemorrhage.

• CT angiography of basilar tip aneurysm.png

  Figure 5B: Computed tomography angiogram in the axial projection demonstrates a focal basilar tip artery aneurysm (arrow).

• Volume rendered CT angiography of anterior communicating artery aneurysm.png

  Figure 5C: Volume rendering of cerebal computed tomography angiogram (CTA) with

CTA Source Images: CTA SI using the images in a CTA, cerebral perfusion can be assessed as low density/dark areas in contrast to hyper-attenuated contrast areas allowing for an estimation of tissue perfusion and a better assessment of tissue at risk compared to NCCT potentially removing the need for a separate CT perfusion study.

Computed tomography perfusion imaging (CTP)

CTP (like CTA) tracks an IV bolus of iodinated contrast over time with sections of the brain imaged repeatedly. This allows the measurements of parameters such as cerebral blood volume, cerebral blood flow, mean transit time (time difference between arterial inflow and venous outflow), time to peak enhancement (time from the beginning of the contrast injection to the maximum concentration in a region of interest). These parameters can be extrapolated to delineate areas of hypo-perfusion and irreversible infarction by creating perfusion maps. CTP has shown incremental increased sensitivity and specificity in diagnosing acute ischemic stroke compared to NCCT or CTA . CTP is easily available and can be performed on a standard helical CT after NCCT. Clinically, in acute stroke CTP provides information on the penumbra (increased mean transit time, moderately decreased cerebral blood flow and normal to high cerebral blood volume due to auto regulation or if blood flow is markedly decreased then decreased cerebral blood volume) and on the infarcted tissue (severe decrease in cerebral blood flow, and blood volume with increased mean transit time) with defined cut-offs for each criteria. A drawback of CTP is the need to analyze several brain slices for accurate flow data, requiring a multi detector CT with higher slice row, (currently at 2 slices, evolving to 64 slices) and high radiation exposure. Different techniques are employed such as Dynamic Contrast-enhanced CT and Perfusedblood- volume Mapping . Dynamic contrast enhanced CT consists of monitoring the passage of iodinated contrast bolus which causes a transient increase in attenuation which is in linear relation to the amount of contrast in the region used to generate curves for arterial and venous Regions of Interest which are converted using mathematical models into the perfusion parameters and color coded perfusion maps.

File:CT perfusion in M1 artery occlusion.png

Figure 6B: This perfusion map gives a quick visual estimate of the anterior cerebral artery (ACA) and middle cerbral artery (MCA) collaterals on the convexity in a case of M1 artery occlusion. This patient underwent CT perfusion study in setting of acute aphasia and hemiparesis. A prominent mismatch is present on CT perfusion with delay in time to peak (TTP) but relatively preserved rCBV, (cerebral blood volume) indicating adequate collateral supply with the exception of basal ganglia (M1 thrombus will nearly always take out the lenticulostriate vessels which are effectively end-arteries without significant collaterals).

Perfusion maps can give a quick visual read for color changes indicative for perfusion deficits or through measurements (usually not required). Perfused-blood-volume Mapping consists of subtracting the unenhanced CT data from the CTA source image data giving cerebral blood volume data with the advantage of allowing evaluation of the whole brain. However, since it does not allow determination of the mean transit time, blood flow and hence the ischemic penumbra, it clinically has a lesser use (Figure 6A,B).

• CT perfusion with flow and volume maps in cerebral infarction.png

  Figure 6A: (1) Regional cerebral blood flow map from computed tomography perfusion in a case of left middle cerebral artery infarct shows a large perfusion defect in the left frontal and temporal lobes, evidenced by a lack of color display. (2) Regional cerebral blood volume map demonstrates a penumbra of decreased perfusion (indicated with arrows around blue areas) surrounding the defect (purple), indicating potentially reversible ischemia around the perfusion defect.

MRI is more sensitive than CT in detecting ischemic changes, similar to CT with respect to diagnosing hemorrhage, and it remains the gold standard for imaging in acute focal neurological deficits . Multimodal MRI (different imaging sequences) provides detailed visualization of brain and vessel anatomy, including perfusion data with a higher sensitivity and specificity than CT to detect stroke mimics . The complete MR imaging protocol requires approximately 15 to 20 minutes to perform, and an additional 10 minutes for patient positioning and transfer, as well as calculation of the apparent diffusion coefficient and mean transit time maps . MRI comes with an advantage of decreased exposure to radiation with higher costs, lesser availability, claustrophobia experienced by patients as a drawback, along with absolute contraindications such as metallic implants or pacemakers.