Lawrence N. Tanenbaum, MD

Readily available in most settings, computed tomography maintains its dominant role in the evaluation of patients with acute stroke. Despite the robust capability of magnetic resonance imaging, CT is the primary modality utilized in the triage of patients in whom thrombolytic intervention is contemplated. CT angiography (CTA) and perfusion CT have augmented the utility of CT in this setting, offering information useful in therapeutic planning.

Decisions about the appropriateness of thrombolytic therapy in patients with acute stroke are largely based on the findings on unenhanced CT (Figure 1). CT assessment remains the “gold standard” for the presence of hemorrhage. Despite the superior sensitivity of diffusion MRI (DW MRI) to the presence of acute stroke, lack of immediate, 24-hour availability limits utilization in most clinical settings. Due to importance in therapeutic decision-making, it is critical that the reader is sensitive to the often-subtle changes that characterize acute infarction on CT (Figures 2-4).


CT angiography is increasingly employed in the evaluation of patients with acute stroke to characterize the presence and level of vascular thrombosis. 1 At many institutions, a central arterial occlusion will be treated with intra-arterial thrombolytic therapy. Patients with CTA studies that are normal or that indicate the presence of a peripheral branch occlusion will be treated intravenously or not at all.

Patients are injected with 50 cc of a high-concentration contrast agent (eg, 370-400) at rates of 2.5 to 4 cc/sec, well tolerated in the setting of acute stroke. Acquisition protocols vary by generation of scanner (Figure 5-7). CTA studies are feasible at all hours using modern scanner operator interfaces that allow rapid, semiautomated, online creation of overlapping limited volume maximum intensity projections (OLIVE MIP) also known as sliding thin slab MIP, or MPVR (Figure 8).

The scanning technologist can render these images in minutes, without close physician supervision and without interrupting the flow of patient scanning in a busy department. Automated volume rendering engines available with some CT systems add ready access to three-dimensional depiction (Figure 9).

Figure 3. Acute stroke. Acute (left) and follow-up CT of the head in a patient with right hemispheric ischemic symptoms. Note the subtle obscuration of the anterior portion of the right sylvian cortex and lentiform nucleus that becomes clearer on the follow-up study.

Perfused blood volume (PBV) studies can be created at the scanning console from the source data obtained for the CT angiogram. CTA source images are reformatted (if necessary based on method of CTA acquisition) into 3-mm slices at 1.5-mm intervals. The processed images are then viewed at windows and levels typical for routine head CT (eg, 90 WW 40 WL). These images of the “blood pool” are more sensitive to the presence of infarcted brain tissue than unenhanced CT (Figure 10) and correlate better with eventual infarct size. 2

Map of Brain Perfusion

Figure 4. Immediate (left) and follow-up CT studies in patient with right hemispheric symptoms. Note the indistinctness and loss of density of the right lentiform nucleus and adjacent opercula, which become evident as infarction on the delayed study.

Perfusion studies are obtained by monitoring the first pass of a standard high-concentration iodinated contrast agent (370-400 mg I/mL) through the cerebral vasculature (Figure 11). The contrast bolus causes a transient rise in attenuation proportionalto the amount of tracer in a given region (Figure 12). Integration of data over the first pass of the contrast agent allows creation of maps of brain perfusion. Commercially available perfusion CT software provides information about parameters such as cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT) in a clinically feasible time frame. Determination of CBV is augmented by normalization for the pixel value of venous blood. The MTT calculation uses a complex deconvolution algorithm to adjust for the arrival time of arterial blood (Figure 13). The net result is accurate, xenon equivalent calculation of CBF (CBV/MTT) processed in a semiautomated fashion directly on the scanner console. 3,4

Perfusion techniques are utilized for the evaluation of acute and subacute stroke, offering the most sensitive measure of the extent of brain tissue under ischemic conditions and the best assessment of tissue viability.

In the acute setting, the deficit on a perfusion study is often greater than that seen on studies such as unenhanced CT, PBV CT, and diffusion MR imaging. A reduction in CBF is a typical accompaniment of acute stroke. Depending on severity, this will manifest as a compensatory increase or a resultant decrease in CBV, as well as a regional prolongation of MTT and time to peak (TTP) (Figures 14, 15). Outside of the setting of acute stroke, perfusion imaging can yield useful information about the functionality of collateral circulation and thus the significance of vascular occlusive disease (Figure 16).

Figure 8. Overlapping limited volume MIP renderings. Serial OLIVE MIP images demonstrate the presence of a left MCA occlusion with moderate filling of collaterals.

Quantitative assessment of CBF yields information critical to assessing brain tissue viability and hemorrhagic risk, critical in thrombolytic therapy decision-making. Subtracting the volume of brain with restricted diffusion from the perfusion indicated volume of tissue under ischemic conditions yields the commonly accepted MR paradigm for tissue at risk for extension of infarction. When commercially available, sodium imaging may supplant diffusion as the best marker of infarcted tissue. The CT parameter that best estimates infarction volume is less clear. The deficit on PBV CT or the region of uncompensated reduction in CBV may be the best CT estimate of the volume of infarcted brain.


Figure 9. Acute left hemispheric stroke. Unenhanced CT (left) reveals a hyperdense left MCA. Volume rendering (center) and OLIVE MIP (right) confirm the presence of a left ICA and MCA occlusion with poor collateralization.

CT maintains its preeminent role in the evaluation of patients with symptoms suggesting acute stroke. CT angiography, perfused blood volume CT, and first pass perfusion techniques have significantly augmented the role of CT for these patients, offering structural and functional information critical in therapeutic decision-making. The multitasking, automated functionality of modern, user-friendly scanner interfaces facilitate performance of advanced postprocessing techniques even in the busiest of settings.

Figure 10. Perfused blood volume. Same patient as in figure 9-left hemisphere ischemic stroke. Unenhanced CT (left) is unremarkable with a well-defined lentiform nucleus and sylvian cortex. Perfused blood volume study (right) reveals relative lack of enhancement of the lentiform nucleus.
Figure 11. Perfusion imaging technique.
Figure 12. First pass perfusion. Time intensity curves reveal passage of arterial (A) and venous (V) contrast as well as parenchymal enhancement (P).
Figure 13. Normal perfusion CT study.
Figure 14. a,b. Acute right hemispheric ischemia. a) Axial OLIVE MIP images reveal a right MCA segmental stenosis. b) PCT study manifests a corresponding reduction in CBF with prolonged MTT and slightly reduced CBV.
Figure 15. a,b,c,d. Acute right hemispheric ischemia. a) OLIVE MIP and volume rendered images reveal a right MCA occlusion. PBV study is normal. b) Perfusion study reveals ischemia in the right MCA territory. Note the reduction in CBF with prolongation of MTT and mostly preserved CBV. c) Careful quantitative assessment of CBF reveals reduction of flow below the viability threshold in the region of the right lentiform nucleus with maintenance of adequate levels elsewhere. d) Follow-up study (right) after intra-arterial thrombolysis manifests infarction of the lentiform nucleus with preservation of remainder of MCA territory.
Figure 16. Functional significance of vascular disease. Volume rendering and OLIVE MIP of a patient with a right ICA occlusion and significant stenosis of the left ICA and M1 segments (left). PCT (right) reveals reduced CBF in both MCA territories, worse on the left. MTT is prolonged in both MCA territories, while CBF is preserved aside from a reduction in the left frontal lobe at a site of old infarction.

Lawrence N. Tanenbaum, MD, is Section Chief CT, MR, and Neuroradiology, New Jersey Neuroscience Institute, Seton Hall School of Graduate Medical Education, JFK Medical Center, Edison, NJ.


  1. Tanenbaum LN, Verro P, Borden NM, Sen S, Eshkar NE. The role of CT angiography in acute ischemic stroke: a prospective study. Poster American Academy of Neurology 2000.
  2. Verro P, Tanenbaum LN, Fonzetti P, Borden NM, Sen S, Eshkar NE. CT perfusion deficits in acute cerebral ischemia predict brain at risk. Presentation American Academy of Neurology 2000.
  3. Cenic A, Nabavi DG, Craen RA, Gelb AW, Lee TY. Dynamic CT measurement of cerebral blood flow: a validation study. AJNR Am J Neuroradiol. 1999;20:63-73.
  4. Nabavi DG, Cenic A, Craen RA, et al. CT assessment of cerebral perfusion: experimental validation and initial clinical experience. Radiology. 1999;213:141-149.