Acute Ischemic Stroke : Infarct Core Estimation On Ct Angiography Source Images Depends On Ct Angiography Protocol (Page 9 of 12) in pdf

NEURORADIOLOGY:

Infarct Core Estimation on CT Angiography Source Images Depends on Protocol

Pulli et al

Figure 5

Figure 5: Bland-Altman plots of interrater agreement for CT angiography source images (CTA-SI) according to protocol show that (a) protocol

1 resulted in good agreement, with a mean difference of 2 2.1 mL and narrow limits of agreement (6.7, 2 11.0 mL), while with (b) protocol 2,

the mean difference was 2 7.9 mL, with limits of agreement of 16.7, 2 32.4 mL. Dotted line = line of equality.

that involve craniocaudal imaging ( 15 ),

a situation where blood ﬂ ow via collat-

protocol that is empirically validated to

a shortened delay time ( 16 ), and in-

eral vessels is delayed. Insufﬁ cient de-

provide a good estimate of the infarct

creased table speed ( 32 ).

lay time is also the most likely expla-

core be established at each institution.

Our ﬁ nding that evaluation of acute

nation for the ﬁ ndings of Sharma et al

Unfortunately, there are trade-offs

ischemia with CT angiography source

( 18 ), who concluded that CT angiogra-

between vessel and parenchymal imag-

images is protocol dependent reﬂ ects

phy source images appear to be CBF

ing with CT angiography. The primary

the pathophysiology of ischemic stroke.

weighted instead of CBV weighted. In

goals of our CT angiography protocol

While DW imaging images the differ-

their study, they used a 5–10-second

optimization were to speed evaluation

ences in Brownian motion of protons in

delay time, which, according to our

and to improve visualization of the in-

water, CT angiography source images

data, would be too short. In contrast,

tracranial arteries, allowing for better

provide an approximation of CBV un-

Wittkamp et al ( 33 ) triggered CT image

characterization of vessel occlusions,

der the assumption of a steady state be-

acquisition at peak enhancement of the

stenoses, and aneurysms. The longer

tween arterial and parenchymal contrast

superior sagittal sinus, ensuring sufﬁ -

delay required for parenchymal eval-

material ( 13,14 ). Attenuation values of

cient delay times. Furthermore, they

uation would prevent vessel opaciﬁ ca-

brain tissue on CT angiography source

performed CT angiography after perfu-

tion in the early arterial phase. To solve

images are directly proportional to the

sion CT, so that contrast material from

this problem, two protocols could be

amount of contrast material that has

the perfusion CT study had enough time

implemented, one performed early for

arrived within the parenchyma at the

to reach the ischemic bed by the time

optimal visualization of the intracranial

time of imaging. When a proximal ce-

of CT angiography image acquisition.

arteries, and a second performed with

rebral artery is occluded, the affected

As a result, they found good correla-

an appropriate delay to evaluate the

territory is supplied by the collateral

tion between CT angiography source

brain parenchyma. Our data suggest

circulation, prolonging contrast mate-

images and CBV. Results of initial studies

that a delay of at least 40 seconds may

rial arrival time even in the setting of

( 10,11 ) demonstrating a close approxi-

be sufﬁ cient. While we found an under-

sufﬁ cient blood ﬂ ow. Earlier CT angi-

mation in infarct size between CT an-

estimation of DW imaging volumes us-

ography image acquisition prevents

giography source images and DW im-

ing such a delay, the differences were

contrast material from traversing the

ages also likely used a sufﬁ cient delay

relatively small, and there remained

collateral vessels and reaching the dis-

time. The fact that this issue was not

a strong correlation between the two

tal bed, thereby increasing the area

detected earlier probably relates to the

techniques such that CT angiography

of hypoattenuation. This explains why

inability of older CT scanners to image

source images acquired with this delay

time to imaging was an independent

at speeds similar to those of current

provided a reasonable approximation of

predictor of volume overestimation on

scanners. On the basis of our ﬁ ndings,

the infarct core.

CT angiography source images in this

we suggest that when CT angiography

Our study limitations included its

study. Similarly, atrial ﬁ brillation, as a

source images are used to evaluate the

retrospective design. Therefore, there was

surrogate for low cardiac output, creates

parenchyma during ischemic stroke, a

a shorter time between CT angiography

601

Radiology: Volume 262: Number 2—February 2012

Acute Ischemic Stroke : Infarct Core Estimation On Ct Angiography Source Images Depends On Ct Angiography Protocol Page 9

Related Articles

Related forms

Related Categories