2 Protocol and Assessment of MR Perfusion

140



J.-W. Kang and S.M. Ko



Step 1



Step 2



Step 3



Rest scan

interpretation



Image processing



Quality assess



• Coronary artery

stenosis and plaque

analysis



• Best phases of

motionless

myocardium

• Epi-and

endocardialcontour

(for dynamic study)



• Motion artifact

• Beam hardening

• Cone-beam

• Stair-step

• Image noise



Step 4



Image interpretation



• Transmurality (≥50 % or

<50 %)

• Reversibility

• Myocardial thinning

• Simultaneous vision of

both stress and rest scans



Fig. 10.5 Diagram of the flow chart of qualitative assessment of CT perfusion study



a



c



Fig. 10.6 (a) Dynamic perfusion

scan [(a) and QR code at

Fig. 10.2] and the derived

myocardial blood flow (MBF)

map show the impaired MBF of

the inferior wall of the left

ventricle (arrow) (b). CT

coronary angiography also shows

the severe stenosis of the right

coronary artery (arrow) (c)



b



Step 5

Correlation with rest

scan



Match perfusion defect

and coronary artery

lesion



10



Evaluation of Myocardial Ischemia Using Perfusion Study



Fig. 10.7 Diagram for perfusion

MR study



141



Contrast bolus injection

0.03–0.1 mmol/kg



Start adenosine



4–5 min



Contrast bolus injection

0.03–0.1 mmol/kg



Stop adenosine



Stress

perfusion



12–15 min interval



survey



• Patient preparation

– Patients are advised to avoid caffeine, a nonselective

competitive adenosine receptor antagonist, 24 h before

examination.

– Intravenous access is performed in both antecubital

veins: one for adenosine or other vasodilator infusion

and one for the contrast administration.

– The scan protocol comprises a stress- and a rest-phase

acquisition. Since stress-first-and-rest-second protocol

has the advantage of increased sensitivity of myocardial

ischemia on stress-phase scan, this “stress-first” scan is

usually performed on MR stress perfusion study.

– More than 10 min time interval between two acquisitions is necessary. When the time interval is short, the

contrast used in the first phase may still remain in the

myocardium at the time of the second acquisition,

which may decrease the sensitivity for detecting myocardial ischemia and infarction.

• Pulse sequences

– Most sequences are based on T1 contrast enhancement

with magnetization preparation (inversion or saturation recovery).

– Spoiled gradient echo (TurboFLASH, turbo fast-field

echo, and GRASS) is widely used: the gradient echo

image acquisition with short TR and TE and magnetization preparation. The typical parameters are TR/TE (ms)

of 3/1, flip angle of 15°, 2-dimensional multisection, section thickness of 8–10 mm, bandwidth of 600–800 Hz

per pixel, nonsection-selective saturation recovery, and

image acquisition time of 150–200 ms per section.

– Steady-state free precession (TrueFISP, balanced turbo

field echo, turbo FIESTA) is also used for the MR perfusion study; the typical parameters are TR/TE (ms) of

2/1, flip angle of 40°, 2-dimensional multisection, section thickness of 8–10 mm, bandwidth of 1,000–

12,000 Hz per pixel, nonsection-selective saturation



Viability



Cine, etc



Continuous adenosine infusion

140 ug/kg/min



10.2.1 Protocols



Rest

perfusion



5 min interval



30 min



recovery, and image acquisition time of 130–160 ms

per section. It has higher contrast-to-noise ration than

that of spoiled gradient echo sequence.

– Hybrid echo planar image and gradient echo sequence

are recently introduced. This sequence has the advantage

of shortest image acquisition time than other sequences.

• Acquisition of MR perfusion

– Cardiac localization is performed for defining imaging

plane. Three or four short-axis planes are used for the

perfusion study.

– For the stress perfusion imaging, intravenous adenosine infusion at the rate of 140 μg/kg/min is performed,

and intravenous gadolinium contrast media of 0.03–

0.1 mmol/kg is delivered at the rate of 3–5 mL/s after

4–5 min from start of adenosine infusion. Twenty milliliters of saline chaser at the rate of 3–5 mL/s is

followed.

– For the rest scan, intravenous gadolinium contrast

media of 0.03–0.1 mmol/kg is delivered at the rate of

3–5 mL/s without adenosine infusion. 20 mL of saline

chaser at the rate of 3–5 mL/s is followed. Usually the

time interval between the stress and the rest scan is

between 12 and 15 min.

– Dynamic scans for 3–4 short-axis planes are usually performed during both stress and rest scans.

Approximately 40–60 serial scans from the injection of the gadolinium contrast media are performed

in every other heartbeat. Therefore, 40–60 images of

each short-axis plane are to be acquired (Fig. 10.7).



10.2.2 Assessment of MR Perfusion

10.2.2.1 Qualitative Assessment

• Simultaneous visualization of both rest and stress

images for regions with hypo-intense myocardium compared with normal myocardium is necessary (see

Sect. 10.3).



142



J.-W. Kang and S.M. Ko



a



b



c



Fig. 10.8 Visual assessment of perfusion defect. Perfusion defect of the anterior wall is seen persistently from early phase (a), to mid phase (b)

and late phase (c)



• Playing images in cine mode is essential for differentiating

between image artifact such as dark-rim artifact and the

true perfusion defect (see Sect. 10.4). Dark-rim artifacts

typically occur in a couple of frames during peak contrast

enhancement of the blood pool in the left ventricle and

before peak contrast enhancement in the myocardial tissue.

True perfusion defect is persistent and more prominent during the peak contrast enhancement in the myocardial tissue.

• Standard 17-segmental model of the left ventricular myocardium suggested by the American Heart Association is

used for the location and scoring of the myocardial perfusion status.

• Each myocardial segment is scored for the presence or

absence of the perfusion defect and graded as transmural

if the perfusion defect involves ≥50 % of thickness or

non-transmural. Reversibility is also graded as reversible,

partially reversible, and irreversible or fixed.

• To ensure the perfusion defect is detected, images from

multiple phases must be reviewed. Motion artifacts and

beam-hardening artifacts can mimic perfusion defect (see

Sect. 5.1 of this chapter) (Fig. 10.8).



10.2.2.2 Quantitative Assessment

• Myocardial blood flow and myocardial blood volume can

be derived by the time-intensity curves (TICs) of the



•



•



•



•



myocardium, the left ventricular cavity, and the aorta

using the dynamic CT perfusion study.

Drawing of endo- and epicardial border of each image in

cine acquisition is required for the quantitative analysis.

Blood pool in the left ventricle and epicardial fat should

be excluded.

Standard 17-segmental model of the left ventricular myocardium suggested by the American Heart Association is

used for the location and scoring of the myocardial perfusion status.

Maximal upslope, upslope, time-to-peak, maximum

signal intensity, and myocardial perfusion reserve index

are introduced to the semiquantitative parameter for the

myocardial perfusion status (Fig. 10.9).

Myocardial blood flow may be used for the myocardial

blood flow and volume.



10.3



Representative Cases of CT Perfusion

and MR Perfusion



10.3.1 One-Vessel Disease

Figure 10.10



10



Evaluation of Myocardial Ischemia Using Perfusion Study



Fig. 10.9 Diagram of relative upslope (RU) for myocardial perfusion

reserve index (MPRI) using the time-intensity curve



143



MR signal



RMU=

Blood pool



MU

S0

x 100 %

LVMU

LVS0



LVMU

Myocardial signal

MU

LVS0

S0

TO



a



TTP



Time



b



c



Fig. 10.10 (a) CT angiography of RCA in rest scan shows >70 % stenosis at the PL (arrow). (b) Stress perfusion CT study shows transmural

perfusion defect at the mid-inferior wall (arrows). (c) Rest scan of CT



d



shows reversibility of perfusion defect. MR stress (d) or rest (e) scan also

shows the same perfusion defect pattern of the inferior wall. (f) Coronary

angiography shows severe total occlusion of proximal PL (arrows)