AJR 2003; 180:725-728
© American Roentgen Ray Society
Combining Two Single-Shot Imaging Techniques with Slice-Selective and Non-Slice-Selective Inversion Recovery Pulses: New Strategy for Native MR Angiography Based on the Long T1 Relaxation Time and Inflow Properties of Blood
Matthias Braendli1 and
Georg Bongartz
1 Both authors: Departement of Radiology, Institute for Diagnostic Radiology,
University Hospitals, Kantonsspital Basel, Petersgraben 4, 4031 Basel,
Switzerland.
Received May 8, 2002;
accepted after revision August 27, 2002.
Address correspondence to M. Braendli.
We propose a new strategy to obtain both long T1 weighting (bright signal
of long T1 tissue) and inflow-enhanced contrast as the basis of a native
two-dimensional bright blood MR angiography. For image acquisition, we used a
two-dimensional single-shot technique (true fast imaging with steady-state
free precession [trueFISP] [1]
and fast low-angle shot [FLASH]
[2]) in combination with
preceding inversion pulses (180° prepulses). Image processing included a
subtraction method. The proposed protocols have been tested successfully to
image the pulmonary vasculature in healthy subjects (Fig.
1A,
1B,
1C,
1D).
View larger version (143K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. —Native pulmonary MR angiography of 35-year-old healthy man.
Maximum-intensity-projection image of 20 two-dimensional coronal subtraction
images (D) shows bright signal of blood. Resulting contrast is long T1
weighted. Note bright signal of gastric fluid (arrow).
|
|
View larger version (110K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. —Native pulmonary MR angiography of 35-year-old healthy man.
True fast imaging with steady-state free precession (trueFISP) image without
black blood preparation (double inversion pulse) (matrix, 192 x 256;
TR/TE, 576/1.7; coronal slices, 20; slice thickness, 6 mm) shows bright signal
of blood and fat. Image contrast depends on T2/T1. There is no inflow
enhancement because trueFISP is not flow sensitive.
|
|
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C. —Native pulmonary MR angiography of 35-year-old healthy man.
TrueFISP image with preceding black blood preparation (double inversion pulse)
(matrix, 192 x 256; 576/1.7; coronal slices, 20; slice thickness, 6 mm)
shows accentuated short T1 weighting. Note dark signal of blood and
still-bright signal of fat (mediastinal and thoracal wall).
|
|
View larger version (102K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1D. —Native pulmonary MR angiography of 35-year-old healthy man.
Subtraction of trueFISP images with black blood preparation from trueFISP
images without black blood preparation (matrix, 192 x 256; 576/1.7;
coronal slices, 20; slice thickness, 6 mm) shows bright signal of blood.
Because trueFISP is practically not flow sensitive, signal enhancement is
mainly due to long T1 contrast. There is no inflow enhancement as in Figures
4C and
5C.
|
|
View larger version (80K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4C. —Native pulmonary MR angiography of 31-year-old healthy man.
Subtraction of non-slice-selective turbo FLASH from slice-selective turbo
FLASH image shows bright signal of blood. Bright signal is due to T1 weighting
and inflow enhancement. Mediastinal fat and fat of thoracal wall are
subtracted.
|
|
View larger version (67K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5C. —Native pulmonary MR angiography of 27-year-old healthy man.
Inversion recovery-prepared trueFISP image (matrix, 192 x 256; 800/1.8;
inversion time, 400 msec; axial slices, 20; slice thickness, 6 mm) shows that
resulting contrast depicts long T1 weighting and inflow enhancement without
short T1 weighting. Note subtracted mediastinal and subcutaneous fat.
|
|
An inversion recovery pulse can be combined with almost any scanning method
[3]. The combination with a
single-shot technique (trueFISP and turbo FLASH) allows for fast breath-hold
imaging. Thereby, the data acquisition occurs while the longitudinal
magnetization relaxes back to equilibrium
[4]. The contrast is given by
the value of the longitudinal magnetization when the central k-space lines are
acquired. The resulting contrast of an inversion recovery single-shot sequence
is both short T1- and long T1-weighted images
[3]. Blood has a relatively
long T1 time (
1200 msec at 1.5 T) and is therefore bright in inverse T1
weighting. An ideal approach to selective bright blood imaging should be to
keep the long T1 weighting of an inversion recovery sequence but to eliminate
the short T1 weighting (e.g., the unwanted bright signal of fat). In this
context, a subtraction technique is helpful. To this purpose, a second data
set was acquired, which was mainly short T1-weighted. Then, this second data
set was subtracted from the first data set. The resulting contrast showed the
desired long T1 weighting.
This mainly short T1 weighting of the second data set was achieved as
follows: whenever a tissue is stimulated by subsequent 180° pulses at
identical intervals, the system gradually arrives at a steady state. Thereby,
the longitudinal magnetization in the steady state just after the inversion
pulse (Mo') depends on the T1 constant of the tissue stimulated and can
be inferred from the equation (Fig.
2):
where P denotes the regular pulse interval. Solving the equation for
Mo' yields
View larger version (13K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2. —Graph shows relaxation curve of various tissues with
different T1 relaxation times (200, 400, and 800 msec) in steady state (after
about five inversion pulses). Mo'= longitudinal magnetization in steady
state just after inversion pulse. PI = pulse interval of inversion pulse (800
msec in example).
|
|
This relationship indicates that Mo' depends on T1; the shorter T1,
the larger -Mo' (Fig. 3).
Because signal is proportional to Mo', the resulting signal will be
additionally T1-weighted.
View larger version (7K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3. —Graph shows longitudinal magnetization in steady state just
after inversion pulse (-Mo') as function of T1 relaxation time for given
regular pulse interval of 800 msec. Shorter T1 means larger -Mo'.
|
|
Therefore, the following feasible procedure results in getting long T1
weighting: First, a single-shot technique (turbo FLASH, inversion recovery
trueFISP) with slice-selective inversion pulses is performed (one slice per
shot). This yields an image set that is both short T1- and long T1-weighted as
mentioned previously (Figs. 4A
and 5A,
5B,
5C). Because the inversion
pulses are slice-selective, they affect the currently acquired slice without
affecting the slices to be acquired later. Furthermore, inflow of unexited
blood into the excited slice results in signal enhancement. The signal of
blood in the slice-selective images is, therefore, long T1 and short T1, as
well as inflow-enhanced (depending on flow velocity and slice thickness).
View larger version (91K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A. —Native pulmonary MR angiography of 31-year-old healthy man.
Inversion recovery snapshot fast low-angle shot image ([turbo FLASH] Siemens,
Erlangen, Germany) (matrix, 96 x 128; TR/TE, 527/1.5; inversion time,
400 msec; axial slices, 20; slice thickness, 6 mm) shows slice acquisition
after slice-selective inversion pulses. Resulting contrast shows short T1
weighting, long T1 weighting, and inflow enhancement.
|
|
View larger version (82K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A. —Native pulmonary MR angiography of 27-year-old healthy man.
Inversion recovery-prepared true fast imaging with steady-state free
precession (trueFISP) image (matrix, 192 x 256; TR/TE, 800/1.8;
inversion time, 400 msec; axial slices, 20; slice thickness, 6 mm) shows slice
acquisition after slice-selective inversion pulses. Resulting contrast shows
short T1 weighting, long T1 weighting, and inflow enhancement.
|
|
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B. —Native pulmonary MR angiography of 27-year-old healthy man.
Subtraction of non-slice-selective inversion recovery-prepared trueFISP from
slice-selective inversion recovery-prepared trueFISP image shows bight signal
of blood, which is due to long T1 weighting and inflow enhancement. Note
subtracted mediastinal and subcutaneous fat.
|
|
Subsequently, the same sequence with identical parameters is repeated, this
time using non-slice-selective inversion pulses that are pulsed at regular
intervals. Now the non-slice-selective inversion pulses affect not only the
currently acquired slice but also the slices to be acquired later. Therefore,
these pulses cause the system to arrive gradually at a steady state. The
non-slice-selective pulses must be continued to maintain the steady state.
Otherwise longitudinal magnetization will recover, and the steady state will
not be maintained. To bring the system close to the steady state before
imaging, approximately the first five inversion pulses are performed without
data acquisition; the subsequent pulses are followed by data acquisition. The
resulting images show an accentuated short T1 contrast (Figs.
4B and
5B). Because the pulses are
not slice-selective, there is no inflow dependence.
View larger version (140K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B. —Native pulmonary MR angiography of 31-year-old healthy man.
Turbo FLASH image (matrix, 96 x 128; 527/1.5; inversion time, 400 msec;
axial slices, 20; slice thickness, 6 mm) shows slice acquisition after
non-slice-selective inversion pulses. Note dark signal of blood and still
bright signal of fat (mediastinal and thoracal wall).
|
|
By subtracting the image set 2 from the image set 1, one can eliminate the
short T1 weighting and retain nothing but the long T1 weighting and inflow
enhancement of the image set 1 (Figs.
4C and
5C).
We tested the feasibility of the proposed strategy in three healthy male
volunteers who were 27-35 years old. Informed consent was obtained before the
examination. All examinations were performed on a 1.5-T MR imaging system
(Symphony; Siemens, Erlangen, Germany) equipped with a quantum gradient
system. A phased array body coil was used for all MR imaging procedures. On
the assumption that the individual heart cycle was approximately constant in
each of our healthy subjects, we used the ECG signal to trigger the inversion
pulses and the subsequent single-shot acquisitions (one slice per shot).
Hence, the inversion pulses were pulsed at approximately the same interval
that was sufficient to maintain the desired steady state in the case of
non-slice-selective inversion pulses. On the other hand, all slices were
acquired at the same phase of the heart cycle, which reduced motion artifacts.
ECG-triggering may not be practicable for imaging patients with arrhythmia
with the probable disadvantage of motion-flow artifacts. Imaging was performed
at end expiration for better reproducibility of the diaphragm position as a
condition for optimal subtraction. The subtraction of the non-slice-selective
short T1-weighted images from the slice-selective short T1- and long
T1-weighted images gave a resulting contrast that was long T1-weighted. In
addition, inflow of unexited blood into the slice resulted in signal
enhancement in the case of slice-selective pulses as mentioned previously. The
signal of blood in the slice-selective images was therefore all long T1-,
short T1-, and inflow-weighted (depending on flow velocity and slice
thickness). Hence, the subtraction images showed an intensely bright blood
contrast, whereby the long T1 of blood and flow contributed to signal
enhancement. Theoretically, a thrombus (slow flow, T1 shortening) should be
visible as a signal void.
Finally, we add the following modification of the suggested protocol: the
combination of trueFISP and black blood trueFISP. Instead of the protocol
using inversion recovery prepared trueFISP or turbo FLASH sequences
respectively, this modified protocol yields mainly T1-weighted images without
inflow enhancement (Figs. 1A,
1B,
1C,
1D).
Black blood preparation is an effective feature of the Siemens'
applications to suppress blood signal. The preparation is made of a double
inversion pulse once slice-selective and once non-slice-selective. The two
inversion pulses follow each other immediately. When applying regularly pulsed
black blood preparations, we found that the non-slice-selective components
produce a steady state of longitudinal magnetization as mentioned previously.
The second slice-selective inversion pulse inverts the longitudinal
magnetization in the currently acquired slice back to positive values
(Fig. 2). The resulting images
are again short T1-weighted because Mo' depends on T1, and signal is
proportional to Mo'. By subtracting this image set
(Fig. 1C) from a simple
trueFISP sequence (Fig. 1B), we
found that the resulting images are long T1-weighted but without inflow
phenomena because trueFISP is not flow sensitive
[3] (Figs.
1A and
1D).
The three proposed protocols (by means of inversion recovery trueFISP,
turbo FLASH, and trueFISP with trueFISP with black blood preparation) have
been tested successfully in healthy subjects for the particular case of native
pulmonary MR angiography. We believe the same technique will also apply to
other vascular areas.
References
- Oppelt A, Graumann R, Barfuss H, et al. FISP: eine neue schnelle
Pulssequenz für die Kernspin-tomographie.
Electromedica
1986;54:15
-18
- Haase A, Matthaei D, Bartkowski R, et al. Inversion recovery
snapshot FLASH MR imaging. J Comput Assist Tomogr
1989;13:1036
-1040[Medline]
- Vlaardingerbroek MT, den Boer JA. Magnetic resonance
imaging. New York: Springer, 1999:339
-344
- Scheffler K, Hennig J. T(1) quantification with inversion recovery
trueFISP. Magn Reson Med
2001;45:720
-723[Medline]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Top
References
