Purpose: The purpose of this work is to obtain a fast method for T1 relaxation time measurements
at 3 Tesla.
Methods: The schematic pulse sequence diagram shown in figure 1 is implemented on a Philips
3T intera scanner. The method is a variation on the methods published by Ordigde et
al. [1] and Clare & Jezzard. [2].
Figure 1
After a global inversion pulse a single shot EPI sequence with 60 slices is used to
sample the inversion recovery curve. The first slice is acquired at TI=20 ms after
the inversion pulse. The slice to slice timing DT=60 ms. At the next repetition of
the sequence, the order in which the slices are acquired is shifted 5 slices (now
starting with slice 6 and finishing with slice 5) to obtain a second sample point
on the recovery curve for each slice. The shifting procedure was repeated 12 times
resulting in 12 sample points per slice. Other imaging parameters were: TR=10s, TE=26
ms, matrix=128×128, FOV=220mm (4 subjects) / 256mm (3 subjects), SENSE-factor=2.2
and flip angle=90 degrees. The total acquisition time for 60 slices of 2mm was 140s.
T1 parameter-maps were measured in seven subjects after written informed consent was
obtained. After temporal reordering of the slices, the data was fitted to I(t)=I0(1–2exp(-t/T1)+exp(-TR/T1)) to obtain T1-values on a pixel by pixel basis. A separate
T1 weighted anatomical scan was used to define regions of interest containing gray
matter, white matter and CSF in each subject separately. The anatomical scan was also
segmented using the FSL segmentation tool [3]. Both ROIs and segments were used to
obtain average T1-values.
Results: Figure 2 shows two slices from a T1 parameter-map obtained from one of the subjects.
The calculated average T1 values are summarized in table 1 together with values from
the literature.
Table 1
|
N=7
|
T1 (ms) ROIs
|
T1(ms) segments
|
T1 (ms) [2]
|
T1 (ms) [4]
|
|
GM
|
1421±106
|
1318±106
|
1160
|
1331
|
|
WM
|
780±25
|
830±42
|
860
|
832
|
|
CSF
|
4043±361
|
2761±578
|
3700
|
-
|
Figure2
Conclusion: The average T1-values are in close agreement with the values published in the literature.
The differences between the ROI and segment based averages can be explained by partial
volume effects. In conclusion, a method is implemented at 3 Tesla that provides T1
relaxation times in less than 2,5 minutes.
References:
[1] R. Ordidge et al., MRM 16:238–245, 1990.
[2] S. Clare and P. Jezzard, MRM 45:630–634, 2002.
[3] http://www.fmrib.ox.ac.uk/fsl/
[4] J. Wansapura et al., JMRI 9:531–538, 1999.
