Steady state free precession (SSFP) signal has multiple signal components. The FID (F0) and the spin echo (F-) components can be used to measure T2 ¹. It was also demonstrated that the various components could be acquired within the same TR ². Triple Echo Steady-State (TESS) used this technique to acquire the F0, F- and the echo shifted (F+) components within one TR for T1 and T2 relaxometry ³.

Acquiring multiple echoes in one TR is unfavorable to signal components that decay over more than one TR cycle before acquisition (such as F- and F+). We propose here a novel SSFP sequence that acquires the F0, F- and F+ component in an interleaved way for T1 and T2 relaxometry.

Figure 1 shows the 3D iTESS sequence. The unbalanced SSFP sequence has a net crusher moment Q (≠ 0) over every TR. F0, F- and F+ were acquired in an interleaved way. The gradients in the phase encoding (PE) and readout (RO) direction were balanced to reduce motion sensitivity. Unwanted components were dephased by designing appropriate crushers applied to any direction but the slice select (SS) direction is preferred due to its lowest resolution ⁴. Let A = the SS gradient moment, the crusher moments are (Figure 1):

F+: B=-Q–A/2, C=2*Q-A/2 [1]

F0: D=-A/2, E=Q-A/2 [2]

F-: F = Q-A/2, G = -A/2 [3]

TE was identical among the three echoes to eliminate the T2* decay effects. The readout bandwidths were also identical. The sequence was implemented on a 3T MRI scanner (TIM TRIO, Erlangen, Germany).

Four tubes of MnCl₂ solution with concentration 0.125/0.25/0.375/0.5 mM were fixed in an agar phantom. Using IR-TSE and SE, the T1 values in ms were found to be 654.9±15.7, 418.8±28.2, 295.5±9.3, 233.9±15.6, the T2 values in ms were 41.3±0.84, 28.2±0.65, 20.0±0.59, 15.6±0.35 respectively. iTESS was used to image the phantom and acquired the three signals using different FAs. The measured signals were compared to the analytical results. Imaging parameters used were: TR/TE = 8/4 ms, bandwidth = 723 Hz/pixel, matrix size = 256*152*64, resolution = 0.83*0.83*3 mm³; FA was varied between 5^o~40^o.

The signals of iTESS were validated against their corresponding analytical equations ⁵ using one tube with T1/T2 = 418.8/28.2 ms. The signals dependence on FAs were compared to the corresponding signals measured from the sequence. T1 and T2 relaxometry were then performed. Here, the signals acquired at a flip angle of 10^o/15^o were used. The T1 and T2 maps were found using the golden section search algorithm ³ implemented in MATLAB (Mathworks, Natick, USA).

Figure 2 shows the phantom images from the F0, F- and F+ components at a flip angle of 10^o. Figure 3 shows how the analytical solutions for the three signal components compared to the experimental results at various FAs. The analytical solutions matched the experimental SSFP signals very well.

The T1 and T2 maps computed from the signals acquired using iTESS are shown in Figure 4. Figure 5 summarizes how the values compared to the reference standards. The slightly higher standard deviations of T1 and T2 values derived from iTESS were likely due to its low SNR.

In this study, a novel SSFP sequence with interleaved acquisition for F0, F- and F+ was proposed and applied to T1/T2 relaxometry. The good agreement in Figure 3 shows that signals acquired from iTESS are described by the corresponding signal equations, implying that steady state is maintained when crushers were placed at different TR cycles according to Eq.[1-3]. Though iTESS takes longer time to image than TESS, it has several advantages over TESS. First, the TEs of the three echoes are identical. The T2* decay of the three echoes can be canceled out. In TESS, the T2* effect cannot be ignored since the three echoes have different TEs. Second, iTESS has a shorter TR. As the effective TE of F+ is (nTR + TE), the short TR decreases susceptibility induced signal void. Short TR also reduces motion sensitivity. Third, in TESS, the crushers are applied along the RO direction, which has the highest spatial resolution. Strong crushers are needed, which would introduce unwanted diffusion effect that needs to be taken into account in the SSFP signal model since TR is long. Meanwhile, the crushers in iTESS can be put in the SS direction and can be small. Together with the short TR, the diffusion effect can be neglected.

In conclusion, a new interleaved SSFP sequence was proposed. It can be and successfully applied to simultaneous T1/T2 relaxometry.