A steady state free precession signal consists of multiple components. These components can be separately acquired within one TR [1]. A long TR is needed to accommodate the multiple readouts, and is unfavorable to components which are formed over multiple TR cycles. DESS (dual echo in the steady state) is an example where the FID (S+) and spin echo (S-) components are sampled in one TR [2]. Here, S+ increases with TR while S- decreases with TR. The longer TR (compared to FISP or PSIF) increases the sequence’s sensitivity to motion and diffusion. Crusher gradients must be carefully designed to separate the two signals and minimize diffusion sensitivity [3].

Here, we propose to acquire the multiple components in the steady state signal by interleaved acquisition of SSFP signal components through appropriate gradient design using DESS as an example. In this case, we show that the two major components (S+ and S-) in the SSFP signal can be separately acquired in alternate TR and form images similar to those conventionally acquired by DESS.

Sequence: Figure 1 shows the proposed (three dimensional) sequence (interleaved DESS or iDESS). The crusher dephasing S- during the acquisition of S+ is moved to the start of the next TR cycle when acquiring S-. Properties of images formed from the two signals remain unchanged compared to FISP and PISF. The sequence is implemented on a 3T MRI system (TIM TRIO, Siemens, Erlangen)

Experiment: the sequence is tested on a phantom consisting of 2 test tubes with T1/T2 values of 654.9/41.3ms,418/28.8ms. Images from the FID and SE signals acquired with iDESS are acquired at several flip angles (10°, 15°, 20°, 25°, 30°). The signals are compared with the images separately acquired using the FISP and PSIF sequences (using the same flip angles) to check if interleaved acquisition affects the steady state signal.

iDESS is then tested in 3 healthy volunteers (IRB approved) for two applications of DESS: knee and vessel wall imaging. Informed consents are obtained from the healthy volunteers. High resolution knee imaging is performed in two volunteers. A custom made knee coil was used. Imaging parameters were: flip angle 40°, FOV=200×125×96mm³, spatial resolution=0.4×0.4×3mm³, TR/TE=11.6/5.75ms, bandwidth =199 Hz / pixel, NEX=2, acquisition time=7.9mins, partition resolution was zero-padded to 1.5mm. High resolution vessel wall imaging of the femoral arteries are performed in another 3 volunteers. A 16 channel body coil and the spine coil are used for reception. Imaging parameters used were: flip angle=32°, FOV=360×180×360mm³ , spatial resolution = 0.8×0.8×3mm³ , TR/TE=8.54/3.36ms,bandwidth=744Hz/pixel,NEX=1,iPAT=2,aquisition time =4.2mins, partition resolution was zero-padded to 1.5mm.

Figure 2 shows the results from the phantom experiment. The S+ and S- signals acquired by iDESS and those from FISP and PSIF using different flip angles are nearly identical.

Figure 3 shows the separate and then the combined images of the knee obtained using iDESS. Note the bright synovial fluid typical of DESS images of the knee.

Figure 4 shows the grey and dark blood signal of the femoral arteries, as observed in [4]

The study showed that components in the SSFP signal can be separately acquired in an interleaved way. In-vivo images of the knee and femoral arteries from iDESS have similar properties of DESS reported. Weak components in the SSFP signal would benefit most from this approach. While the technique may increase the SAR value, it helps reduce the sequence’s signal loss due to the increased sensitivity to motion and diffusion in DESS.