Pulse sequences such as the dual echo steady state (DESS) or triple echo steady state (TESS) schemes have enabled rapid measurements of transverse relaxation times (T₂) [1,2]. In contrast to DESS, TESS is robust against systematic T₁ bias and static field homogeneities, thus delivering accurate T₂ maps in vitro [2]. However, it has been shown that TESS may result in errorneous T₂ offsets when applied to articular cartilage [3]. Cartilage is a multi-compartment tissue consisting of 65-80% water and the remainder of macromolecules and lipids [4]. To this end, chemical shift-related signal fluctuations in the TESS echo modes need to be considered and the accuracy of T₂ quantification re-assessed. Accordingly, the purpose of this work was to investigate the accuracy of TESS for T₂ quantification of multi-compartmental tissue using computer simulations, in vitro and in vivo experiments.

Numerical simulations: Bloch simulations were performed using the exact RF pulse shapes and other experimental parameters of the in vitro experiments. A simulated DESS protocol served as comparison. The multi-compartment effect was studied by simulating the following four scenarios: (1) single-compartment tissue; (2) two-compartment tissue, chemical shift 3.5 ppm, ratio 1:10; (3) two-compartment tissue, chemical shift 3.5 ppm, ratio 1:5 and (4) three-compartment tissue, chemical shift -0.4/4.0 ppm, ratio 1:1.5:10. Simulations were repeated for a range of echo spacing values.

In vitro experiments: Phantom samples were prepared to match the four multi-compartment scenarios: (1) Gd-doped water; (2) water-oil emulsion, 10% fat fraction; (3) water-oil emulsion, 20% fat fraction and (4) rapeseed oil with its three largest peaks corresponding to the three compartments. The samples were placed in a 1.5 T whole-body MRI scanner (Achieva, Philips Healthcare, Best, the Netherlands) with a standard 4-element receiver coil array. TESS scan parameters were: flip angle 15°, FOV 100x100x30 m³, voxel size 1.0x1.0x3.0 mm³. The first echo was fixed at TE 3.6 ms and the echo spacing was varied for different scans according to dTE 3.0, 3.2, … 6.0 ms and correspondingly TR 13.3, 13.7, … 19.3 ms. As reference, a single slice from a single-echo spin echo (SE) and multi-echo spin-echo (CPMG) sequence were acquired with TR 5000 ms and echo times ranging from 10 to 100 ms (otherwise identical parameters).

In vivo experiments: Sagittal and axial human knee images were acquired with scan parameters, FOV 200x200x30 mm³, voxel size 1.5x1.5x3.0 mm³ and otherwise identical parameters as in vitro.

Post-processing: T₂ maps were generated from the reference data using a standard exponential fit. Mean T₂ ± standard deviation were calculated in regions of interest (ROI).

Fig. 1 shows the signal intensities of the TESS echo modes as a function of echo spacing for the different scenarios. The single-compartment scenario showed slowly and little varying modes. The multi-compartment scenario, however, exhibited a strong echo time dependency, resulting in oscillatory fluctuations of the modes and deviations from the expected signal intensities as evident from simulations (Fig. 1a) and measurement (Fig. 1b). This effect was not observed in DESS (Fig 1c). Estimated T₂ values of the scenarios are given in Table 1. TESS produced varying results dependent on the choice of echo times. Fig. 2 displays T₂ maps of the knee acquired using the CPMG sequence as reference and using TESS with varying echo times. The T₂ values of cartilage and muscle derived from TESS were approximately 3 times larger compared to the corresponding CPMG values. The choice of echo spacing caused variations of up to 32% (Table 1).In comparison to T₂ quantification with DESS, the accuracy of TESS is influenced by the presence of multiple compartments in tissue causing chemical shift induced echo signal fluctuations. Interestingly, these errors could also be observed in pure fat. Although in vivo imaging was limited to the knee in order to avoid additional errors from motion, the presented findings are applicable to the characterization of a variety of tissues. Especially for hepatic and gastrointestinal applications the here described effect of the fat compartment is most likely to be more pronounced.While T₂ quantification of TESS is feasible in homogeneous tissue, multi-compartment effects limit its accuracy and therefore the interpretation of the measured T₂ values must be undertaken with caution.