Noninvasive measurement of skeletal muscle oxygen consumption and metabolism plays an important role in the clinical diagnosis of diseases such as peripheral arterial occlusive disease (PAOD) [1] and diabetes mellitus [2]. Recently, MRI based methods for measuring muscle oxygen extraction fraction (OEF) have been reported [3]. Asymmetric spin-echo (ASE) sequence combining with a susceptibility model [4] is the most widely used approach for measuring OEF. However, conventional ASE sequence uses single-shot (SS) EPI for data acquisition, which suffers from the problem of severe susceptibility artifacts and distortion due to the relatively long echo train length (ETL). One solution is to employ multi-shot (MS) EPI instead of SS EPI for data acquisition. With the use of MS-ASE technique, much higher spatial resolution could be achieved for lower extremity muscle imaging.The purpose of this study is to compare the performances of MS-ASE and SS-ASE sequence for muscle OEF measurement.

Sequences: The diagram of ASE sequence is shown in Fig.1. The ASE sequence is a modification of the spin-echo EPI sequence in which the 180° refocusing pulse has a time offset of τ from TE/2. By varying τ while keeping TE constant, the MR signal change is governed by the R2’ decay, which is directly related to the concentration of deoxyhemoglobin. In this study, a triple-echo ASE sequence with 20 varied 180° pulse shifts (from -10 ms to 9 ms with an increment of 1 ms) was
implemented to acquire the source images for muscle OEF quantification. Detailed scanning parameters for SS-ASE and MS-ASE are summaried in Table 1.

Phantom Studies: To compare the image quality of SS-ASE and MS-ASE sequence, we performed a comparison study using a quality control phantom. MRI measurements were carried out on a 3.0 Tesla MR scanner (Achieva, Philips Medical Systems, Best, Netherlands) with gradient strength = 80 mT/m and gradient slew rate = 200 T/m/s. A 32-channel Cardiac coil was used for signal reception.

In vivo Studies: Approved by the local institutional human study committee, eight healthy volunteers (mean age 23.5 ± 1.8 years) participated in this study. Muscle images were acquired with identical scanning parameters as phantom studies. A fat-saturation preparation was applied to reduce the chemical shift artifacts. To reduce the impact of the intravascular signal, a pair of small flow dephasing gradients (b = 40 s/mm²) was applied. The total acquisition time for SS-ASE and MS-ASE were 50 s and 200 s, respectively.

Quantitative Analysis: The measurement of muscle OEF was
derived from a theoretical model proposed by Yablonskiy and Haacke [4], in which a set of randomly orientated cylinders was assumed to characterize the signal behavior in static dephasing regime. A nonlinear least-squares curve fitting function was used to fit this model.

Images of the quality control phantom acquired using SS-ASE and MS-ASE sequences are shown in Fig.2. Less image distortion was observed in the MS-ASE image (τ = 0 ms). For muscle imaging, the SS-ASE image (Fig.3a) showed strong spatial blurring, this blurring was dramatically reduced in the MS-ASE image (Fig.3b) due to higher spatial resolution and shorter ETL. MS-ASE images (τ = 0 ms) showed obviously better image quality with sufficient depiction of different muscle groups. Representative lower extremity muscle OEF maps derived from SS-ASE and MS-ASE sequences are shown in Fig.4. The lower extremity muscle OEF derived from MS-ASE was 0.323 ± 0.028 in gastrocnemius muscle, and 0.332 ± 0.033 in soleus muscle, which is consistent with reported literature values. No significant difference was found between MS-ASE based OEF and SS-ASE based OEF (P = 0.33).With phantom and in vivo studies we compare the performances of MS-ASE and SS-ASE sequence for muscle OEF measurement. The MS-ASE technique could reduce artifacts and distortion caused by susceptibility differences, and limit spatial blurring dramatically, which is promising for evaluation of skeletal muscle oxygen metabolism in some diseases.