As a non-invasive medical imaging technique with high resolution and no radiation, magnetic resonance imaging (MRI) has been widely used in assessment of the fetus [1]. Magnetic resonance imaging (MRI) of the fetal brain provides a safe and powerful way to examine the brain anatomy and diseases during early development. Fetal MRI has been widely applied to investigate structural abnormalities of the fetal brain, such as ventricular dilatation, cerebellar dysplasia, and neuronal migration disorder [2]. The sequences generally used for fetal brain MRI are fast T2-weighted imaging(T2WI), T1-weighted imaging(T1WI), and diffusion weighted imaging (DWI). However, fetal brain images are known to have poor contrasts that are challenging for anatomical definition and detection of abnormalities. The main factors underlying this difficulty are inadequate contrast and artifacts associated with respiration of pregnant mother and movement of the fetus. The resolution of T1WI is lower than that of T2WI. It has been considered difficult to produce high-quality T1WI images of the fetal brain. T1WI contrast may be used to visualize regions of protein, calcification, hemorrhage, and fat. Further, T1WI is suitable for displaying brain structure and evaluating the maturation of myelination. Myelination, lamination, and migration are essential in brain development. These processes must occur correctly to form normal brain structure. The formation of normal brain structures, such as basal ganglia, central sulcus, lateral fissure, optic chiasm, gray matter, as well as myelination, reflect normal brain developmental processes, including lamination, migration, and myelination. Myelination is associated with decreased T1 relaxation time, which changes from hypointense to hyperintense relative to gray matter on T1WI. Thus, the T1WI sequence can be used to assess whether brain maturation is progressing normally [3]. Specific milestones have been established for when changes in white matter intensity (relative to gray matter intensity) are expected.3 Therefore, finding high-resolution T1WI by which to assess brain structure and level of myelination is important for assessment of the fetal nervous system. However, due to unpredictable fetal movement and abdominal motion of the pregnant mother, the scanning sequence and parameters are complicated and need to be appropriately defined.
Efforts have been made to maximize T1W image contrast in the fetal brains. Gholipour et al. conducted a thorough review of fetal MRI imaging techniques. Various acquisition schemes are used for fetal T1-weighted imaging [4]. Inversion recovery (IR) pulse-prepared single-shot turbo-spin-echo (TSE) acquisition exhibits high image quality and robustness to fetal movements and maternal breathing motion. Although a long echo train could be used to shorten acquisition time, this would increase T2 blurring. The specific absorption ratio (SAR) is another concern in the TSE sequence, especially at 3T scanner. Gradient-echo based acquisition can significantly reduce SAR. However, this approach may be limited from either mixed weighting of T1 and T2* for spoiled acquisition or banding artefact for balanced acquisition [4]. To enhance T1 contrast, an IR preparation pulse is applied to gradient sequences [5]. To decrease motion artifacts while imaging the fetus and eliminate the need for sedation, T1-weighted images are typically acquired through a two-dimensional turbo fast low-angle shot (FLASH) sequence. FLASH uses radio frequency excitation pulses with a low flip angle (less than 90º) and subsequent reading gradient reversal to produce a gradient echo signal [6]. In gradient-echo sequences all slices are acquired sequentially. Subsequently, given slight fetal movement at any time during the acquisition period, all slices are degraded by motion artifacts. However, as the voxel size is not equal among slices, segmentation or another image post-processing are not available. Inversion-recovery-prepared half-Fourier acquisition single-shot turbo spin echo (IR-HASTE) acquisition is another optimal choice for T1 imaging. IR-HASTE is based on single-slice acquisition. Fetal motion typically affects the particular slice that is being acquired while motion occurs [7]. The use of large echo train lengths with short echo time (TE) results in blurring and loss of contrast. Scan efficiency is improved to directly increase robustness against motion in 2D fetal MRI. 3D images are reconstructed using three-plane post reconstruction. An alternative way to reduce motion sensitivity is to use motion insensitive encoding trajectories, such as radial acquisition. One of the clinically available sequences is radial non-Cartesian T1-weighted gradient-recalled echo, known as radial VIBE (volumetric interpolated breath-hold examination, named radial VIBE by Siemens Healthcare as product sequence) [8–10]. This sequence provides several advantages over conventional 3D T1-weighted GRE imaging in its robustness to pregnant women’s respiratory motion and unpredictable motion by the fetus [11].
Therefore, to reduce motion artifacts, a 3D radial VIBE sequence [12] was used and optimized in our study. Because the T1 values in white matter and gray matter differ between fetuses and newborns, and T1 values change with development of the fetal brain, imaging parameters could be optimized to provide better T1 contrast for the fetal brain. The T1 contrast of the 3D radial VIBE sequence primarily depends on imaging parameters, including flip angle (FA) and repetition time (TR). In our research, the shortest TR was utilized to minimize scan time. Therefore, different flip angles were utilized to find the most suitable parameter for the 3D radial VIBE sequence and compared with the turbo FLASH (TFL) sequence.