MR imaging of the fetal brain at 1.5T and 3.0T field strengths: comparing specific absorption rate (SAR) and image quality

Abstract

Objective

1) To evaluate the feasibility of fetal brain magnetic resonance imaging (MRI) using a fast spin echo sequence at 3.0T field strength with low radio frequency (rf) energy deposition (as measured by specific absorption rate – SAR) than 1.5T; 2) and to compare image quality, tissue contrast and conspicuity between 1.5T and 3.0T MRI.

Methods

T2 weighted images of the fetal brain at 1.5T were compared to similar data obtained in the same fetus using a modified sequence at 3.0T. Quantitative whole body SAR and normalized image signal to noise ratio (SNR); and a nominal scoring scheme based evaluation of diagnostic image quality and tissue contrast & conspicuity for specific anatomical structures in the brain, were compared between 1.5T and 3.0T.

Results

12 pregnant women underwent both 1.5T and 3.0T MRI examinations. The image SNR was significantly higher (p=0.03) and whole body SAR was significantly lower (p < 0.0001) for images obtained at 3.0T compared to 1.5T. All cases at both field strengths were scored as having diagnostic image quality. Images from 3.0T MRI (compared to 1.5T) were: 1) equal (57%; 21/37) or superior (35%; 13/37) for tissue contrast; and 2) equal (61%; 20/33) or superior (33%, 11/33) for conspicuity.

Conclusion

It is possible to obtain fetal brain images with higher resolution and better SNR at 3.0T with simultaneous reduction in SAR compared to 1.5T. Images of the fetal brain obtained at 3.0T demonstrated superior tissue contrast and conspicuity compared to 1.5T.

INTRODUCTION

While ultrasound has been the primary modality used in prenatal diagnosis the role of magnetic resonance imaging (MRI) in fetal diagnostic evaluation is increasing [1]. Fetal MRI has multiple advantages over ultrasound in this field.[2–36] Such advantages include: 1) improved soft tissue contrast and characterization[4, 21, 37–39]; 2) access to functional data [e.g. diffusion weighted imaging (DWI),[40–48] perfusion weighted imaging (PWI)[12, 32, 49, 50], blood oxygenation dependent (BOLD),[51–55] MR spectroscopy (MRS)[10, 56–65]]; and 3) larger field of view (FOV). Fetal MRI has been shown to be useful in diagnosis of fetal pathologies [66–68]. In particular, fetal MRI has been shown to be superior to ultrasound in evaluating the fetal central nervous system.[4–7, 16, 17, 69–77]

Three decades have passed since the first application of MRI in the human pregnancy.[78] Currently, 1.5 Tesla (T) is the field strength of choice for MRI in pregnancy, although fetal MR imaging at 3.0T has also been reported in several studies.[55, 73, 79–82] Adult, pediatric, and neonatal populations routinely undergo clinical MR examinations at 3.0T.[83–86] The major advantage of 3.0T MRI (vs. 1.5T) is the increased signal to noise ratio (SNR), which is linearly proportional to the imaging field strength. Demonstrated benefits of such increased signal include: 1) higher imaging resolution; 2) shorter imaging time; 3) improved tissue biochemical profiling through MRS; and 4) improved functional data.[87–90] However, susceptibility related artifacts are more pronounced at 3.0T MRI than 1.5T.[88–90] More importantly, 3.0T MRI has an operating frequency of 128 MHz (vs. 64 MHz of 1.5T), leading to higher radiofrequency (rf) energy deposition.[91–93] Specific absorption rate (SAR) measures energy deposition, and is defined as rf power absorbed per unit mass of the tissue (Watts/Kg).[94–96] The issue of increased SAR in pregnancy has limited the use of 3.0T MRI in fetal imaging.[73, 82] Yet, radiofrequency energy deposition (SAR) can be reduced through pulse sequence modifications.[97–100] Therefore, we hypothesized that by using modified pulse sequences, imaging the fetal brain at 3.0T (vs. 1.5T) could be performed at a higher resolution with improved image quality, while simultaneously reducing the SAR.

The objectives of this study were to: 1) evaluate the feasibility of fetal brain magnetic resonance imaging (MRI) using a fast spin echo imaging sequence at 3.0T field strength with simultaneous reduction in radio frequency (rf) energy deposition; 2) quantitatively compare the image quality with conventional 1.5T MRI; and 3) compare tissue contrast and conspicuity for specific anatomical structures in the brain between 1.5T and 3.0T MRI.

MATERIALS AND METHODS

Pregnant women (19–40 weeks of gestation) scheduled to undergo a 1.5T MR exam for clinical indications were approached to also undergo a 3.0T exam afterwards. All women recruited as part of this study were referred for a clinical MRI through Hutzel Women’s Hospital in Detroit, MI. The 1.5T MRI scans were performed at Children’s Hospital of Michigan, Detroit, MI, while 3.0T MRI scans were performed at Wayne State University’s Magnetic Resonance Research Facility at Detroit Medical Center, Detroit, MI. All women were enrolled in a research protocol approved by the Human Investigation Committee of Wayne State University, and all participants provided written informed consent for the use of MR images for research purposes.

MRI Examination

To compare the image quality and SAR between 1.5T and 3.0T MR, the T2 weighted single shot fast spin echo sequence (SSFSE) was chosen, since it is the most frequently acquired sequence for evaluating fetal anatomy, and has the highest SAR values in typical fetal MRI protocols.[101]

Clinical MRI scans were performed on a 1.5T General Electric Signa system (Milwaukee, WI) with an 8 channel cardiac array and spine receive coils. Following the scout/localizer scans, anatomical data of the fetal brain were acquired using a T2 weighted SSFSE technique with the following imaging parameters (Table 1): repeat time (TR) 1192–1240 milliseconds (msec), echo time (TE) 90 – 240 msec, slice thickness of 4 mm, voxel size 0.93 x 0.93 mm² – 1.3 x 1.3 mm², and flip angle 90°. Images were obtained sequentially in three planes (axial, coronal, sagittal) relative to the fetal brain. Acquisitions were repeated when fetal motion was encountered.

Table 1.

MR imaging parameters for the single shot fast spin echo (SSFSE) sequence at 1.5T and 3.0T

Field Strength	TE (msec)	TR (msec)	Resolution (mm3)	Flip Angle (degree)	Band width (Hz/pixel)
1.5 T	90 – 240	1192–1240	(0.93 to 1.3) x (0.93 to 1.3) x 4	90	244
3.0 T	139–140	2600–5000	(0.87 to 1.1) x (0.87 to 1.1) x (3 to 4)	75	369 or 372

All 3.0T MRI scans were performed on a 3.0T Siemens Verio system (Erlangen, Germany) with a 6 channel body flex array and spine receive coils. An additional 2 channel flex extremity receive coil was used in some patients having a larger abdominal girth. Following scout/localizer scans, a T2 weighted single shot turbo spin echo sequence with half-Fourier reconstruction (HASTE) was acquired for anatomical purposes. This sequence implementation is essentially the same as SSFSE on the GE system.[102] The following imaging parameters were used for the HASTE sequence: repeat time (TR) 2600–5000 milliseconds (msec), echo time (TE) 139–140 msec, slice thickness of 3 or 4 mm, voxel size 0.87 x 0.87 mm² – 1.1 x 1.1 mm², and flip angle 75° with hyper-echo acquisition. Images were obtained sequentially in three planes (axial, coronal, sagittal) relative to the fetal brain. Acquisitions were repeated when fetal motion was encountered.

Data Analysis

T2 data was first reviewed for general image quality, including image or motion artifacts. From data collected at a given field strength, the best dataset (defined as a volume without motion or image artifacts) for each anatomical plane relative to the fetal brain was chosen for further analysis. Each fetus had 3 datasets collected at 1.5T that were compared with the corresponding datasets acquired at 3.0T.

a. Signal to noise ratio (SNR)

Quantitative comparison of image quality was performed by computing the SNR of the fetal brain. For a given fetus, SNR measurements were computed for each T2 dataset. For all cases, SNR was measured by defining a region of interest (ROI) within the white matter, and mean signal was measured from this ROI. Noise was estimated by measuring the signal standard deviation (SD) from a homogenous white matter region within the ROI. SNR was then defined as: ratio of the mean value of the signal from the ROI and SD measured from the homogeneous signal region. For a given dataset, a central slice within the multislice data was first chosen, such that both cerebral hemispheres were clearly visualized. Next, the mean and SD values from two ROIs (one from each cerebral hemisphere) were obtained, and then averaged to minimize the bias due to coil drop-off or non-uniform excitation.[88, 90, 103, 104] Compared to the 1.5T scan, higher resolution data were acquired at 3.0T with a corresponding higher pixel bandwidth. Therefore, the SNR measures were normalized to a fixed voxel volume of 1 mm³ and a fixed pixel bandwidth of 244 Hz/pixel, so that a direct comparison could be made between SNR measures of the 1.5T and 3.0T data. Finally, SNR measures from the three T2 data acquisitions were averaged to obtain a single SNR measure for each fetus at a given field strength.

b. Specific absorption rate (SAR)

Whole body SAR values, as estimated by the MR system console, were noted from the DICOM[105] (Digital Imaging and Communications in Medicine) image header. For a given patient, SAR values from the three separate views were noted and averaged to obtain a single SAR measure for each patient at a given field strength. It is noteworthy that whole body SAR values are calculated based on maternal physical parameters, which are also the criteria used to assess safety in rf dosimetry studies for fetal imaging.[106–108]

c. Diagnostic image quality and scoring system

Diagnostic image quality was assessed in a blinded fashion for all 1.5T and 3.0T MR images by a senior pediatric neuroradiologist (SM) with more than 10 years of experience with fetal MRI. The following nominal scoring scheme was used: 1) Score 1: images of diagnostic quality without any artifacts; 2) Score 2: images of diagnostic quality, but with minor artifacts or low SNR; and 3) Score 3: images of non-diagnostic quality. Three datasets were evaluated for each fetus, and the best score was used to represent the overall data quality for that fetus respective to the field strength of the image.

d. Comparison of tissue contrast and conspicuity between 1.5T and 3.0T MRI

High soft tissue contrast occurs when different tissues are reflected by different intensity levels in the images.[37] Conspicuity is the property of being clearly discernible.[109, 110] Both tissue contrast and conspicuity of anatomical structures in the fetal brain were compared in a blinded fashion between 1.5T and 3.0T MRI by a senior pediatric neuroradiologist (SM) with more than 10 years of experience with fetal MRI. Specifically, tissue contrast for the fetal cortex, basal ganglia, dentate nucleus, and germinal matrix were evaluated due to their inherent discernibility. Conspicuity was evaluated for the fetal optic chiasm, basilar artery and vein of Galen due to their small size. Assessment was performed using the following nominal scoring scheme (performance of 3.0T relative to 1.5T): 1) Score 0: inferior; 2) Score 1: same as 1.5T data; 3) Score 2: superior; and 4) NA: not applicable (some structures cannot be visualized due to either early gestational age or pathology). After reviewing all T2 data acquired in different orientations at a given field strength, a single score was assigned (one for tissue contrast and one for conspicuity) for a given fetus.

e. Statistical Analysis

Normality was assessed using the Kolmogorov-Smirnov test and visual plot inspection. Differences in distributions of normalized SNR and SAR were tested using either the paired t-test or its non-parametric equivalent, the Wilcoxon signed rank test, as appropriate. A 5% threshold was used in determining statistical significance. All analyses were performed using SAS version 9.3 (Cary, N.C.).

RESULTS

Twelve pregnant women prospectively underwent both a 1.5T and 3.0T fetal MR examination. Patients were referred for MRI examination due to the presence of fetal congenital anomalies, which included: Dandy Walker malformation (n = 2), mild cerebral ventriculomegaly, hydrocephalus, mega cisterna magna, myelomeningocele, hypoplastic left heart syndrome, distended small bowel loops, congenital heart disease including left heart syndrome, diaphragmatic hernia, cytomegalovirus infection, and monochorionic/diamniotic twins (with one viable fetus). The median (range) age of mothers was 24 (19–34) years. The median (IQR) gestational age at the time of 3.0T scan was 31.4 (27–34.2) weeks, and the median (IQR) time interval between the 1.5T and 3.0T scans was 2.5 (0.75 –5.25) days. All but three women were imaged within 3 days of the 1.5T scan; these women underwent a 3.0T MRI at 12, 19, and 20 days after the first scan.

a. SNR and SAR

Placement of the ROI on the T2 datasets for SNR measurements are shown in Figure 1. For all cases, normalized SNR and SAR measurements for fetal images obtained at 1.5T and 3.0T are depicted in Table 2. The SNR per unit voxel volume of 1 mm³ (arbitrary units, a.u.) was significantly higher for images obtained using 3.0T than those obtained at 1.5T [median (IQR):4 (3.1–5.6) vs. 3.35 (2.5–3.65), respectively; p = 0.03]. Conversely, the whole body SAR value was significantly lower for images obtained at 3.0T than those obtained at 1.5T (mean ± SD: 0.6 ± 0.12 vs. 1.6 ± 0.2 Watt/kg, respectively; p < 0.0001) (Figure 2). Even when excluding the three subjects with more than 3 days between the 1.5T and 3.0T scans, such differences remained significant (SNR; p = 0.03 and SAR; p < 0.0001).

Figure 1.

Placement of the ROI on T2 MRI datasets. ROIs were drawn on data acquired in three orientations for SNR measurements. Gestational age (weeks) A: 34 1/7; B: 34 4/7; C: 22 3/7. Images shown were acquired at 3.0T field strength.

Table 2.

Comparison of normalized signal to noise ratio (SNR, arbitrary units) per unit voxel volume of 1 mm³ and specific absorption rate (SAR, Watt/Kg) between T2 weighted single shot fast spin echo sequence data (n=12 subjects) obtained at 1.5T and 3.0T fetal MRI field strengths.

	1.5T		3.0T
Subject #	SNR (a.u.)	SAR (Watt/kg)	SNR (a.u.)	SAR (Watt/kg)
1	3.2	1.1	5.5	0.6
2	3.4	1.8	2.1	0.3
3	2.0	1.8	3.1	0.7
4	1.5	1.8	3.7	0.5
5	3.4	1.5	5.3	0.7
6	3.4	1.7	5.8	0.5
7	5.3	1.7	5.7	0.6
8	2.3	1.7	4.3	0.7
9	2.7	1.5	3.6	0.6
10	3.9	1.5	2.9	0.7
11	3.3	1.5	3.1	0.8
12	3.9	1.5	6.0	0.6
Mean (SD)		1.6* (0.2)		0.6* (0.1)
Median (IQR)	3.35† (0.9)		4† (2.4)

^*P-value < 0.0001

^†P-value = 0.03

Figure 2.

Comparison of normalized SAR and SNR values between T2 weighted single shot fast spin echo (SSFSE) sequence data obtained at 1.5T and 3.0T fetal MRI field strengths. (a.u., arbitrary units).

b. Diagnostic Image Quality

Anatomical data sets in different orientations were acquired from all fetuses at both field strengths. The best score was used to represent the overall data quality for a specific fetus. Scores assigned for diagnostic image quality are shown in Table 3. All cases at both field strengths were scored as having diagnostic quality present (Score of 1 or 2). Thus, there was no case which received a Score of 3 (non-diagnostic quality). Of all the cases, 83.3% (10/12) demonstrated equal diagnostic quality between both field strengths: 1) 75% (9/12) received a Score of 1 at both 1.5T and 3.0T (indicating diagnostic image quality without any artifacts); and 2) 8.3% (1/12) received a Score of 2 at both 1.5T and 3.0T (indicating diagnostic image quality, but with minor artifacts or low SNR). In 16.7% (2/12) cases, images from 3.0T received a Score of 2, while the corresponding images from 1.5T received a Score of 1. Nevertheless, for a given fetus, data from at least one anatomical orientation had diagnostic quality present at both imaging field strengths.

Table 3.

Scores assigned for diagnostic image quality for fetal images obtained at 1.5T and 3.0T MRI. The best score was used to represent the overall data quality for a specific fetus.

Diagnostic Image Quality Score
Subject #	3.0T	1.5T
1	2	1
2	2	1
3	1	1
4	1	1
5	1	1
6	1	1
7	1	1
8	1	1
9	1	1
10	2	2
11	1	1
12	1	1

Score 1: images of diagnostic quality without any artifacts

Score 2: images of diagnostic quality, but with minor artifacts or low SNR

Score 3: images of non-diagnostic quality

c. Tissue Contrast

Tissue contrast was evaluated and compared between field strengths for four anatomical structures in the fetal brain: cortex, basal ganglia, dentate nucleus and germinal matrix (Table 4). While the cortex could be visualized and evaluated in all 12 fetuses, the basal ganglia and dentate nucleus were each visualized and evaluated in 11 fetuses, and the germinal matrix in 3 fetuses. Some anatomical structures were not visualized due to either early gestational age or the pathologic abnormality, and were scored as NA (not applicable). Thus, a total of 37 scores were assigned. Images of the cortex, basal ganglia, dentate nucleus, and germinal matrix obtained from 3.0T were equal (57%; 21/37), or superior (35%; 13/37) to that of 1.5T for tissue contrast. In 8% (3/37), tissue contrast of the dentate nucleus was inferior on 3.0T (vs. 1.5T) MRI.

Table 4.

Comparison of tissue contrast and conspicuity for anatomical structures in the fetal brain between 1.5T and 3.0T MRI data. Images are scored with regards to how 3.0T compares to 1.5T MRI (reference).

	Tissue contrast				Conspicuity
	Cortex	Basal Ganglia	Dentate Nucleus	Germinal Matrix	Optic Chiasm	Basilar Artery	Vein of Galen
Inferior (Score 0)	0	0	3	0	1	1	0
Equal (Score 1)	6	7	5	3	6	7	7
Superior (Score 2)	6	4	3	0	4	3	4
NA	0	1	1	9	1	1	1

d. Conspicuity

Conspicuity was evaluated and compared between field strengths for three fetal anatomical brain structures: optic chiasm, basilar artery and vein of Galen (Table 4). In one fetus, these structures could not be visualized due to early gestational age. Thus a total of 33 scores were assigned. Images of the optic chiasm, basilar artery, and vein of Galen obtained from 3.0T were equal (61%; 20/33) or superior (33%; 11/33) to that of 1.5T for conspicuity. In 6% (2/33), conspicuity of the optic chiasm and basilar artery was inferior on 3.0T (vs. 1.5T) MRI.

Comparisons between MR images of the fetal brain obtained at 1.5T and 3.0T field strengths are shown in Figures 3 through 5. Figure 3 compares images from a 26 week fetus that was scanned on the same day at both field strengths. Figure 4 depicts 1.5T images of the fetal brain and the corresponding 3.0T images of three different fetuses at varying gestational ages. Figure 5 compares the conspicuity and contrast of the germinal matrix, optic nerve, and basilar artery, as well as the migrational pattern seen between 1.5T vs. 3.0T MRI. Superior tissue contrast and conspicuity was observed in the 3.0T images.

Figure 3.

Comparing images of the fetal brain at 26 weeks of gestation obtained at 1.5T and 3.0T MRI (same fetus): 1.5T (top row, A–C) and 3.0T (bottom row, D–F) in all three orientations: Axial (A,D); Sagittal (B,E); and Coronal (C,F). Both 1.5T and 3.0T scans were performed on the same day. The images from 3.0T show superior tissue contrast and conspicuity to that of 1.5T.

Figure 5.

Comparison of 1.5T (top row) vs. corresponding 3.0T (bottom row) MR images of the fetal brain, showing the advantages of increased resolution at 3.0T. 1) Blue arrows (A, D) show the pattern of migration more clearly at 3.0T; 2) green arrows (B, E) show increased contrast in the germinal matrix; and 3) yellow and red arrows (C, F) show clear delineation of the optic nerve and basilar artery, respectively, at 3.0T. Images in the top row and the corresponding images in the bottom row are from the same fetus. Gestational age (weeks) at the time of scan were: (A: 22, D: 22 3/7); (B and E: 26 1/7) ; (C: 35 1/7, F: 35 4/7).

Figure 4.

1.5T MR images of the fetal brain (top row) and the corresponding 3.0T images (bottom row) across different gestational ages. Data were obtained from 3 different fetuses. Images in the top row and the corresponding images in the bottom row are from the same fetus. Gestational age (weeks) at the time of scan were: (A: 22, D: 22 3/7); (B: 27, E: 27 2/7); and (C and F: 35 1/7). Superior tissue contrast and conspicuity is demonstrated in the 3.0T images.

DISCUSSION

The principal findings of this study are: 1) SNR was significantly higher for images obtained using 3.0T than those obtained at 1.5T; 2) the average whole body SAR value was significantly lower for images obtained at 3.0T than those obtained at 1.5T; 3) all cases at both field strengths were scored as having diagnostic quality present, and 83.3% of cases demonstrated equal diagnostic quality between both field strengths; 4) images from 3.0T MRI (compared to 1.5T) were equal (57%) or superior (35%) for tissue contrast; and 5) images from 3.0T MRI (compared to 1.5T) were equal (61%) or superior (33%) for conspicuity of anatomical structures. Such superior conspicuity could be attributed, in part, to higher resolution imaging with 3.0T MRI, and changes in the tissue relaxation times between 1.5 and 3.0T. The smaller voxel size allowed for clearer visualization of certain anatomical structures (e.g. optic nerve) (Figure 5) and thus, improved conspicuity. Tissue T2 relaxation values are known to decrease with increasing field strength. [111–113] Due to the use of almost equivalent echo times between both field strengths, this effectively led to an increased T2 weighting at 3.0T compared to 1.5T, which may have contributed to the improved tissue contrast and conspicuity.

Signal to noise ratio (SNR) in MR imaging is roughly linearly proportional to the imaging field strength, and is one of the main incentives for moving to fetal imaging at 3.0T. Practically, the factor 2 gain in SNR is not always seen at 3.0T due to various factors, such as rf inhomogeneity (excitation and reception) and field appropriate sequence modifications performed for optimizing image quality. In the study herein, additional sequence modifications were employed for SAR reduction. Yet, despite this, we observed a significantly higher SNR at 3.0T than at 1.5T. The higher SNR allowed for higher resolution acquisition at 3.0T (compared to 1.5T), which improved the definition of smaller anatomical structures, such as the optic chiasm.

SAR is a crucial factor to consider when imaging the fetus at 3.0T. Importantly, when imaging the fetal brain at 3.0T, there were lower SAR values (despite better SNR) compared to that of 1.5T (by almost a factor of 2). The SAR values compared in the current study are average whole body SAR values as calculated by the MRI scanner, which assumes certain standard adult imaging conditions.[114] SAR is also dependent on multiple factors, such as body shape, surface area, composition, and spatial location within the scanner. Recent rf dosimetry studies of pregnant human models report that, provided the scanner calculated maternal whole body SAR is < 2W/kg, the local fetal SAR values and tissue temperature increase at 3.0T field strength are well within the safety limits.[106–108] Moreover, the lower the maternal whole body average SAR, the lower is the rf energy deposition in the fetus.[106–108] The SAR parameter also has the following dependency on MR imaging parameters and patient weight:[92, 93, 114, 115]

$$\text{SAR}\mathrm{\alpha }{(\theta )}^2\ast (1/\text{TR})\ast (1/{\mathrm{\tau }}_{\text{rf}})\ast (1/\mathrm{W}),$$

where θ is the rf pulse flip angle of the imaging sequence, τ_rf is the rf pulse duration, TR is the repetition time of the MR imaging sequence, and W is the weight of the patient. Therefore, SAR is directly proportional to the imaging flip angle, and inversely proportional to the rf pulse duration and TR of the sequence. In this study, a HASTE sequence with hyperecho option was used for imaging at 3.0T.[100, 116–118] Along with the use of lower flip angle pulses of 75°, this hyperecho option helped to reduce rf energy deposition.[100, 116–118] In addition, rf pulses with a longer pulse duration (using the ‘low SAR’ option on the scanner) and a longer TR were used for the HASTE sequence at 3.0T. All of these factors contributed towards significantly lower SAR values at 3.0T compared to 1.5T. The SAR values at 3.0T reported herein, are also lower than those reported by Victoria et al in a recent review paper.[82] Indeed, in the study herein, the maximum SAR for the HASTE sequence at 3.0T field strength was 25% less than the 2 W/Kg limit corresponding to the ‘normal’ operating mode of clinical MRI scanners.[119, 120] While the lower nominal excitation flip angle does affect the SNR to some extent, this did not significantly reduce the quality of the images at 3.0T, which is evident from the diagnostic image quality scores.

Some limitations of this work include the small sample size and variation in the time duration between 1.5T and 3.0T studies. However, even when excluding the three subjects with more than 3 days between the 1.5T and 3.0T scans, the differences in SNR and SAR between field strengths remained significant and did not alter our findings.

MR imaging at 3.0T offers tremendous advantages in terms of SNR and improved spectral separation in MRS. This could allow the use of faster and more sensitive advanced sequences to image the human fetus (e.g. brain) at 3.0T, such as susceptibility weighted imaging, diffusion weighted imaging, and magnetic resonance spectroscopy, all of which are typically low in SAR.[121]

CONCLUSION

This is the first study in the human fetus to systematically compare SNR, SAR and image quality between 1.5T and 3.0T MRI. With appropriate sequence adaptations, examining the fetal brain using 3.0T MRI results in higher image resolution and SNR, with simultaneously lower radio frequency energy deposition than that of 1.5T. Moreover, 3.0T images demonstrate superior tissue contrast and conspicuity than images obtained using 1.5T MRI in approximately one-thirds of cases.