Signal-to-noise ratio increase in carotid atheroma MRI: a comparison of 1.5 and 3 T

Abstract

Objectives: This study reports quantitative comparisons of signal-to-noise ratio (SNR) at 1.5 and 3 T from images of carotid atheroma obtained using a multicontrast, cardiac-gated, blood-suppressed fast spin echo protocol.

Methods: 18 subjects, with carotid atherosclerosis (>30% stenosis) confirmed on ultrasound, were imaged on both 1.5 and 3 T systems using phased-array coils with matched hardware specifications. T(1) weighted (T(1)W), T(2) weighted (T(2)W) and proton density-weighted (PDW) images were acquired with identical scan times. Multiple slices were prescribed to encompass both the carotid bifurcation and the plaque. Image quality was quantified using the SNR and contrast-to-noise ratio (CNR). A phantom experiment was also performed to validate the SNR method and confirm the size of the improvement in SNR. Comparisons of the SNR values from the vessel wall with muscle and plaque/lumen CNR measurements were performed at a patient level. To account for the multiple comparisons a Bonferroni correction was applied.

Results: One subject was excluded from the protocol owing to image quality and protocol failure. The mean improvement in SNR in plaque was 1.9, 2.1 and 2.1 in T(1)W, T(2)W and PDW images, respectively. All plaque SNR improvements were statistically significant at the p<0.05 level. The phantom experiment reported an improvement in SNR of 2.4 for PDW images.

Conclusions: Significant gains in SNR can be obtained for carotid atheroma imaging at 3 T compared with 1.5 T. There was also a trend towards increased CNR. However, this was not significant after the application of the Bonferroni correction.

Figure 1

Illustrating the circular regions of interest (ROIs) used for signal-to-noise ratio calculations on a representative proton density weighted image of the cylindrical phantom. These were automatically located on the radial line between the centre of the phantom and the centre of mass of the signal on that side (shown by the two lines and stars), with the outer edge of the ROI 2% from the average edge of the phantom.

Figure 2

Example of demarcation of regions of interest on images. 1.5 T T₁ weighted images (a) without and (b) with regions of interest drawn for lumen, outer vessel wall and within the sternocleidomastoid muscle. 3 T T₁ weighted images (c) without and (d) with the same regions of interest drawn.

Figure 3

Box plot illustrating the signal-to-noise ratio (SNR) difference between 1.5 and 3 T for the three image types (T₁ weighted, T₂ weighted and proton density weighted). Plots show the median and interquartile range as well as the range for each parameter. SNR measurements for plaque and sternocleidomastoid muscle are given (right- and left-sided measurements illustrated separately).

Figure 4

Example of 1.5 and 3 T black-blood carotid images of both carotid arteries for T₁ weighted, T₂ weighted and proton density-weighted images. There is bilateral atheromatous plaque present, which is demonstrated on the multicontrast 1.5 T images; however, this is more clearly seen at 3 T, and the plaque morphology is easier to distinguish.