Design and evaluation of a 32‐channel phased‐array coil for lung imaging with hyperpolarized 3‐helium

Abstract

Imaging with hyperpolarized 3-helium is becoming an increasingly important technique for MRI diagnostics of the lung but is hampered by long breath holds (>20 sec), which are not always applicable in patients with severe lung disease like chronic obstructive pulmonary disease (COPD) or a-1-anti-trypsin deficiency. Additionally, oxygen-induced depolarization decay during the long breath holds complicates interpretation of functional data such as apparent diffusion coefficients. To address these issues, we describe and validate a 1.5-T, 32-channel array coil for accelerated 3 He lung imaging and demonstrate its ability to speed up imaging 3 He. A signal-to-noise ratio increase of up to a factor of 17 was observed compared to a conventional double-resonant birdcage for unaccelerated imaging, potentially allowing increased image resolution or decreased gas production requirements. Accelerated imaging of the whole lung with one-dimensional and two-dimensional acceleration factors of 4 and 4 3 2, respectively, was achieved while still retaining excellent image quality. Key words: MRI; parallel imaging; helium; lung; phased array coil Lung imaging using hyperpolarized 3 He has received increased attention in the last several years for studying lung diseases (1-12). So far, functional imaging, e.g., the apparent diffusion coefficient (ADC) or methods to observe the microstructure of the lung (12,13), requires long breath holds (up to 30 sec), which are the main barrier to applying this technology to patients with severe lung disease. Improvements in array coil methodology are a promising approach for increasing signal-to-noise ratio (SNR) and allowing significant acceleration of the image-encoding process. Lee et al. (14,15) presented a 24-channel array and later a 128-channel array for 3 He imaging, which demonstrated the feasibility of this approach. The acceleration afforded by parallel imaging is particularly important for hyperpolarized gas imaging since the nonequilibrium state of the magnetization allows faster imaging, often without incurring the standard SNR penalty proportional to the square root of imaging time (16). To analyze the improvements of these techniques in more detail, a 32-channel phased array was designed and built. For comparison, imaging protocols such as used in the European Community Framework V project ''Polarized Helium Imaging of the Lung'' were applied, as well as newly developed protocols, which utilize the extra sensitivity of the array. The sensitivity improvements were compared to a single-channel volume coil and the benefits of parallel imaging for lung diagnostics were analyzed by theoretical geometry-factor (g-factor) maps for the array and demonstrated in accelerated imaging. MATERIALS AND METHODS CoilsThe phased-array coil (Fig. 1) consists of 32 loops mounted on a clamshell former, which was presented by Schmitt et al. (17) for cardiac imaging. The 32 receive elements were milled from 35-lm copper-plated 1.5mm-thick FR4 circuit board material and arranged in two groups of 16 ele...

A flexible 32‐channel receive array combined with a homogeneous transmit coil for human lung imaging with hyperpolarized 3He at 1.5 T

“…For a given volunteer, no significant SNR difference is observed between scans with different acceleration factors along the R/L direction. This is consistent with previous findings (5,9) and highlights the fact that the SNR loss incurred due to a reduced total number of acquisitions can be compensated by an increased flip angle. The mean SNR for R ¼ 4 in the S/I direction (Fig.…”

“…In the R ¼ 4 (R/L) scan of the male volunteer, the maximum observed g-factor over all slices was g max ¼ 2.07. These g-factors are comparable with, and in some cases slightly less than those obtained with a coil design for the same application which relies on a pairwise decoupling scheme (9). While the pair-wise approach of the earlier design results in a maximum correlation of 43%, compared with 71% for the shielding decoupling scheme used here, the parallel imaging performance of the present coil does not appear to be compromised, as the mean correlation remains low.…”

“…This effect needs to be considered when performing dynamic inhalation studies using parallel imaging techniques (9). In that case, the lung will be imaged in all states of inflation and the g-factor distribution does not remain constant during the course of the experiment, which can present a confounding factor, particularly when performing a quantitative analysis of pixel intensities, e.g., to detect phenomena such as gas trapping.…”

“…This is because signal can be increased at lower N by increasing the flip angle (5,8). Initial results presented with 24, 32, and 128 channel arrays were promising, demonstrating parallel imaging of hyperpolarized gas with high SNR (5,(9)(10)(11).…”

Flip‐angle measurement by magnetization inversion: Calibration of magnetization nutation angle in hyperpolarized 3He magnetic resonance imaging lung experiments

“…(ii) A dedicated 32-receive channel phased array for 3 He imaging with Helmholz-type transmitting loops (24). Here, the transmit coils consist of two loops with the dimensions of 34 Â 37 cm 2 .…”