Introduction

Magnetic resonance imaging (MRI) is a proven noninvasive diagnostic test for the detection, evaluation, assessment, staging, and follow-up of disorders of the knee¹. At 3T, high-resolution proton density-weighted (PDW), T2-weighted (T2W) and T1-weighted (T1W) images are often acquired with two-dimensional (2D) fast (turbo) spin-echo (FSE or TSE) sequences at different orientations in routine knee exams^1,2. Additional sequences may be used for their sensitivity for specific abnormalities. One example of these sequences is 2D gradient recalled echo (GRE) imaging in radial planes to better detect meniscal disorders^3,4. Fat suppression is commonly used in one or more of the sequences to enhance the diagnostic yield, for instance, to identify marrow abnormalities¹.

Spiral imaging has many favorable characteristics such as fast imaging, high signal-to-noise ratio (SNR) efficiency and low sensitivity to motion. Spiral imaging has been implemented for high resolution brain imaging^5-7 along with the advancements in spiral Dixon technique⁸. The goal of this study is to investigate the feasibility of spiral knee imaging with in-plane resolution as high as 0.4 x 0.4 mm².

Materials and Methods

Spiral pulse sequences were implemented on a 3T scanner (Ingenia Elition, Philips Healthcare, Best, The Netherlands). Data of four volunteers were collected using a 16-channel knee transmit/receive coil. Informed consent was obtained in accordance with the guidelines of the institutional review board. Sagittal PDW, sagittal T2W, coronal T1W and radial plane images were acquired. For all fully sampled spiral scans, data were acquired at each nominal echo time (TE) with two TE shifts, 0 and 1.15 ms, for Dixon water-fat imaging. Field maps of the static field inhomogeneity were obtained from separate low-resolution Cartesian scans, with which water and fat images were reconstructed using a joint water-fat separation and deblurring method⁸. Cartesian scans with parameters close to the clinical routine knee exams were also performed for comparison. Some of the scan parameters are summarized in Table 1.

Sagittal PDW and T2W imaging
A 2D dual spin-echo (SE) technique with hybrid spiral readouts was used to simultaneously collect data for PDW and T2W at the first echo and the second echo respectively⁷. The first echo was acquired with spiral-out trajectories and the second echo was acquired with spiral in-out trajectories that has twice the readout length of the first echo to improve the SNR of the T2W images. In-plane flow compensation was applied for the second echo. A saturation band was also added superior to the imaging field of view (FOV). The acquired spiral in-out k-space data were divided into the spiral-in part and the spiral-out part, which were reconstructed separately. The deblurred and water-fat separated spiral-in and spiral-out images were added together to form the final water and fat images.

The readout for both Cartesian TSE PDW and T2W sequences was in the anterior- posterior (AP) direction so that the flow artefacts from the popliteal artery would be in the head-foot (HF) direction to minimize the effects on the center knee area. Spectral attenuated inversion recovery (SPAIR)⁹ was applied to suppress the fat signal in the Cartesian T2W sequence.

Coronal T1W imaging
A 2D SE sequence with spiral out readout⁵ was employed to acquire T1W images. Unlike 2D spiral SE brain imaging⁵, flow is not a concern in coronal knee exams. Hence, through-plane crushers were removed and in-plane crushers were minimized to obtain a shorter TE of 8 ms. For similar reasons, the readout for Cartesian TSE T1W sequence was in HF direction unlike PDW and T2W sequences.

Radial Plane imaging
The medial and lateral menisci were imaged in separate scans. In each scan, cross-sectional images were obtained in 9 radial slices that are perpendicular to the meniscus in an interleaved mode^3,4. A low flip angle of 15⁰ was used with a repetition time (TR) of 415 ms to mitigate the vertical low-signal band at the center of each image.

Results and Discussion

The spiral water-only and water-fat combined images are shown and compared with the Cartesian images in Figures 1-4. The overall contrast, SNR and sharpness of spiral PDW and T1W images are comparable to the Cartesian reference images. The spiral images have more textured appearance in bone marrow, which might be partially due to the increased T2* weighting in spiral readouts. The flow artefacts from the popliteal artery in the spiral Dual SE are mitigated, with only minor residues seen in T2W images (Figure 2). Cartesian T2W images have slightly higher SNR than spiral T2 images because, the Cartesian sequence has shorter TE and longer total readout time. It is straightforward to increase the number of spiral interleaves if higher SNR is preferred, at the cost of slightly longer scan-times. Alternatively, 2D spiral TSE¹⁰ with short echo train length (ETL) can be used to improve the SNR efficiency. Spiral images in radial plane demonstrate similar visibility of the menisci compared to the Cartesian scans (Figure 4), with only 38% of scan time.

Conclusion

This work has demonstrated the feasibility of using spiral Dixon imaging for high resolution knee MRI exams.

Acknowledgements

This work was funded by Philips Healthcare.