Magnetic Resonance Imaging of the Zone of Calcified Cartilage in the Knee Joint Using 3-Dimensional Ultrashort Echo Time Cones Sequences
Introduction Osteoarthritis (OA) is a prevalent joint disorder affecting approximately 360 million people worldwide, with 50% of individuals aged 60 years and older experiencing its symptoms. The early stages of OA are often clinically silent, making early diagnosis challenging. Articular cartilage, which plays a crucial role in joint function, is composed of four distinct zones: the superficial zone, the transitional zone, the radial zone, and the zone of calcified cartilage (ZCC). The ZCC serves as a critical interface between articular cartilage and bone, facilitating solute transportation and force transmission. Age-related changes in the ZCC, such as increased mineral content and thickness, have been implicated in the degeneration of adjacent non-calcified cartilage. Additionally, micro-cracking and vascular invasion of the ZCC are believed to contribute to the pathogenesis of OA. Given that pathological changes in the ZCC are reversible in the early stages of OA, direct imaging and quantification of the ZCC could be invaluable for early diagnosis and treatment monitoring.
Magnetic resonance imaging (MRI) is a widely used technique for evaluating knee articular cartilage, particularly in assessing surface irregularities and cartilage thickness. However, conventional MRI sequences are ineffective for imaging the ZCC due to its high mineral content and intrinsically short T2 relaxation times, which result in no or low signal. The advent of ultrashort echo time (UTE) MRI has revolutionized the imaging of short-T2 tissues, including cortical bone, menisci, ligaments, tendons, and calcifications. UTE sequences minimize signal decay during acquisition, enabling the visualization of tissues with short T2 relaxation times. Recent studies have demonstrated the potential of UTE MRI for imaging the ZCC, with spectroscopic UTE imaging (UTESI) estimating T2 values in the ZCC to be between 1 and 2 ms in both volunteer and cadaveric knee joints. However, imaging and quantifying the ZCC in the whole knee remain challenging due to the need for high spatial resolution and the time-consuming nature of UTE MRI.
This study aimed to develop a 3-dimensional ultrashort echo time cones (3D UTE-Cones) imaging protocol for whole knee ZCC imaging on a clinical 3T scanner. The study evaluated the feasibility of direct imaging of the ZCC in both cadaveric knee specimens and in vivo healthy knees, with the goal of quantifying T2* relaxation times in the ZCC.
Methods Ethical approval was obtained for this study, and written consent was provided by the volunteers. The study included 12 cadaveric knee joints (5 males, 7 females; mean age of 47.85 ± 22.21 years) and 10 in vivo healthy knees from 10 volunteers (6 males, 4 females; mean age of 32.90 ± 8.39 years). The cadaveric knee joints were obtained in accordance with protocols approved by the Committee for the Oversight of Research for the Dead (CORID) and the Institutional Review Board (IRB). All samples underwent a single freeze-thaw cycle and were stored at -80°C.
Imaging was performed on a 3T whole-body GE scanner equipped with an 8-channel transmit-receive knee coil. A fat-saturated dual-echo 3D UTE-Cones sequence was used to image the ZCC. The sequence employed a short rectangular pulse excitation followed by 3D spiral trajectories with conical view ordering, providing isotropic imaging with high in-plane resolution and thicker slices. The imaging parameters included a field of view (FOV) of 8 cm × 8 cm × 3 cm, a matrix size of 256 × 256 × 30, a flip angle of 10°, a bandwidth of 125 kHz, and a repetition time (TR) of 90 ms. For the cadaveric knee joints, six groups of dual-echo UTE-Cones sequences with different echo times (TEs) (0.1/6.6, 0.2/6.6, 0.4/6.6, 0.6/6.6, and 2.2/6.6 ms) were used for T2* measurement, with a total scan time of 20 minutes. For the volunteer knees, the scan time was 10.6 minutes, using three different echo time groups (0.032/4.3, 0.3/6.6, and 0.8/8.8 ms). Echo subtraction images were generated by subtracting the second echo image from the first echo image to enhance short T2 contrast.
Conventional MRI sequences, including proton-density weighted fast spin echo (PD-FSE), T1-weighted fast spin echo (T1-FSE), fat-saturated T2-weighted fast spin echo (T2-FS-FSE), and Carr-Purcell-Meiboom-Gill (T2 CPMG) sequences, were also performed for comparison.
Image analysis was conducted using a semi-automated MATLAB code developed in-house. Regions of interest (ROIs) were delineated within the inner edge of the ZCC to minimize partial volume effects. Single-component T2 and T2 values were calculated from the fat-saturated 3D dual-echo UTE-Cones and T2 CPMG data. The mean, standard deviation (SD), 95% confidence intervals (CI), and standard error of the mean (SEM) of T2 values were recorded for analysis.
Results The clinical FSE and CPMG sequences showed no signal in the ZCC due to its short T2 relaxation time. The T2-weighted CPMG sequence revealed T2 values of 35.63 ± 7.54 ms in the deep layer of articular cartilage. In contrast, the 3D UTE-Cones sequence generated high-quality MR images that adequately sampled the signal decay pattern of the ZCC. The single-exponential fitting curve of the ZCC was accurately obtained with an R2 value of 0.989 on average from 12 cadaveric and 10 volunteer knee joints.
For the cadaveric knee joints, the ZCC exhibited a short T2 ranging from 0.62 to 2.55 ms, with a mean ± SD of 1.49 ± 0.66 ms and a 95% CI of 1.20–1.78 ms. For the volunteer knees, the short T2 ranged from 0.93 to 3.52 ms, with a mean ± SD of 2.09 ± 0.56 ms and a 95% CI of 1.43–2.74 ms. The SEM of T2* values in the ZCC ranged from 0.38 to 0.93 ms for the volunteer knees and from 0.53 to 1.21 ms for the cadaveric knee joints.
Discussion The 3D UTE-Cones sequence offers several advantages over traditional 2D UTE sequences for imaging the ZCC. First, the 3D UTE-Cones sequence uses a short rectangular pulse for signal excitation, which is less prone to eddy current artifacts compared to the slice-selective half-pulse excitation used in 2D UTE sequences. Second, the 3D UTE-Cones sequence allows for high-resolution volumetric imaging and quantification of the ZCC in the whole knee, overcoming the limitations of single-slice 2D UTE sequences, which may suffer from out-of-slice long-T2 signal contamination and partial volume effects. Third, the 3D UTE-Cones sequence provides higher signal-to-noise ratio (SNR) due to its volumetric coverage, which is crucial for accurate quantification of the ZCC given its low proton density.
The 3D Cones k-space trajectory used in this study offers better SNR performance and reduced scan time compared to conventional 3D radial k-space trajectories. In this study, high SNR ZCC imaging was achieved with total scan times of 10.6 minutes for volunteers and 20 minutes for cadaveric knee joints. The preliminary results demonstrate that the 3D UTE-Cones sequence, combined with a multi-echo acquisition strategy, enables volumetric mapping of T2* relaxation times in the ZCC of whole knee joints on a clinical 3T scanner.
The T2 values obtained in this study are consistent with previous findings, although the T2 values in the cadaveric knee joints were slightly lower than those reported in earlier studies. This discrepancy may be attributed to the older age of the cadaveric samples, many of whom had suffered from OA. The decrease in T2 values in OA can be explained by the hypermineralized ZCC, which is harder than neighboring subchondral bone and may accelerate wear rates and alter loading patterns, contributing to joint tissue destruction. Additionally, the loss of water trapped within collagen fibrils in the ZCC may result in a relative increase in shorter T2 component intensities, further reducing T2 values.
This study has several limitations. First, no histology or polarized light microscopy (PLM) was performed to correlate with the quantitative 3D UTE-Cones findings. Future studies should investigate the relationship between imaging results and histopathological scores. Second, the magic angle effect, which influences T2 values due to the different orientation of collagen in various areas of the ZCC, was not investigated. Measurements in different areas of the whole knee joint could help elucidate the angular dependence of T2 values. Third, although the 3D UTE-Cones sequence provides high image resolution, partial volume effects may still exist. Finally, the total scan times of 10.6 minutes for volunteers and 20 minutes for cadaveric knee joints may be too long for clinical applications. Further optimization of the imaging protocol, including thinner slices, advanced reconstruction techniques, and high-performance localized coils, is needed to reduce scan time and improve image quality.
Conclusion The high-resolution 3D UTE-Cones sequence is a promising technique for direct imaging and quantification of the ZCC in whole knee joints on a clinical 3T scanner. This non-invasive method allows for the detection of T2 relaxation times in the ZCC, which may be useful for the early diagnosis of articular cartilage degeneration and OA, as well as for monitoring treatment. Further research is needed to optimize the imaging protocol and explore the potential of UTE-T2 mapping as a tool for detecting early changes in articular cartilage.
doi.org/10.1097/CM9.0000000000000103
Was this helpful?
0 / 0