There is strong interest in measuring cartilage load to assess the causes and treatments of joint disorders such as osteoarthritis. One approach is to map T2 in loaded joints as a surrogate for cartilage deformation due to load^1-3. Open, vertical gap scanners offer the potential to image weightbearing joints, but rapid T2 mapping sequences are required because of participant movement. The effect of loading on T2 in cartilage has not been assessed in weightbearing joints. To assess feasibility of T2 mapping using DESS (double-echo steady state) as a surrogate for deformation of knee cartilage¹ in an open upright MR scanner using the DESS sequence (double-echo steady state). Experiments were performed on a 0.5 Tesla upright MR scanner (MR Open, Paramed, Italy) which allows a full range of poses from supine to sitting to standing (see Figure 1). A normal female volunteer was scanned with a dual-channel knee coil in a supine position (unloaded, or loaded on a single knee with 0.15x body weight using a hanging-weight loading rig), or a standing position. A 3D DESS T2 mapping sequence with global T1 estimator⁴ was implemented to provide full sagittal coverage of the knee joint (matrix=256x256x24, FOV=22x22x12 cm, TE/TR = 6/16 ms, NA=1, scan time= 1min38sec, T1 setpoint for fitting procedure = 1.2 sec). A MSE (multi-spin-echo) sequence⁵ was used to collect gold-standard T2 data in a single slice parallel to a DESS slice passing through a tibial plateau (nonselective composite refocusing pulse, descending/alternating gradient spoilers, 16 echoes, matrix=256x128, FOV=36x18 cm, slice thickness=5mm, NA=2, TR=1sec, echo spacing = 15ms, scan time = 4min21sec, monoexponential fit). All images were postprocessed with a non-local means filter denoising⁶ algorithm.

In regions where cartilage surfaces were in contact, median T2 in both tibial and femoral cartilage was less for standing than for both the unloaded and loaded conditions in the supine posture (Figure 2). Figure 3 shows that the standing-posture T2 distributions (femoral/tibial) in contact areas skews lower than the supine cases; a two-sample Kolmogorov-Smirnov test (threshold p < 0.05) indeed shows that the standing-posture distributions are significantly different than the supine-posture distributions. The decrease in T2 may be due to the consolidation of the cartilage (more immobile water molecules following loading) and resulting efflux of water from the tissue.

The differences in T2 distribution between supine loaded and unloaded were slighter. A 0.15x body-weight load on one leg without preloading the knee was chosen to minimize the chance for participant movement, but may not have applied a large enough load to expect significant changes in T2 values. The supine-posture T2 distributions in the femoral cartilage contact areas were tested to be significantly different from each other, which was not the case in the tibial cartilage.

Visual comparison of the DESS T2 maps and histograms versus the MSE gold-standard T2 data (Figure 4) show similar spatial patterns, with a similar range in T2 values mostly between 10 and 50 ms. However, the correspondence between techniques is poorer in certain regions especially for the standing position in femoral cartilage. This may be due to increased B0 inhomogeneity at the posterior surface of the knee, which may increase off-resonance effects in the non-selective refocusing and thereby increase the MSE signal decay near this region. Another potential source of difference is the higher partial volume effect in the MSE data, which has a 2.7x larger voxel size than the DESS data. The DESS/MSE comparison was performed in a single slice only, since MSE operates in a single slice at a time (due to nonselective refocusing) and is therefore impractical for whole volume coverage (4min21sec per MSE slice, whereas DESS provides full coverage of 24 3D sections in 1min21sec).

To our knowledge, we have demonstrated knee cartilage T2 mapping in a true standing position for the first time, and have shown preliminary evidence that there may be differences between loading the joint in the standing position versus the supine loaded case which is the only option for more conventional closed-bore higher-field MR scanners. A larger decrease in T2 is expected in true weightbearing than in supine loading, due to the higher loads¹. The volumetric DESS T2 maps were acquired with short acquisition time which is critical for imaging weightbearing postures, while maintaining a range of T2 values that were similar to gold-standard MSE T2 results. This initial work paves the way for further cartilage T2 mapping studies in symptomatic subjects in true physiological load-bearing conditions.