Osteoarthritis (OA) is a debilitating disease that is a burden on individuals and society as a whole[1]. However, there are no reliable methods to detect the early onset of disease or track slight disease changes over time in vivo.

OA has three main stages of progression in cartilage: (1) glycosaminoglycan depletion; (2) a breakdown of the collagen matrix; and (3) the cartilage begins to fissure and loose volume and thickness. Measures that track changes in each of these stages of OA will enable a better understanding of the disease and enable studies into treatments of the disease. We aim to measure all three of these changes in a single protocol.

MRI Sequences exist that show promise in measuring each of these stages. However, they have not previously been combined into a single protocol. Here we present a protocol that measures cartilage glycosaminoglycan content using T1ρ, cartilage matrix integrity using T2 and ADC measures derived from a modified double-echo in the steady-state (DESS) sequence, and cartilage volume, thickness, and surface defects using a phase-cycled balanced steady state free precession (bSSFP) sequence for high resolution morphology.

All images were acquired with a Siemens (Erlangen, Germany) 7 Tesla MRI machine with a 28 channel QED (Ohio, U.S.A.) knee coil.

Sequence parameters are summarized in table 1. Further method details follow.

T1ρ for glycosaminoglycan evaluation

A 3D T1ρ prepared spoiled gradient echo sequence was used to acquire T1ρ maps in cartilage[2]. T1ρ maps were calculated offline using MathWorks (Massachusetts, U.S.A.) MATLAB. Sequence parameters are found in Table 1. An example image is shown in Figure 1.

Modified DESS for collagen matrix evaluation

A 3D modified DESS sequence was used to acquire images at two different flip angle/diffusion gradient settings resulting in four images (two images for each set of sequence parameters)[3]. Sequence parameters are found in Table 1. T2 and ADC maps were calculated using an extended phase graph signal model[4] in MathWorks (Massachusetts, U.S.A.) MATLAB. An example T2 map and ADC map overlaid on a DESS image are shown in Figure 2 and 3 respectively.

Phase-Cycled bSSFP for morphology measurements

A bSSFP sequence was modified to enable phase cycling. Four different images were acquired at different phase cycling to shift the banding artifact. All four images were combined in MATLAB using a root sum of squares combination method to reduce the appearance of banding. Sequence parameters are found in Table 1. An example image is shown in Figure 4.

Initial images are promising in the proposed whole knee imaging protocol to assess OA in the cartilage of the knee.

T1ρ measurements are consistent with those previously reported. However, the current T1ρ sequence does not sufficiently cover the knee in the desired scan time (unilateral knee coverage is desired in about 30 minutes). Other studies have acquired data in the coronal and axial planes, effectively doubling the scan time, to cover patellar cartilage as well as femoral condyle cartilage. The proposed protocol focuses on the femoral condyle cartilage regions at the cost of missing the patellar cartilage. Further work will be extending the coverage of the T1ρ sequence through sequence development of this and the other sequences without sacrificing the timing of the protocol.

Modified DESS measurements of T2 are consistent with those previously reported. The ADC measurements do not directly agree with standard diffusion measurements. However, further work is being done to validate the T2 and ADC measurements along with improving the ADC estimates.

The phase cycled bSSFP images provide high resolution morphological images. Different layers of cartilage can even be discerned in some areas. Future work is to enable semi-automated segmentation using these images. The anisotropic resolution is to facilitate better segmentation.

Here we present a protocol for early OA cartilage evaluation at 7T in about 30 minutes. This protocol may be used in studies to better understand early OA in the knee.