Diffusion Weighted MRI of the Fibrous Structure of the Meniscus of the Knee: In Vitro Studies at 11.7T
Qun He1,2, Jihye Baek1, Daryl D'Lima1, Jiang Du1, Nikolaus M. Szeverenyi1, and Graeme Bydder1

1University of California, San Diego, San Diego, CA, United States, 2Ningbo Jansen NMR Technology Co., Ltd., Cixi, Zhejiang, China, People's Republic of

Synopsis

In vitro high resolution diffusion weighted imaging of the meniscus of the knee at 11.7T shows detailed structure of fiber groups that has not previously be seen with MRI. There were clearly differentiated from cartilaginous areas of the meniscus. The diffusion contrast adds to that which results from the magic angle effect. This acts as a T2 filter. The diffusion contrast has more flexibility than magic angle contrast since it is not dependent on B0 and in principle any diffusion direction can be used. The results may be translatable to high field clinical systems.

Purpose

To image bovine meniscus using high field diffusion weighted microscopy in order to define the fibrous groups within it so that these features can be related to function and changes in disease.

Introduction

The meniscus of the knee is a complex tissue which distributes mechanical load from the body and thereby protects the adjacent articular cartilage. This is achieved with a mixed cartilaginous and fibrous structure with the cartilaginous regions resisting compression, and the fibrous regions resisting tension. Classical anatomic descriptions of the fibers based on light and electron microscopy are well known1, but the fiber patterns described are not seen with MRI in vivo, and do not correspond well with the patterns seen with in vitro using magic angle imaging. In order to more fully demonstrate the fibrous structure of the meniscus in vitro studies of bovine menisci were performed at 11.7T using diffusion weighted imaging. Hitherto demonstration of meniscus fibers has only been possible using the magic angle effect and this constrains the orientation of the sample and limits the options for achieving useful contrast.

Materials and Methods

Six bovine menisci were harvested and frozen in accordance with institutionally approved protocols. An 11.7T system, Bruker BioSpec (Billerica, MA) with 750 mT/m gradients was used. 2D multislice multiecho (MSME) diffusion weighted (b=1000s/mm2) pulse sequences (TR= 1000 ms, TE=16 ms) of voxel size 60-100 x 60-100 x 600 μm were employed with gradient orientations in the X,Y and Z planes. Specimens were examined in 72 and 38 mm T/R resonators, a four element rat brain array coil and a solenoid (12 mm diameter). The plane of the meniscus was perpendicular or parallel to B0.

Results

A complex network of linear, oblique and curved fibers orientated in different directions was observed (Fig. 1). With B0 and the diffusion gradient parallel to the circumferential fiber direction, fibers running in radial and vertical directions were highlighted. Vertical fibers were seen extending outwards to the lamellar layers (white arrow). The vertical fibers found in the central region of the meniscus were generally thicker. Superficial cartilaginous lamellar regions showed a uniform signal with little or no evidence of fibers (arrowhead).

Overlapping and crossing circumferential fibers were also observed (Fig. 2b). These were orientated in different oblique directions instead of in one circumferential or hoop directions as conventionally described. They formed a meshwork like pattern in the axial plane. Blood vessels showed a low signal on diffusion weighted images (white arrow). On unsensitized images(e.g. Fig. 2a), circumferential fibers were seen, but only a small fraction of these was visible compared to what was seen on diffusion weighted images (e.g. Fig. 2b).

Discussion

In this study collagenous fibers showed contrast due to a combination of the magic angle effect and diffusion weighting. The choice of B0 orientation determines the magic angle effect and as a consequence introduces T2 weighting. With MSME diffusion weighted images of the meniscus, this means that the signal from fibers which are approximately parallel to B0 is filtered out, while that from radial and vertical fibers remains. Diffusion gradient introduces further contrast depending on the intensity and direction of the diffusion gradient. By carefully choosing these two parameters, it is possible to selectively highlight fibers and achieve greater contrast than by use of the magic angle effect alone as shown in the comparison between Fig. 2a and Fig 2b. On the diffusion weighted images, vertical, oblique, circumferential and radial fibers were seen. Complex overlapping and braiding of fiber groups was observed in both the radial and axial planes with high contrast. These fiber groups do not feature in classical illustrations of the fiber structure of the meniscus. Cartilage was most evident in the femoral and tibial lamellar regions and centrally where compressive forces are more significant.

Conclusion

High resolution diffusion weighted MRI provides a powerful contrast mechanism for demonstrating the structure of the meniscus and provides detail of fiber groups which is not available using the magic angle effect alone. Some of the features may be visible using clinical systems operated at higher field strengths.

Acknowledgements

The author acknowledges support from Ningbo Jansen NMR Technology Co., Ltd.

References

1. Petersen W, Tillmann B. Collagenous fibril texture of the human knee joint meniscii. Anat Embryol (Berl). 1998;197:317-324.

Figures

Fig. 1 Bovine meniscus in radial plane, diffusion weighted MSME sequence. B0 is perpendicular to image plane and diffusion gradient is perpendicular to image plane. A complex network of vertical and oblique fibers is observed (a surgically induced tear extending from the femoral surface is also seen).

Fig. 2 Bovine meniscus in axial plane, diffusion weighted MSME sequence. B0 is vertical. a) Unsensitized image b) Diffusion weighted image with gradient perpendicular to image plane. Crossing and braiding of the circumferential fibers is seen in (b).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4494