Advanced MRI of articular cartilage

Abstract

Musculoskeletal MRI is advancing rapidly, with innovative technology and significant potential for immediate clinical impact. In particular, cartilage imaging has become a topic of increasing interest as our aging population develops diseases such as osteoarthritis. Advances in MRI hardware and software have led to increased image quality and tissue contrast. Additional developments have allowed the assessment of cartilage macromolecular content, which may be crucial to the early detection of musculoskeletal diseases. This comprehensive article considers current morphological and physiological cartilage imaging techniques, their clinical applications, and their potential to contribute to future improvements in the imaging of cartilage.

Figure 1. Standard clinical MRI of the knee using 2D fast spin echo T₁, T₂ and proton density weighting for distinct contrast

(A) Depicts a T₁-weighted coronal image with characteristically dark fluid regions. (B) A T₂-weighted, fat-suppressed sagittal image showing relatively dark cartilage (solid arrow) high signal in fluid-filled regions (dashed arrow) and excellent fluid-cartilage contrast. (C) A proton density weighted, fat-suppressed coronal image of the same knee, showing a higher cartilage signal than the T₂-weighted image.

Figure 2. 2D and 3D imaging using fast spin echo

(A) Depicts traditional 2D fast spin echo (FSE) coronal proton density imaging of the knee with high signal-to-noise ratio and impressive tissue contrast. (B) An example of 3D FSE proton density imaging in the coronal plane. 3D FSE allows for isotropic resolution and reformations, but increased blurring and decreased signal from subchondral bone compared to 2D FSE (white arrow) are limitations of this technique.

Figure 3. Sagittal images of the knee using IDEAL

(A) Shows an IDEAL image with both fat and water. (B) IDEAL fat image. (C) IDEAL water image. Arrows denote cartilage and subchondral bone changes characteristic of patellofemoral osteoarthritis.

Figure 4. MRI from a patient with metallic fixation screws in the tibia

(A) An example of a routine clinical scan with significant metal-induced artifact. The same image acquired using 2D FSE is shown in (B). Metal artifact correction with slice-encoding for metal artifact correction (C) and multiacquisition variable-resonance image combination (D) are also shown. These techniques minimize artifact and allow improved visualization of soft tissues surrounding metallic implants. Images courtesy of Christina Chen (Stanford University, CA, USA), Brian Hargreaves (Stanford University) and Kevin Koch (GE Healthcare Applied Sciences Laboratory, WI, USA).

Figure 5. Imaging differences at 3 and 7 T

(A & B) Show sagittal and axial images acquired at 7 T using a 28-channel coil. Markedly increased SNR is observed. (C & D) Representative of the same images acquired at 3 T. Increased field strength and improvements in coil technology allow for improved SNR and shorter scan times. SNR: Signal-to-noise ratio. Images courtesy of Ravinder Regatte (New York University, NY, USA).

Figure 6. A small cartilage fissure in the medial femoral condyle is shown with VIPR ATR images in the sagittal (A) and coronal (B) planes

Images courtesy of Rick Kijowski (University of Wisconsin, WI, USA).

Figure 7. T₂ mapping of medial femoral articular cartilage in (A) healthy volunteer and (B) a patient with Kellgren-Lawrence grade 1 osteoarthritis

The increased T₂-relaxation time is seen in Figure 8B with significantly more red present in the T₂ cartilage map (arrows).

Figure 8. Examples of medial femoral cartilage in T₂, T₁rho and sodium images

(A–C) Standard images, with maps overlaid on (D–F). Regions of increased T₂- and T₁rho-relaxation times are seen with the high degree of red in (D & E). (F) Shows the sodium signal intensity throughout the image.

Figure 9. Comparison of the delayed gadolinium enhanced MRI of cartilage index obtained in a healthy subject (A) compared with a patient with knee osteoarthritis (B)

Index values are lower in Figure 9B, particularly at the medial tibial plateau, which represents decreased glycosaminoglycan content in articular cartilage and medial meniscus degeneration. Scale in ms. Reproduced with permission from [2].

Figure 10. Images of the meniscus acquired using ultrashort echo times at the decreasing echo times: 45 ms (A), 30 ms (B), 15 ms (C), 16 ms (D), 12 ms (E) and 8 ms (F)

Cartilaginous and fibrous components, particularly at the tissue periphery, are unmasked with the use of ultrashort echo times. This is particularly evident when comparing (A & F). Images courtesy of Christine Chung (University of California San Diego, CA, USA).

Figure 11. Diffusion-weighted imaging in patellar cartilage in vivo

(A) Uses high diffusion weighting. (B) Uses low diffusion weighting. Both imaging techniques afford high resolution, high signal-to-noise ratios, and the ability to calculate apparent diffusion coefficients. Images courtesy of Ernesto Staroswiecki and Brian Hargreaves (Stanford University, CA, USA).