Medical professionals, researchers or MRI technologists with an interest in quantitative MR imaging and image processing of collagen rich musculoskeletal (MSK) tissues such as articular cartilage, meniscus, tendons, ligaments and bone.

By the end of this presentation participants will:

1. Be able to describe the structure and function of collagen rich MSK tissues

2. Understand how collagen presence affects quantitative MRI

3. Know which quantitative MRI parameters are used to study MSK tissues

4. Understand some technical considerations of processing quantitative MRI data

Collagen is an integral component of several MSK tissues including articular cartilage, meniscus, tendons, ligaments and bone. Both the amount and organization of the collagen affects the tissue’s function and MRI relaxation properties. Collagen is a short string-like protein that provides structural support to tissues, analogous to the steel bars (rebar) in reinforced concrete. This highlights the great importance of collagen because the major functions of MSK tissues are to facilitate motion and transmit forces. Collagen itself has a very short T₂ relaxation time, making it challenging to image. Its proportional amount and orientation dictate the overall T₂ relaxation time of the particular tissue. The magic angle effect must be considered when imaging MSK tissues because orientation of collagen fibers at approximately 55° from the main magnetic field results in an increase in the T₂ relaxation time in that region¹.

The orientation of collagen in MSK tissues not only matters for MRI imaging but also for overall function. The orientation is directly related to how forces are transmitted through the tissues. Therefore, by understanding this relationship one can easily remember the collagen orientation.

Articular Cartilage: Articular cartilage covers the surfaces of long bones; it provides a smooth bearing surface that allows near frictionless motion and resists both compressive and shear forces. It is primarily composed of collagen, proteoglycans and water and is often described as having three zones. The superficial zone, approximately top 20%, has the highest proportion of collagen content (about 85% dry weight)² and the collagen fibers are oriented tangentially to resist shear forces. In the middle zone (approximately mid 50%)2, the collagen fibers are randomly oriented and, along with proteoglycans and osmotic swelling pressure, resist compressive forces. Finally, the deep zone (approximately bottom 30%)2, the collagen fibers are oriented perpendicular to the bone-cartilage interface and act to anchor the cartilage to the bone.

Meniscus: The menisci of the knee improve congruence between the femur and tibia, increasing the area over which forces can be transmitted. They have a semi-lunar shape and a wedge-like cross section. Collagen makes up over 80% of the dry weight with most fibers oriented in the circumferential direction². This organization generates hoop stresses in the meniscus when axial loads are applied.

Tendons: Tendons attach muscle to bone and therefore must transmit the large forces generated by muscle. They are predominantly composed of highly aligned collagen fibers along the length of the tissue.

Ligaments: Ligaments attach bone to bone and serve the purpose of guiding motion. They are also composed of aligned collagen fibers but are more randomly oriented than tendons. This means they have more ‘stretch’ than tendons.

Bone: The role of bone is to facilitate motion and provide protection to organs and tissues. While bone is composed of mainly minerals, it also contains a great amount of collagen (approximately 45% by weight)².

It should be noted that collagen content and structure may be altered with disease. For example, in articular cartilage, collagen content decreases and remaining collagen becomes disorganized osteoarthritis. This highlights that collagen is an important biomarker of tissue health and is why imaging collagen with MRI is so important in MSK.

Collagen is the most common protein in the body³. Each collagen molecule consists of three polypeptide chains that together form a triple helix. Collagen is quite stiff, making it an ideal reinforcing material. Water is present both between and outside the chains; not surprisingly, it has been shown that the outer water protons are considerably more mobile than the inner water protons⁴. This results in complex, multicomponent relaxation behaviour.

We must remember that collagen is just one component of MSK tissues; thus, relaxation behaviour becomes even more complex. For example, as mentioned above articular cartilage is composed of collagen, proteoglycan and water and it has been shown to have three components at approximately 2.3, 25.2 and 96.3 ms⁵. If we fit a single T₂ relaxation time using a mono-exponential fit, as is commonly done in MSK work, we are obtaining an overall T₂ relaxation time not specific to the collagen in the tissue. Because of differences in content, relaxation times vary greatly between MSK tissues. For example, at 3 Tesla, approximate T_2* relaxation times are in the range of 1, 2, 10, and 20 ms, for bone, tendon, meniscus and cartilage, respectively^6-9.

When collagen is oriented at approximately 55° from the main magnetic field (B₀), an erroneous increase in T₂ is observed; this is called the Magic Angle Effect. It is most commonly observed in T₂-weighted scans of the knee in the posterior femoral condyle cartilage region. The magic angle can be a source of artifact in T₂, T_2* and T_1ρ relaxation time mapping but can also be an advantage when trying to determine the orientation of collagen fibers in a tissue^{1, 10}.

Quantitative MRI is proving to be very useful in research studies of MSK diseases such as osteoarthritis. T₂, T_2* and T_1ρ relaxation times are higher in the cartilage and meniscus of patients with osteoarthritis and ACL injury^{6, 11, 12}. While standard quantitative MRI sequences can be used to study cartilage, ultra-short echo time sequences are required for studying bone, tendons and ligaments. Ultra-short echo time sequences can be used to study the meniscus; however, modified standard sequences are likely also sufficient given its ~10ms relaxation time. Other quantitative MRI approaches such as sodium MRI, susceptibility mapping and magnetization transfer are also valuable in studying MSK tissues, although their use has been limited to date^13-15.

Image processing is one of the greatest challenges of quantitative MRI of MSK tissues. The first step is ensuring that the source images have sufficient SNR for the subsequent analysis; it has been shown that this value is approximately 20¹⁶. The next step is segmenting the desired tissues from the source images. With three-dimensional sequences we obtain a vast amount of data but extracting it can be very difficult since no reliable automated segmentation tool exists for MSK. As a result most studies analyze a single slice or a few slices in each tissue or plate; however, by doing this it is possible that a diseased focal region is excluded from the analysis. Next, a signal model is used to generate the quantitative parameter maps. It is essential to assess the quality of the data fitting procedure; often this is not reported but it should be carried out and it would be very useful if it were reported. Once a map is obtained again, there is a vast amount of data. Most often mean values for the structure or sub-region are calculated; however, some groups do carry out texture or cluster analyses to isolate patterns or defects within the tissues^{17, 18}. While this is not an exhaustive list of post-processing factors to consider, it does highlight some main challenges currently faced by researchers carrying out quantitative MRI research of collagen rich MSK tissues.The presence of collagen complicates quantitative MRI assessments of MSK tissues but it also plays a very important functional role. As long as images are acquired, processed and interpreted with care, valid and reliable data can be obtained for research studies.