Current Status of Imaging of The Musculoskeletal System

Introduction

It can be argued that imaging of the musculoskeletal system has been revolutionised by the advent of magnetic resonance imaging (MRI). Despite this, our armamentarium would be incomplete without plain radiographs, computed tomography (CT), ultrasound (US) and nuclear medicine. The aim of this chapter is to outline the basic principles, strengths and limitations of each technique, and to discuss the role of each tool in current diagnostic strategies.

Conventional Radiography

Despite the technological sophistication of modern cross-sectional imaging, plain radiographs are still a critically important tool in key areas of musculoskeletal medicine. The rapid acquisition of images with high spatial resolution and contrast between soft tissue and bone means radiographs remain the first-line imaging tool for identifying fractures, and characterising arthritis and primary bone tumours.

There are two main diagnostic limitations to the technique. The first results from the planar representation of three-dimensional structures: for instance, intact bones are projected over fractures, marginal joint erosions may not be projected in profile and intact dense cortical bone may obscure medullary tumours. The impact of superimposition can be reduced by ensuring that each point of interest is imaged, as standard, with a minimum of two radiographs performed using projections that are perpendicular to each other. For some sites, such as the clavicle, two perpendicular projections are not possible. For other bones, such as the scaphoid, in which the long axis is curved and oblique to all three orthogonal planes, then between four and six standard projections are advocated. For the staging of complex fractures, or for the detection of fractures in complex structures, multiplanar CT is more convenient and more accurate. For the scaphoid, in particular, where fractures are frequently missed, there is good evidence that patients would be best served with routine early limited MRI, if their plain radiographs are interpreted as normal. Similarly, the detection of marginal erosions in the hand can be increased by performing oblique radiographic projections, but US and MRI are more accurate investigations.

The second major limitation is the relative insensitivity of plain radiographs to bone loss from aggressive bone lesions. Some authors suggest that there needs to be a loss of at least 50% of local bone mass before a lesion can be reliably detected. A normal plain radiograph does not exclude metastases and can be falsely reassuring to the uninitiated. It is for this reason that the standard of care for the staging of multiple myeloma has changed from skeletal surveys to whole-body MRI.

The contrast resolution of soft tissues is limited, but useful information can be gleaned from radiographs both directly and indirectly. Joint effusions and tendons can be identified on lateral projections of the elbow, knee and ankle because of the contrast in attenuation compared with surrounding fat ( Fig. 38.1 ). Soft-tissue calcification ( Fig. 38.2 ) and ossification, both in and outside joints, is often obvious on a radiograph and impossible to see on MRI.

Indirect assessment of ligaments can be performed by positioning the patient in order to stress the suspect structure. Flexion and extension views of the cervical spine assess stability of the atlantoaxial joint, clenched-fist views can be used to assess the integrity of the scapholunate ligament, and the unstable acromioclavicular joint can be distracted by getting the patient to hold a weight.

Weight-bearing is an important technical component in radiographs of joints in the lower limb; in particular, joint space narrowing is more accurately depicted in the knee and the severity of valgus or varus deformities become more apparent. On the other hand, while weight-bearing views of the foot and ankle are required to record angular measures of planovalgus and cavovarus, the diagnosis of these conditions is made by clinical examination.

The diagnosis of significant ligamentous injuries can be made with radiographs by identifying avulsion fractures. Avulsion injuries often produce thin shell-like crescents of bone, adjacent to the origin or insertion of the ligament. These may be easily missed on MRI due to partial volume effects and limited bone oedema, but they are easy to see on a conventional radiograph ( Fig. 38.3 ).

A further relative disadvantage of radiography is risk associated with ionising radiation. For radiographs of the extremity, doses are small, in the order of 0.001 mSv, or the equivalent of 3 hours of background radiation. Often the diagnostic yield is high and therefore the risk to benefit ratio favours the use of radiography. In contrast, the radiation dose for a lumbar spine examination is 1.5 mSv, the equivalent of 6 months of background radiation, and the diagnostic yield is low. Conventional radiographs of the spine have little role to play in identifying sources of back pain, and almost no use in identifying the source of a radiculopathy. For this reason some authorities argue that the only diagnostic indication for plain radiographs of the spine is for the detection of osteoporotic vertebral fractures.

Ultrasound

US has revolutionised the practice of musculoskeletal radiology. In the US room, the operator can take a targeted history, examine the patient clinically, perform the diagnostic US examination, advise the patient, and administer diagnostic or therapeutic treatments as appropriate.

High-frequency linear probes (10 MHz and above) provide soft-tissue images, compiled from reflected echoes at tissue interfaces, with vastly superior spatial resolution compared with MRI. Passive and active movement of joints, muscles and tendons allow a dynamic examination and direct correlation with the patient's symptoms. Impingement of tendons and ligament laxity can be directly visualised and recorded. Hyperaemia can be demonstrated and quantified with Doppler flow. Although it is debatable whether or not US-guided corticosteroid injections are more effective than clinically guided ones—there are a few well-designed studies comparing the two that demonstrate marginal or no benefit to US-guided injections—the real advantage comes from the diagnostic component of the injection. If a patient fails to respond, even temporarily, to an US-guided local anaesthetic injection of the subdeltoid bursa this means that rotator cuff impingement is unlikely to be a source of the patient's symptoms. It is not possible to be so certain with a clinically guided injection.

US is a useful first investigation for patients presenting with a soft-tissue mass ( Fig. 38.4 ). The presence of the mass can be confirmed and, if present, non-aggressive cystic lesions, nerve sheath tumours, vascular malformations, haematomas and bursae can often be diagnosed with confidence. US-guided biopsy of aggressive, or non-specific lesions, once the patient has been reviewed at a multidisciplinary sarcoma meeting, can maximise the yield from percutaneous biopsy by avoiding areas of obvious tumour necrosis and ensure that adjacent compartments are not violated.

While US is limited as a tool for assessing bone, it is excellent at assessing the periosteum and cortex for identifying fractures and aggressive superficial bone lesions ( Fig. 38.5 ).

The main limitation of US in the musculoskeletal system is the absorption of the high-frequency beam at depth. For intervention, this can be overcome by using a low-frequency curvilinear probe, which has the advantage of a larger field of view and better visualisation of the needle at lower frequencies. This is satisfactory for hip injections and biopsy of deep lesions, but the lower frequency, and therefore resolution, may limit characterisation of the lesion.

Anisotropy is a US artefact affecting the imaging of tissues with a parallel linear fibrillar structure. Such tissues include tendons, ligaments and muscle, so this is a frequent pitfall in musculoskeletal US. If the US beam is not perpendicular to the orientation of the bundles of fibres in these tissues, a hypoechoic appearance results, which can be misinterpreted as degeneration or a tear. An experienced operator recognises this artefact, which can be avoided through the use of scrupulous US technique.