One of the most important parameters in tumor imaging is FOV. Although it is applicable to all cross-sectional imaging, it has received relatively little attention until recently [7]. Referring providers often order an MR study by anatomic area. MR imaging for a small mass around the hip may be ordered as a pelvis study or a lesion in the intertrochanteric region as a femur; examinations which are typically performed with a large (36–40-cm) FOV. Imaging the hip, for example, would typically be evaluated with a 20-cm FOV, allowing increased spatial resolution by as much as a factor of four, allowing markedly improved anatomic delineation of adjacent vessels, nerves, fascial planes, and other anatomic structures.

As a general guide, FOV is dictated by the size and location of the mass being studied and must be large enough to evaluate the entire lesion and allow appropriate local staging. Cases which require evaluation of an entire long bone, such as with suspected osteosarcoma, small FOV images can be supplement with limited large FOV images. Similarly, for example, when no discrete lesion is identified at initial imaging and a superficial unencapsulated lipoma is suspected, limited large FOV imaging of both sides is useful for comparison to allow confident identification of asymmetry and a definitive diagnosis.

MR imaging is invariably performed using a multiplanar technique. In the vast majority of locations neurovascular structures and muscles can be optimally identified and followed in the axial plane. Consequently, we generally use this as the primary imaging plane. This is augmented with long-axis images giving the best assessment of the entire lesion in “profile,” assisting in its characterization as well as its relationship to the important surrounding anatomic structures; information required for diagnosis and surgical planning. Lesions such as those of the chest wall may be problematic due to its curved nature, and images in the standard coronal or sagittal plane may be less useful than imaging perpendicular to the chest wall. To avoid confusion during image acquisition, unusual oblique planes must be clearly defined to the imaging technologist.

Most musculoskeletal masses are well evaluated with T1-weighted and fat-suppressed fluid-sensitive images. We generally prefer fat-suppressed proton-density–weighted FSE images to fat-suppressed T2-weighted FSE images due to the increased signal-to-noise ratio of the former. It is important to remember that the main purpose of fat suppression in fluid-sensitive imaging is enhancement of lesion conspicuity. In reality, conspicuity is typically not an issue in tumor imaging, and non-fat-suppressed T2-weighted images may be a useful adjunct for comparison of signal intensity to that of internal controls of muscle (low), fat (intermediate), and fluid (high). This is often helpful in better assessing the nature of the mass; for example, myxoid lesions generally follow fluid signal intensity, whereas collagenous/fibrous lesions generally have low to intermediate signal intensity. As FSE imaging tends to blur the signal intensity difference between fat and water, strongly T2-weighted parameters are required in such cases for non-fat-suppressed T2-weighted sequences (TR > 4000/TE100–110).

Gradient-echo (GRE) imaging has less fluid conspicuity but is particularly useful in identifying evidence of prior intralesional hemorrhage, revealing hemosiderin deposition as a result of its greater magnetic susceptibility [8]. In addition to hemosiderin, gradient-echo imaging is useful in identifying increased susceptibility artifacts from metal and air [8].

We find short τ inversion-recovery (STIR) images especially useful in areas where fat suppression is heterogeneous. Heterogeneous fat suppression can be particularly problematic in evaluation of off-center extremity musculoskeletal masses, as well as in areas where air pockets form adjacent to curves in the extremity or torso, such as adjacent to the perineum. STIR imaging is more susceptible to degradation from patient or respiratory motion and has lower signal-to-noise ratio than spin-echo imaging; however, it can increase lesion conspicuity and is a useful adjunct [9–11].

The use of contrast material for evaluation of musculoskeletal tumors is well established, and its efficacy in improving diagnostic accuracy is well documented [10, 12, 13]. Contrast material is also useful in assessment of the vascular anatomy as well as lesion vascularity [12, 14]. We find contrast material useful not only in giving greater confidence in diagnosis, but also in identifying the most suitable location of the vascularized solid elements that will provide diagnostic tissue from subsequent biopsy if required.

Contrast-enhanced imaging is especially useful in assessment of hematomas and in distinguishing the hemorrhagic component of a neoplasm from the solid elements that may not be easily distinguished on conventional imaging. Subtraction imaging is a relatively recent innovation that has received little attention in the literature [15–17]; however, we find it invaluable in such cases, since this technique eliminates the possibility of misinterpreting the T1 shortening associated with hemorrhagic change as vascular enhancement.

The above MR imaging techniques are applicable for the vast majority of cases; however, the relatively long imaging acquisition times of these standard sequences limit their use in anatomic areas that are susceptible to respiratory motion [10]. Recent advances in software and imaging applications now allow a variety of techniques that can capture a complete set of images in a single breath hold [18]; techniques that not only allow rapid image acquisition but are extremely fluid sensitive. Many musculoskeletal radiologists are not familiar with these techniques, although they are routinely used by our abdominal and thoracic imaging colleagues.

We generally prefer the gradient version of single-shot imaging (true FISP [true fast imaging with steady-state precession], FIESTA [fast imaging employing steady-state acquisition], balanced FFE [fast field echo]), which provides high-signal-intensity flowing blood and increased signal-to-noise ratio. When used without fat-suppression, its mixed T1/T2 weighting can be confusing due to the increased signal intensity from fat, water, and flowing blood; however, its quick acquisition time makes it a valuable adjunct.

Rapid fluid-sensitive images can also be acquired using a spin-echo single-shot technique (HASTE [half-Fourier acquisition single-shot turbo spin-echo], turbo spin-echo, single-shot FSE). These sequences are motion insensitive, ideal for breath-hold imaging, and exquisitely fluid sensitive [19–21]. The most significant limitation of HASTE and similar sequences is the decreased signal-to-noise ratio, typically requiring thicker section thickness and larger-FOV imaging.

Chemical shift imaging is a well-established technique based on the observation that at gradient-echo imaging, the signals from similar quantities of fat and water in a single voxel will reinforce each other when in phase and cancel each other when out of phase cancel each other out [22, 23]. This allows qualitative and quantitative (measured as percentage change in signal intensity on opposed-phase relative to in-phase images) assessment of the amount of microscopic fat in any lesion. Initially used in separating adrenal adenoma from other adrenal masses [24], this technique has proved itself useful in assessment of musculoskeletal osseous masses [25]. In a retrospective study of 50 osseous pelvic lesions Kohl et al. [25] found chemical shift imaging to be highly sensitive in identifying malignant disease and, despite its lower specificity, found its use could have eliminated the need for biopsy in more than 60% of patients with benign disease. Its application for soft tissue is more limited, but it is useful in distinguishing reactive marrow changes from tumor and microscopic fat in higher-grade liposarcoma [26].

Diffusion-weighted imaging has been well documented to be a useful technique in MR imaging of the brain and has more recently received greater acceptance in the musculoskeletal community [27, 28]. Diffusion-weighted imaging allows qualitative and quantitative assessment of the movement of water molecules through tissue. This technique can identify when that normal random motion is restricted, such as by an abnormally increased number and/or size of cells. This restricted motion of water is detected by application of diffusion-probing gradients of different strength (expressed as the b value) and can be visually assessed as signal intensity, with restricted diffusion showing increased signal intensity and measured as the apparent diffusion coefficient (ADC), with restricted diffusion demonstrating decreased signal on ADC maps [22].

Although qualitative assessment of diffusion-weighted imaging readily recognizes restricted diffusion and ADC values are reproducible [29], the quantitative distinction between benign and malignant based on ADC value is not yet well-defined, with variable results reported [30–33]. There is no universally applicable “one size fits all” ADC value. Razek et al. [34] recently suggested a value of 1.34 × 10⁻³ mm²/s as a threshold for distinguishing between benign and malignant masses. Using this value, they obtained a sensitivity, specificity, and accuracy of 94%, 88%, and 91%, respectively. A more recent study by Surov et al. [35] documented the variability of ADC value with tumor type, noting that if a value of 1.34 × 10⁻³ mm²/s were used in their assessment, 35.1% of their malignancies would have been incorrectly categorized. Such results underscore the influence of tumor morphology (cellularity and matrix composition) in diffusion-weighted imaging, as has been previously noted [32, 36]. Although the value of diffusion-weighted imaging in distinguishing benign from malignant lesions is as yet somewhat unsettled, it can identify restricted diffusion, which is also quite useful in identifying small pelvic lymph nodes [18].

Another quantitative method for assessment of soft-tissue tumors is proton MR spectroscopy. Unlike the previous quantitative techniques, there is no real qualitative component. In musculoskeletal spectroscopy, choline, a marker for cell membrane turnover, is measured and is recognized to be elevated in malignancies. Although spectroscopy can provide useful information, it is a technically demanding and time-consuming technique. Accordingly, it is not usually used in routine practice.

As previously noted, CT is particularly useful in assessment and characterization of mineralization, especially in areas where the osseous anatomy is complex. It allows distinction of ossification from calcification and identification of characteristic patterns of mineralization. CT is also superior to radiography in detecting the zonal pattern of mineralization, essential to radiologic diagnosis of early myositis ossificans, a pattern that can be identified at CT, while radiographs remain nonspecific [37].

This is a relatively new technology that has proved itself as a useful adjunct in evaluation of musculoskeletal masses. Using the differences in energy attenuation of soft tissue at 80 and 140 kVp, dual-energy CT allows distinction of urate crystal deposits from other soft-tissue calcifications [38]. Although it may seem that the diagnosis of chronic tophaceous gout should be obvious, such is not invariably the case. The diagnosis can be challenging, especially in those without a history of gout or hyperuricemia, and cases of gout referred to tertiary-care centers as a suspected soft-tissue sarcoma are not rare [39]. Dual-energy CT has also proved to be a useful method to evaluate metal implants by generating images acquired by monoenergetic high energy quanta, reducing metal artifact [40]. Utilization of this technique can significantly reduce metal artifact in the assessment of metal implants, improving the diagnostic value of imaging [40]. Most recently, it has shown application in the assessment of marrow edema [41, 42] and has been investigated in the distinction of marrow edema from intramedullary tumor invasion [43].

The rapid image acquisition capability of modern scanners also allows accurate assessment of lesion vascularity. In a comparison with MR angiography by Mori et al. [44], CT angiogaphy with three-dimensional reconstruction was found equivalent to MR angiography in its ability to demonstrate neurovascular involvement. CT was also superior to MR in its ability to identify cortical/ marrow involvement. Finally, CT may be the most appropriate modality for very large patients and patients with pacemakers when MRI is not feasible or contraindicated.