Radiography Limitations and Pitfalls

Abstract

Radiography is often the first-line imaging modality in the workup of musculoskeletal conditions. Radiographs are widely available, inexpensive, and well tolerated and can be rapidly and easily obtained. Although advanced imaging modalities such as magnetic resonance imaging and computed tomography are still often required for more detailed assessment of structures such as bone marrow and various soft tissues, radiographs still play an important complementary role. As with any imaging modality, radiography has inherent limitations. For example, soft tissue injuries are not accurately evaluated on the radiograph due to poor soft tissue resolution. Radiography is also limited in the assessment of conditions such as early osteomyelitis and undisplaced acute fractures, both of which can be radiographically occult. Certain scenarios make it challenging or impossible to accurately interpret radiographs, for example, when an external cast obscures the bone or when a background of osteopenia results in a paucity of osseous detail. Other potential pitfalls in relation to radiographic technique, patient positioning, and anatomical area of coverage can be encountered.

1 Introduction

It is widely acknowledged that radiography should be the initial imaging modality in the evaluation of most suspected musculoskeletal lesions. They are often adequate for diagnosis, although advanced imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI) are still often required for more detailed assessment of structures such as bone marrow and various soft tissues. Nevertheless, radiographs play an important complementary role to these newer cross-sectional imaging techniques. For example, they can provide a big picture view of bony or joint abnormalities, which allows better assessment of conditions such as the inflammatory arthritides (e.g., by showing distribution and pattern of joint involvement). Radiographs can also demonstrate calcifications and ossifications, which might not be convincingly seen on MRI. In fact, radiography still remains the most specific imaging modality for the diagnosis of bone tumors.

Radiographs are commonly obtained for acute musculoskeletal trauma, infection, chronic arthropathies, and bone or soft tissue tumors. They are also performed for follow-up imaging after treatment such as fracture fixation and joint replacement. Radiographs are widely available, inexpensive, and well tolerated and can be rapidly and easily obtained. As with any imaging modality, radiography has advantages and disadvantages. Radiographs are of limited value in the evaluation of soft tissue injuries, for example, musculotendinous, cartilaginous, or ligamentous injuries. These soft tissue structures are not clearly seen on the radiograph and are better assessed on MRI or ultrasound (US) imaging, both of which have superior soft tissue resolution. Radiography is also limited in the assessment of conditions such as osteomyelitis and non-displaced acute fractures, both of which can be radiographically occult. Certain scenarios make it challenging or impossible to accurately interpret radiographs, for example, when an external cast obscures bone or when a background of osteopenia results in a paucity of osseous detail. Apart from the aforementioned intrinsic limitations of the radiographic modality, other potential pitfalls in relation to radiographic technique can be encountered. Proper positioning of the patient is crucial in obtaining a radiograph of diagnostic quality. An adequate coverage of the area of interest and an adequate number of views are also necessary for proper evaluation.

2 Pitfalls Related to Radiographic Image Acquisition

Radiography is often the initial modality used in the imaging workup of patients with musculoskeletal complaints. It is the workhorse in the emergency department, where it is able to support the high patient throughput. Despite the time pressure, the radiographer has to be meticulous during the acquisition of radiographs as a multitude of potential pitfalls may occur, limiting the accuracy of the evaluation and ultimately negatively impacting upon the clinical management of patients.

2.1 Adequacy of Coverage and Views

When radiographic imaging is requested, one of the fundamental responsibilities of the radiologist is to ensure adequate coverage and sufficient views of the anatomical region of interest. For example, a full radiographic series of the cervical spine should include an anteroposterior view, a lateral view, and an open-mouth odontoid view. On the lateral view, the entire cervical spine should be visualized with the base of the skull seen superiorly and the cervicothoracic junction (C7-T1 level) seen inferiorly. If the routine lateral view is insufficient, attempts should be made to better visualize the cervicothoracic junction (e.g., performing a swimmer’s view) (Fig. 1.1). Missing a significant injury as a result of inadequate coverage on imaging evaluation is virtually indefensible in the court of law. The example of the cervical spine radiograph is particularly pertinent due to the medicolegal implications of a missed unstable cervical spine injury, which can result in devastating neurological sequelae. CT has been shown to have superior sensitivity and has largely superseded radiography in the detection of cervical spine injuries in patients who have high risk of injury (Holmes and Akkinepalli 2005). Obtaining high-quality radiographs with adequate coverage tends to be challenging in these patients due to difficulties in positioning. However, radiography is still used for screening low-risk patients with an indication for imaging, and adequate coverage of anatomy remains crucial.

Fig. 1.1

A 47-year-old man who presented with neck pain following a motor vehicle accident. (a) Lateral radiograph of the cervical spine shows inadequate coverage of anatomy. The cervical spine inferior to the C5 level, as well as the cervicothoracic junction, is obscured by the shoulder girdle. (b) A swimmer’s view was performed as a complementary study and ensured complete coverage of the region of interest

Another potential pitfall is the failure to cover separate associated anatomical regions which may be involved while imaging the primary area of interest. For example, in a patient with injury to the medial ankle structures, a Maisonneuve injury may be missed if imaging does not include the proximal fibula (Pankovich 1976) (Fig. 1.2). Associated injuries should be suspected based on the injury mechanism and the imaging of the relevant area obtained, if indicated. In another example, a calcaneal fracture is usually due to an axial loading force and should raise the suspicion of spinal injury, especially at the thoracolumbar junction.

Fig. 1.2

An 87-year-old woman who presented with right ankle pain after a fall. (a) Frontal radiograph of the right ankle shows a spiral fracture of the distal tibia. (b) Frontal radiograph of the right leg shows a proximal fibular fracture (arrow). This Maisonneuve injury would have been missed if imaging coverage of the proximal leg was not performed

In general, orthogonal views are sufficient in the radiographic imaging of the axial and appendicular skeleton (Fig. 1.3). However, additional views may be required based on the complexity of the anatomy, especially if the structure of interest has a complex shape. Recognition that there are insufficient views can help to identify this pitfall and prevent potential missed diagnoses. In the knee, for example, a skyline view may demonstrate an avulsion fracture of the medial aspect of the patella in relation to transient patellar dislocation (Fig. 1.4). This finding would have been missed on routine anteroposterior and lateral views of the knee. In the case of the scaphoid bone, the standard posteroanterior and lateral radiographs of the wrist are usually insufficient for this complex-shaped bone, with additional views needed for adequate evaluation. Although previous studies show varying recommendations on the number and specific views required in the scaphoid series, a posteroanterior view of the wrist with ulnar deviation and slight tube angulation is usually part of the imaging series (Malik et al. 2004; Shenoy et al. 2007; Toth et al. 2007) (Fig. 1.5).

Fig. 1.3

A 31-year-old woman who presented with left lateral ankle pain. Frontal and lateral radiographs of the left ankle were obtained. A minimally displaced distal fibular fracture is seen as a lucent line only on the lateral radiograph (arrow). It is occult on the frontal radiograph. This common injury illustrates the importance of having sufficient radiographic views for proper evaluation

Fig. 1.4

An 18-year-old man who presented with left knee pain after a collision with another player during a football match. (a) Frontal and (b) lateral radiographs of the left knee show no fracture or dislocation. (c) Skyline view of the knee, however, shows a bony fragment at the medial aspect of the patella (arrow), suggestive of an avulsion fracture. (d) Axial fat-suppressed T2-W MR image of the left knee shows marrow edema at the site of the avulsion fracture (arrow) and disruption of the medial patellofemoral ligamentous structures (arrowheads), which are typical features of a transient patellar dislocation. Associated kissing bone contusion was present at the lateral femoral condyle (not shown)

Fig. 1.5

A 40-year-old man who presented with left wrist pain after falling onto his outstretched left hand. (a) Frontal and (b) lateral radiographs of the left wrist show no fracture or dislocation. (c) Posteroanterior view of the left wrist taken in ulnar deviation and slight tube angulation, usually part of the scaphoid imaging series, shows an undisplaced fracture of the waist of the scaphoid (arrow)

Sometimes, stress views are warranted for evaluation of ligamentous injuries. Examples include the clenched-fist view for assessment of scapholunate ligament integrity and the weight-bearing view for assessment of the coracoclavicular ligament. These views may demonstrate widening of the scapholunate interval and coracoclavicular distance, respectively, which indicate significant injury (Lee et al. 2011; Eschler et al. 2014). Without stress views, these injuries would likely be radiographically occult and hence be difficult to detect.

2.2 Radiographic Technique and Positioning

Radiographic technique refers to the selection of exposure factors, the kilovolt peak (kVp), and the milliampere-second (mAs), which determine the properties of the X-ray beam. The kVp influences the penetrative ability of the X-ray beam, while the mAs influences the quantity of radiation delivered. Good radiographic technique requires the proper selection of exposure factors such that there is optimal beam penetration of the anatomy, as well as optimal quantity of radiation reaching the detector at a radiation dose which is as low as reasonably achievable (ALARA).

In recent years, the field of diagnostic radiology has seen the transition from film/screen radiographic systems to digital imaging. For the film/screen systems, whether a film was under- or overexposed was easily appreciated, since the image would either be too white or too dark, respectively. In digital systems, however, the wide dynamic range of the detector and the ability to automatically post-process images to achieve optimal brightness allow images of acceptable quality to be produced over a larger range of exposures as compared to the film/screen systems (Murphey et al. 1992). Digital radiography is thus more forgiving with suboptimal exposures and has significantly reduced the number of rejected films. Even with the advantages of digital radiography, radiographic technique still affects the image quality and the radiation dose to the patient. Underexposure results in a reduction of the signal-to-noise ratio and manifests as increased quantum mottle, which might render the image unsuitable for diagnostic interpretation (Fig. 1.6). Conversely, overexposure in the digital system does not affect image quality but delivers a larger quantity of radiation than is necessary, resulting in excessive radiation dose to the patient. An optimum exposure should be given with each radiographic study, in line with the ALARA principle and producing an image of sufficient quality.

Fig. 1.6

A 29-year-old man who presented with low back pain following a motor vehicle accident. (a) Frontal and (b) lateral radiographs of the lumbar spine were obtained. Underexposure is evident, as a result of suboptimal radiographic technique. There is increase in quantum mottle especially on the lateral view, causing loss of radiographic detail and affecting the diagnostic quality of the study. This case illustrates how inappropriate selection of exposure factors can result in a poor-quality radiograph

Each digital radiographic system is unique, due to differences in design and detector type. The optimal exposure settings are thus unique to each system. A technique chart should be available for each radiographic system, containing specific optimal exposure settings for each radiographic position in every region of the body. Exposure adjustment systems should then be applied to fine-tune the exposure settings based on the patient’s weight and thickness of the body part to be imaged, the methods of which are beyond the scope of this chapter (Ching et al. 2014).

Standardized positioning of the patient results in radiographic views which are reproducible and optimal for interpretation. If proper positioning is not achieved during image acquisition, alignment of bones might not be accurately evaluated. For example, the posteroanterior view of the wrist should be obtained with the wrist in a neutral position, as only with proper positioning can the ulnar variance be accurately demonstrated. Supination of the forearm decreases ulnar variance, while pronation increases ulnar variance (Epner et al. 1982) (Fig. 1.7). Another example is in the radiographic evaluation of the ankle mortise. Accurate assessment of the alignment of the ankle mortise requires internal rotation of the leg and is not accurately assessed on standard anteroposterior views of the ankle (Takao et al. 2001).

Fig. 1.7

Poor positioning affects the radiographic orthogonal views of the wrist in a 24-year-old man. (a) Posteroanterior and (b) lateral radiographs of the right wrist show inappropriate positioning of the wrist on the lateral view. After acquisition of the standard posteroanterior view, the wrist was left on the imaging plate with forearm supination performed for acquisition of the lateral view. Note the resultant identical appearances of the ulna on both images, making this a suboptimal radiographic study due to the lack of proper orthogonal views. (c) Posteroanterior and (d) lateral radiographs of the right wrist in another patient show appropriate positioning of the wrist, which remains in neutral position on both views. The ulnar variance is also influenced by the position of the wrist, increasing on pronation and decreasing on supination. Ulnar variance is therefore only accurately assessed on the standardized posteroanterior view of the wrist in the neutral position. It is noteworthy that dynamic stress maneuvers can also alter bony alignment. For example, the (e) frontal and (f) clenched-fist stress radiographs of the left wrist of a 40-year-old woman show an increase in ulnar variance on the stress view, which is a normal finding

2.3 Radiographic Artifacts

Various technical artifacts can be produced during the acquisition and processing of radiographs (Shetty et al. 2011; Walz-Flannigan et al. 2012). Detailed discussion of this subject is beyond the scope of this chapter, but these artifacts are generally obvious and do not pose clinical diagnostic problems. Occasionally though, with relevance to musculoskeletal imaging, technical artifacts secondary to dirt or dust may mimic foreign bodies. Encountering any of these technical artifacts would usually trigger a repeat of the radiographic examination. Nevertheless, a properly trained radiographer will be able to prevent many of these technical artifacts. Artifacts which are external to the patient may cause the underlying anatomical structures to be obscured. An external cast on a postreduction radiograph is a frequently encountered example. X-rays have difficulty penetrating the cast, resulting in reduced radiographic detail of the underlying bones (Fig. 1.8). Alignment of fractures and assessment of fracture healing can therefore be difficult to assess on these limited postreduction radiographs.

Fig. 1.8

A 66-year-old woman who presented after falling onto her outstretched left hand. (a) Frontal radiograph of the left wrist shows fractures of the distal radius and ulnar styloid process. (b) Postreduction frontal radiograph shows the presence of a backslab, which obscures the bony details

3 Limitations of Radiographic Imaging of Non-osseous Structures

One of the intrinsic limitations of radiography is its poor soft tissue contrast. Delineation of soft tissue structures on the radiograph may be difficult or impossible, making radiography a generally unreliable imaging modality for the assessment of soft tissue lesions. Unless there is gross morphological alteration (e.g., tendon rupture) (Fig. 1.9) or typical pattern of calcification or ossification (e.g., myositis ossificans) (Fig. 1.10), soft tissue abnormalities invariably go unnoticed on the radiograph. Radiography is generally an insensitive imaging modality compared to other cross-sectional imaging modalities, such as MRI and US imaging, in the early stages of disease processes involving the soft tissues. Early radiographic signs tend to be subtle and easily overlooked, whereas in advanced disease, osseous changes are usually readily observed. However, at the stage when such typically irreversible advanced osseous changes take place, the optimal time for therapeutic intervention might have been missed.

Fig. 1.9

A 41-year-old woman who presented after a fall with left ankle pain and had significant limited range of motion. (a) Lateral radiograph of the ankle shows background calcaneal enthesopathy, with an avulsed fragment off the dorsal aspect of the calcaneus (arrow). This finding, together with concordant clinical examination findings, allowed the diagnosis of a high-grade Achilles tendon tear to be made. (b) Sagittal T1-W MR image of the left ankle shows disruption of the fibers of the Achilles tendon in keeping with a complete tear. The associated avulsion fracture of the dorsal aspect of the calcaneus is seen, and on the (c) sagittal fat-suppressed T2-W MR image, there is corresponding marrow edema (arrowheads)

Fig. 1.10

A 17-year-old man who had a history of trauma to the left thigh 2 weeks before presentation. (a) Frontal and (b) lateral radiographs of the left femur show faint curvilinear sheet-like calcifications (arrows) adjacent to the mid-shaft of the femur. In the context of recent trauma, this appearance is highly suggestive of myositis ossificans. Note the external artifacts due to the patient’s clothing

3.1 Intra- and Periarticular Structures

3.1.1 Articular Cartilage

The articular cartilage is essentially radiolucent and invisible on the radiograph. This makes its radiographic assessment particularly challenging. Radiographs are insensitive in the direct detection of chondral lesions, whether degenerative or traumatic in etiology. A purely chondral lesion caused by acute trauma, without involvement of the subchondral bone, cannot be appreciated on the radiograph (Fig. 1.11). The radiographic diagnosis of osteoarthritis is based on the observation of indirect features, such as the presence of marginal osteophytes, narrowing of joint space, as well as subchondral sclerosis and cyst formation. Radiographs have high specificity in the detection of advanced osteoarthritis, with the combination of indirect features allowing an easy diagnosis to be made. However, in early osteoarthritis, radiographs have low sensitivity and tend to underestimate the extent of cartilage degeneration (Blackburn et al. 1994). Despite the absence of radiographic signs of osteoarthritis, many symptomatic patients have been shown on arthroscopic evaluation to have significant degeneration of articular cartilage (Kijowski et al. 2006). In addition, the degree of joint space narrowing in patients with known osteoarthritis is a poor predictor of the actual state of the articular cartilage (Fife et al. 1991) (Fig. 1.12).

Fig. 1.11

A 42-year-old man who presented with left knee pain. (a) Frontal radiograph of the left knee shows minimal degenerative changes with preservation of the joint spaces. Coronal (b) PD-W and (c) fat-suppressed T2-W MR images of the left knee show a focal chondral defect (arrows) at the articular surface of the lateral femoral condyle. This finding cannot be appreciated radiographically since cartilage is essentially radiolucent

Fig. 1.12

A 44-year-old man who presented with chronic left knee pain. (a) Frontal and (b) lateral radiographs of the left knee show mild osteoarthritis with small marginal osteophytes and relative preservation of the joint spaces. A joint effusion is also noted. (c) Coronal and (d) axial fat-suppressed PD-W MR images of the left knee show partial- to full-thickness articular cartilage loss (arrows) involving the lateral tibiofemoral and patellofemoral joint compartments and confirm the presence of a joint effusion. As illustrated, the degree of joint space narrowing is not a sensitive method for predicting the actual state of the articular cartilage

Currently, radiography is still widely used in the imaging follow-up of patients with established osteoarthritis. Nevertheless, with advances in pharmacological and surgical therapies, a more precise imaging modality is required for articular cartilage evaluation. MRI is currently the gold standard in the imaging evaluation of articular cartilage. It has the ability to assess both the morphology and the biochemical integrity of the articular cartilage and will play an increasingly important role in the noninvasive evaluation of articular cartilage both before and after therapeutic intervention (Gold et al. 2009; Crema et al. 2011).

3.1.2 Synovium and Joint Fluid

The synovium is normally not visible on the radiograph. When synovitis occurs, whether inflammatory such as in the inflammatory arthritides or infective in the case of septic arthritis, the radiographic appearance is usually normal early in the course of the disease. However, at this early stage, some changes may already be appreciated on MRI or US imaging. Synovial hypertrophy, hypervascularity and enhancement, and possible adjacent soft tissue and bone marrow signal changes are features which are usually radiographically occult (Fig. 1.13).

Fig. 1.13

A 36-year-old man who presented with left elbow pain after a traumatic injury. Initial radiographic evaluation of the left elbow was negative for fracture or dislocation. (a) Lateral radiograph shows that the posterior fat pad is just visible (arrow), raising suspicion of a joint effusion. (b) Sagittal contrast-enhanced fat-suppressed T1-W MR image of the left elbow confirms the presence of a joint effusion and shows enhancement of the synovium (arrowheads), consistent with synovitis. No fracture was identified. In the absence of other clinical features of infection, transient inflammatory synovitis of traumatic etiology was the primary diagnosis. The patient made a full recovery with conservative management

Small joint effusions may not be appreciated on radiography. Even if seen radiographically, without information on the state of the synovium, periarticular soft tissues, and bone marrow, this finding may not be useful in narrowing the differential diagnoses (Fig. 1.14). On the other hand, MRI, with or without intravenous contrast administration, is able to show the state of the surrounding structures, allowing better assessment of the underlying pathology. US imaging is highly sensitive in the detection of joint effusions and is able to provide real-time imaging guidance for diagnostic joint aspiration.

Fig. 1.14

A 38-year-old man who presented with left groin pain after a period of intense physical activity. (a) Frontal radiograph of the pelvis is essentially normal. (b) Coronal turbo inversion recovery magnitude (TIRM) MR image of the pelvis shows a left hip joint effusion (arrow), which was radiographically occult. This represented a reactive effusion secondary to a left groin muscular strain (not shown), and the patient subsequently made a full recovery with conservative management

Radiographs have been used for more than a century in the imaging evaluation of the inflammatory arthritides, and they are able to demonstrate the osseous changes which indicate advanced disease. Currently, however, when modern therapy allows prevention or delay of irreversible joint destruction, imaging modalities with higher sensitivity for inflammatory changes (i.e., MRI and US imaging) are superior to radiographs in guiding treatment decisions (Szkudlarek et al. 2006; Weiner et al. 2008; Sankowski et al. 2013) (Fig. 1.15). Similarly, in the case of septic arthritis, irreversible joint destruction would have occurred by the time osseous changes are seen on the radiograph. Clinical judgment is of paramount importance in the management of patients with septic arthritis. Joint aspiration has to be performed for the diagnosis to be established, and treatment has to be instituted without delay, as rapid progression to permanent joint destruction may otherwise ensue.

Fig. 1.15

A 35-year-old man who presented with worsening back pain, on a background of chronic back and polyarticular pain. (a) Frontal radiograph of the pelvis shows no significant abnormality. (b) Coronal TIRM MR image of the sacrum shows periarticular marrow edema involving the inferior aspect of both sacroiliac joints (worse on the left), seen as conspicuous hyperintense fluid signal which has high contrast compared with the surrounding fat-suppressed normal marrow. This case highlights the high sensitivity of MRI to the early changes of sacroiliitis, when it is radiographically occult

3.1.3 Ligaments, Tendons, and Other Fibrocartilaginous Structures

Radiographic assessment of ligaments, tendons, and fibrocartilaginous structures relies upon the observation of indirect features which indicate possible underlying pathology involving these structures. These indirect features include joint malalignment, hemarthrosis, calcific deposits, and secondary osseous changes. The latter two are seen in relation to chronic degenerative processes. In acute trauma, injuries to these structures tend to be occult on the radiograph, especially when sprains, strains, or partial tears occur. Extensive ligamentous disruption with joint dislocation is usually visible on the radiograph. Complete tendon rupture with tendon retraction may also be visible radiographically. However, these injuries tend to be obvious on clinical examination.

Gross disruption of the supporting ligaments of a joint, with the presence of dislocation, represents one end of the spectrum of ligamentous injuries and manifests in radiographs as a disruption of the bony alignment. Less severe ligamentous injuries range from sprains to high-grade partial tears and even complete tears of individual ligaments. These injuries are largely not appreciated on radiographs, especially in the acute setting when joint instability might not be accurately assessed on clinical examination, and the bony alignment often remains normal on imaging (Figs. 1.16 and 1.17). Sometimes, the same traumatic mechanism causing ligamentous injury may result in bone abnormalities, which again are indirect features on the radiograph. These bone abnormalities can often be seen on the radiograph but are usually subtle and easily missed, if not suspected. In the example of an injury involving the anterior cruciate ligament of the knee, possible bony abnormalities include avulsion fractures at the femoral or tibial attachment sites, the Segond fracture, and an osteochondral impaction fracture of the lateral femoral condyle (Ng et al. 2011). Detection of any of these bony abnormalities without appreciating the underlying soft tissue injuries is a potential pitfall in the interpretation of the radiograph.

Fig. 1.16

A 25-year-old man who presented with right knee pain after sustaining an injury from a tackle during a football match. (a) Frontal radiograph of the right knee shows no fracture or dislocation. (b) Coronal fat-suppressed PD-W MR image of the right knee shows a grade 2 injury of the medial collateral ligament, evident as a partial disruption of the ligament with surrounding edema (arrowheads). This injury is usually not appreciable on radiography, but may be suggested if bony avulsion occurs at the attachment sites of the ligament

Fig. 1.17

A 35-year-old man who presented with right knee pain after a twisting injury sustained during a game of basketball. (a) Lateral radiograph of the right knee shows no significant bony injury or joint malalignment. A small suprapatellar joint effusion is noted. (b) Sagittal fat-suppressed PD-W MR image of the right knee shows a high-grade injury of the anterior cruciate ligament, evident as a disruption of the fibers at the femoral attachment (arrowheads). This injury is not usually discernable on radiography, although indirect osseous findings (such as Segond fracture) may suggest this injury

A complete tendon rupture with tendon retraction, involving a superficial large tendon such as the Achilles tendon, does not usually pose a diagnostic problem. However, in other locations, for example, the rotator cuff tendons in the shoulder, accurate assessment of tendon tears is impossible on radiographs. The rotator cuff is the archetype of a tendon which is highly susceptible to chronic degeneration. Significant chronic tendinosis and tendon tears of the rotator cuff are typically not visualized on radiographs (Fig. 1.18), though indirect findings such as osseous changes (signifying advanced disease), calcific tendinous deposits, and features of subacromial impingement may be seen. Nevertheless, other imaging modalities such as MRI and US imaging would be necessary for proper evaluation, as radiographic findings alone will not be sufficient to guide clinical management (Seibold et al. 1999). Other examples of fibrocartilaginous structures which are usually not directly visualized on radiography include intervertebral disks, menisci of the knee, glenoid labrum, and triangular fibrocartilage complex of the wrist (Figs. 1.19, 1.20, 1.21, and 1.22). Radiography plays a limited but usually complementary role in the evaluation of these structures.

Fig. 1.18

A 46-year-old man who presented with left shoulder pain and decreased range of motion after a motor vehicle accident. (a) Frontal radiograph of the left shoulder shows no significant abnormality. The humerus is in an internally rotated position. (b) Longitudinal US image of the supraspinatus shows a complete tear of the supraspinatus tendon, with a tendon gap at its attachment to the greater tuberosity (as indicated by the crosshairs). This finding was not apparent on the radiographs. (c) Longitudinal US image of another patient shows a normal fibrillar pattern of the distal supraspinatus tendon (shown for comparison). (d) Frontal radiograph of the left shoulder of a 70-year-old woman with advanced rotator cuff disease shows an obliterated acromiohumeral interval and secondary degenerative osseous changes, features which are radiographically discernible. At this late stage of disease, surgical intervention will not be useful

Fig. 1.19

A 25-year-old woman who presented with low back pain and bilateral lower limb numbness after lifting some heavy loads. (a) Lateral radiograph of the lumbar spine is essentially normal. (b) Sagittal fat-suppressed T2-W MR image of the lumbar spine shows posterior intervertebral disk extrusions at L4–L5 and L5-S1 levels (arrows), worse at the L4–L5 level where there is significant spinal canal stenosis with likely impingement of the cauda equina nerve roots. Note the reduced signal of the desiccated disks at the affected levels, as well as increased marrow signal at the endplates of L4–L5 level consistent with Modic type 1 degenerative signal changes. The other intervertebral disks have a normal appearance. Example of fibrocartilaginous structure (disk) not directly visualized on radiography

Fig. 1.20

A 37-year-old man who presented with left knee pain. (a) Frontal radiograph of the left knee shows no significant abnormality. (b) Coronal fat-suppressed PD-W MR image of the left knee shows a horizontal tear of the medial meniscus involving its inferior surface, evident as a linear area of hyperintensity (arrow). The lateral meniscus has a normal appearance. Example of fibrocartilaginous structure (meniscus) not directly visualized on radiography

Fig. 1.21

A 22-year-old man who had recurrent episodes of anterior shoulder dislocation. (a) Axillary radiographic view of the left shoulder shows normal joint alignment and no significant osseous abnormality. (b) Axial fat-suppressed T1-W MR arthrographic image of the left shoulder shows a Perthes lesion (a variant of the Bankart lesion), manifesting as detachment of the anteroinferior glenoid labrum which remains attached to an intact but lifted periosteum of the anterior glenoid (arrow). The posterior glenoid labrum shows a normal appearance. Example of fibrocartilaginous structure (glenoid labrum) not directly visualized on radiography

Fig. 1.22

A 28-year-old motorcyclist who had persistent ulnar-sided left wrist pain 4 months after being involved in a minor motor vehicle accident. (a) Frontal radiograph of his left wrist is essentially normal, with no fracture or dislocation seen. Negative ulnar variance is noted. (b) Coronal fat-suppressed T1-W MR arthrographic image of the left wrist shows a tear of the radial attachment site of the triangular fibrocartilage complex (TFCC), seen as linear high signal (arrow). Hyperintense contrast agent is seen extending into the distal radioulnar joint, a result of the TFCC tear. Example of fibrocartilaginous structure (TFCC) not directly visualized on radiography

3.2 Other Soft Tissues and Foreign Bodies

Various soft tissues, such as muscle and subcutaneous soft tissue, are usually included in the views obtained on radiography of the musculoskeletal system. Radiographs are largely limited in the assessment of muscle abnormalities, with rare exceptions such as myositis ossificans which shows typical radiographic appearances but may be diagnostically confusing on MRI early in its course (McCarthy and Sundaram 2005). Otherwise, muscle lesions are best assessed on MRI, which can demonstrate alterations in muscle signal intensity characteristics (Theodorou et al. 2012).

However, much information can still be gleaned from the radiographic appearance of the subcutaneous soft tissue. Diffuse processes such as edema and cellulitis may be appreciated by the presence of increased reticular markings and thickening of the overlying skin. However, this appearance is nonspecific, based on radiography alone. Likewise, for focal pathologies such as superficial hematoma, abscess, and a myriad of other soft tissue masses (including intramuscular masses), the radiographic appearance alone is usually nonspecific. Certain radiographic characteristics such as lesion density, the presence of calcification or ossification, and effect on adjacent osseous structures may shed some light on the nature of the soft tissue mass. Thus, although limited on its own, radiography plays a complementary role in the imaging evaluation of soft tissue masses (Gartner et al. 2009).

Radiographs are useful for the detection of suspected radiopaque foreign bodies, as well as soft tissue gas pockets. On the radiograph, gas pockets are visible on a background of soft tissue densities as their hyperlucency provides imaging contrast. Similarly, radiopaque foreign bodies can be seen, as the differences in densities provide good contrast. The higher the radiodensity of a foreign body, the greater its visibility. A limitation of radiography is in the detection of foreign bodies which are weakly radiopaque, especially if the material of the foreign body has a density close to that of the surrounding soft tissue (e.g., wood). These weakly radiopaque foreign bodies will not be appreciated on the radiograph. This potential pitfall should be recognized by the clinician requesting the radiograph, and if necessary, an alternative imaging modality such as US imaging should be considered (Aras et al. 2010) (Fig. 1.23).

Fig. 1.23

A 34-year-old construction worker who presented with pain in the right foot after stepping barefoot onto unknown material at his work site. (a) Lateral radiograph of the right foot does not show any radiopaque foreign body. (b) US image shows a few small foreign bodies within the plantar subcutaneous layer of the right foot, seen as linear echogenic structures (as indicated by the crosshairs). The patient underwent removal of the foreign bodies, which turned out to be wooden splinters. Wood is weakly radiopaque and may be invisible on radiography

4 Radiographically Occult Osseous Abnormalities

One of the strengths of the radiograph is its ability to demonstrate the osseous structures. This gives it relatively good specificity in the evaluation of osseous abnormalities. In most cases of discrete osseous lesions, the radiographic appearance allows categorization into aggressive and nonaggressive entities and so helps narrow the differential diagnoses. In many cases, the radiographic appearance is so characteristic as to allow a diagnosis to be made.

4.1 Destructive Osseous Lesions

Although radiographs display osseous anatomy well, before a destructive osseous lesion is even visible radiographically, there has to be loss of about 50% of the cortical bone mass (Osmond et al. 1975; Taoka et al. 2001). Radiography is thus significantly limited in the detection of lesions early in the course of osseous metastatic disease, since the pathology is predominantly confined to the medullary cavity of the bone at this early stage (Gold et al. 1990). In contrast to radiography, bone scintigraphy is sensitive enough to demonstrate metastatic involvement of cortical bone with a threshold of about 5–10% lesion-to-normal bone ratio (Algra et al. 1991) (Fig. 1.24).

Fig. 1.24

A 53-year-old woman who was recently diagnosed with adenocarcinoma of the right lung and presented with diffuse back pain. (a) Frontal and (b) lateral radiographs of the lumbar spine show spondylotic changes. Bone density is preserved and there is no evidence of osseous destruction. (c) Whole-body Tc-99 m MDP bone scintiscan shows extensive osseous metastases, evident as increased tracer uptake at multiple sites, especially in the axial skeleton including the lumbar spine. This case demonstrates the superior sensitivity of bone scintigraphy compared to radiographs in the detection of metastatic bone disease

Similarly, in osteomyelitis, before thresholds of about 50% of bone mineral content involvement and 1 cm of lesion size are reached, the lesion remains radiographically occult (Pineda et al. 2009). Thus, the radiographic features of osteomyelitis are typically delayed by about 10–14 days from the onset of infection. In the early stages of osteomyelitis, the radiograph can be normal in appearance and so is significantly limited as a diagnostic modality. However, radiographs can still be useful for demonstrating associated findings such as foreign bodies and soft tissue gas in this setting. MRI is exquisitely sensitive to bone marrow changes, making it the imaging modality of choice in the evaluation of vertebral metastases as well as osteomyelitis, both of which mainly involve the bone marrow (Algra et al. 1991; Pineda et al. 2009) (Figs. 1.25 and 1.26).

Fig. 1.25

A 62-year-old man with type 2 diabetes mellitus who presented with an infected ulcer at the left hindfoot. (a) Lateral radiograph of the left ankle shows a radiolucent area posterior to the calcaneum (arrow), corresponding to the known ulcer. No obvious radiographic sign of osteomyelitis (e.g., periosteal reaction and bone destruction) is discerned. (b) Sagittal TIRM MR image of the left foot shows subcutaneous inflammatory signal alteration deep to the ulcer (arrowheads), extending down to the calcaneum which shows abnormal marrow edema (asterisk). (c) Sagittal T1-W MR image shows the characteristic T1-hypointense marrow signal of osteomyelitis involving the calcaneum (asterisk)

Fig. 1.26

A 51-year-old man who presented with back pain of insidious onset. (a) Frontal and (b) lateral radiographs of the thoracic spine appear essentially normal. There is no radiographic evidence of osseous metastasis. (c) Sagittal T1-W MR image of the thoracic spine shows abnormal hypointense marrow signal involving multiple vertebral levels. The patient had extensive osseous metastases secondary to a primary malignancy in the right lung apex, which was incidentally imaged on the frontal radiograph (a) and seen as right lung apical opacities

4.2 Trauma-Related Osseous Injuries

4.2.1 Undisplaced Fractures

An important limitation of radiographs is in the evaluation of acute undisplaced fractures, especially hairline ones. For an acute fracture to be visualized on the radiograph, at least a small amount of displacement or separation is usually necessary for the fracture to manifest as a radiolucent line, sclerotic line, or cortical step. Hence, even with adequate views and proper technique, an acute undisplaced fracture can be radiographically occult. It is important to be aware of this potential pitfall when interpreting radiographs of the acute trauma patient, especially if there are associated soft tissue findings in a seemingly negative radiograph. For example, elevated fat pads in the elbow joint indicate the presence of a hemarthrosis, which could be secondary to an occult undisplaced fracture. If the clinical suspicion for a fracture is high despite a negative initial radiograph, follow-up radiographs can be obtained 10–14 days later. If present, these fractures typically become increasingly visible with time as bone resorption and callus formation occur at the fracture site (Fig. 1.27).

Fig. 1.27

A 6-year-old boy who presented with pain in the left elbow after a fall onto his outstretched left hand. (a) Lateral radiograph of the left elbow shows no displaced fracture. The joint alignment is maintained. There is however a joint effusion, as indicated by elevated anterior and posterior fat pads (arrowheads). This finding should raise the suspicion of an occult undisplaced fracture in the setting of trauma. (b) Follow-up lateral radiograph of the left elbow was performed 10 days after the initial presentation. Despite the presence of an overlying cast, periosteal reaction can be seen along the posterior aspect of the distal shaft of the humerus (arrow), indicating a healing undisplaced fracture

In certain clinical scenarios, confirmation of the presence of a fracture may need to be done rapidly, as an unnecessary delay in the diagnosis would result in significant morbidity. In such cases, follow-up radiographs should not be advocated and advanced imaging evaluation should instead be performed. For example, in the elderly adult with hip pain after a fall, where there is inability to weight-bear and no fracture is seen on radiographs, further advanced imaging such as MRI and CT should be arranged early to establish the presence of an undisplaced femoral neck fracture, which usually requires to be treated surgically (Oka and Monu 2004; Gill et al. 2013; Ward et al. 2013) (Fig. 1.28).

Fig. 1.28

An 83-year-old woman who presented with left hip pain after a fall. (a) Frontal radiograph of the pelvis and (b) lateral radiograph of the left hip show no appreciable fracture or dislocation. There is diffuse reduction in bone density, which limits radiographic sensitivity for minimally displaced fractures. (c) Axial T1-W MR image of the pelvis shows an undisplaced fracture of the neck of the left femur, seen as a hypointense fracture line (arrow). This fracture was radiographically occult. (d) Axial fat-suppressed T2-W MR image of the pelvis at the corresponding level shows associated marrow edema around the fracture site

4.2.2 Stress Injuries

Stress injuries range in a continuum from stress reactions to established stress fractures. They develop as a result of chronic repetitive microtrauma causing fatigue of normal bone and essentially comprise microtrabecular fractures which are not visible on the radiograph. If the inciting activity causing repetitive stress is not stopped to allow bone to heal, these microtrabecular fractures accumulate and eventually result in a full cortical fracture. Formation of periosteal new bone is the earliest radiographic feature of stress fractures, and its appearance may be delayed up to 3 months from the initial injurious stimulus (Jarraya et al. 2013) (Fig. 1.29).

Fig. 1.29

A 24-year-old avid runner who presented with right foot pain. (a) Frontal and (b) oblique radiographs of the right foot show no significant abnormality. Incidental note is made of a type 2 os navicularis. (c) Sagittal T1-W MR image of the right foot shows curvilinear hypointense trabecular fracture lines in the base of the fourth metatarsal (arrow). (d) Sagittal fat-suppressed T2-W MR image of the right foot shows associated marrow edema (asterisk). With the presence of a concordant history, the findings were consistent with a stress injury. (e) Frontal radiograph of the right foot in another patient at a later stage in the natural progression of a stress injury shows typical periosteal reaction adjacent to the neck of the second metatarsal (arrowhead), indicating a healing stress fracture. This is the earliest radiographic feature of stress injury

The earliest stage of stress injuries, termed stress reaction, is radiographically occult but may manifest on MRI as bone marrow edema without a fracture line. Due to nonspecificity of isolated bone marrow edema, CT may also be useful in distinguishing stress reaction from other entities such as osteoid osteoma (Liong and Whitehouse 2012). Since only timely management can interrupt the cycle of repetitive stress, early detection of stress injuries is crucial. An understanding of the limitation of radiography in this respect and the use of more appropriate imaging modalities is vital in establishing the diagnosis of stress injury in the early stages.

4.3 Osteoporosis

Osteoporosis is defined as a reduction in bone mass, with a bone density of less than 2.5 standard deviations below that of a healthy young adult (World Health Organization 2003). It has many different causes, but its appearance on radiography is the same regardless of etiology. The diagnosis is most confidently made on a quantitative technique such as dual-energy X-ray absorptiometry (DXA). However, most cases of osteoporosis are still diagnosed on radiography (Guglielmi et al. 2011). Radiographs are inherently limited in the detection of reduced bone mass, which is only appreciable when about 30% of bone loss has occurred (Harris and Heaney 1969). Radiographic technique may also be a confounding factor, for example, causing bones to appear more radiolucent than usual and giving a false impression of osteoporosis. This is a potential pitfall when radiographs are used in the diagnosis of osteoporosis. A combination of features such as cortical thinning, trabecular changes, and insufficiency fractures is usually needed to provide a higher degree of confidence in the diagnosis of osteoporosis on radiography (Fan and Peh 2016). Another radiographic pitfall in relation to osteoporosis is the limitation in the detection of destructive lesions and fractures. On a background of reduced bone mass, these conditions can be difficult or impossible to appreciate. This is especially so early in the course of destructive processes (e.g., osteomyelitis or osseous metastasis) or when fractures are undisplaced (e.g., hip fracture in the elderly adult) (Figs. 1.28 and 1.30).

Fig. 1.30

A 74-year-old woman who presented with low back pain and tenderness in the upper gluteal region after a fall. (a) Frontal and (b) lateral radiographs of the lumbar spine show diffuse reduction in bone density consistent with osteoporosis. Known chronic osteoporotic compression fractures are seen at the thoracolumbar junction levels. Background spinal degenerative changes are evident, especially at the lower lumbar spine. (c) Coronal T1-W and (d) coronal fat-suppressed T2-W MR images of the sacrum show sacral insufficiency fractures, evident as a typical “H” pattern of T1-hypointense and T2-hyperintense marrow edema involving the bilateral sacral ala and the body of the second sacral segment. This finding is radiographically occult

Conclusion

Radiography as an imaging modality has inherent limitations in the demonstration of lesions involving soft tissues, as well as early in the course of abnormalities involving bone. Pitfalls are also encountered in radiographic acquisition, and if the clinician or radiologist is not cognizant of them, suboptimal diagnosis and patient management can ensue. Nevertheless, the humble radiograph remains a mainstay of diagnostic imaging in the modern day, frequently being the first-line imaging modality or assuming a complementary role to other more advanced imaging modalities.