SEARCH

SEARCH BY CITATION

Keywords:

  • Cone beam;
  • radiography;
  • radiology;
  • computed tomography

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

Radiologic interpretation is a complex process which involves the application of an appropriate algorithm in the study of radiologic images and the ability to understand the meaning and to weight the various findings, ultimately contributing to diagnosis. Prerequisites include the knowledge of orofacial radiologic anatomy and the various pathoses which may arise or manifest in this region of the body. An understanding of the strengths and limitations of the modality employed is also essential. The process of interrogating radiologic images for abnormalities varies, depending on the modality. This paper outlines the basic steps involved in the radiologic examination of abnormalities which affect the jaws, primarily in relation to plain 2-D imaging.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

Radiologic interpretation involves the detailed study of radiologic images, ultimately contributing to diagnosis. It requires the application of an algorithm which demands a certain knowledge base and skill set. This is, in part, dependent on the imaging modality employed and the region examined. Several essential prerequisites allow accurate identification of the normal structures and abnormalities. The ability to understand the features of an abnormality and to weight these findings is also crucial.

A thorough discussion of all these points is not within the scope of this paper and acceptable interpretive standards can only be achieved with clinically based programmes. This paper will focus on identifying some of the key prerequisites and introducing the key steps in the radiologic interpretation of plain 2-D radiographic images of the dental and orofacial structures, such as the intraoral and panoramic radiographic images.

Prerequsites

Radiologic anatomy

The detailed knowledge of anatomy is an obvious prerequisite for all clinicians. A similar understanding of the appearance of anatomic structures in radiologic images is also crucial. This requires a thorough appreciation of anatomy three-dimensionally and how these structures appear radiologically, depending on the imaging modality employed. Knowledge of normal anatomic variants and their radiological appearances is also vital.

In plain 2-D radiographs, awareness of the angle of projection and associated geometry is crucial. For example, the temporomandibular joint is poorly examined with the panoramic radiograph,1–3 related to the obliquity of projection and the limitations of tomography.

Obviously, larger field of view 2-D images will include more structures, requiring a broader knowledge base. It is also important to understand the differences in the appearance of anatomic structures in plain 2-D projections (e.g. intraoral periapical view) compared to a panoramic radiograph, which is essentially a curvilinear tomogram.

The radiologic anatomy as depicted in plain 2-D images differs from that seen in volumetric radiologic studies, such as multislice CT and cone beam imaging.

It is essential that the clinician responsible for the interpretation is able to identify all the anatomic structures which are depicted in the images. The ability to identify the presence of an abnormality is severely compromised if the clinician is not completely familiar with the appearances of all anatomic structures and normal morphologic variants. It must be emphasized that not all abnormalities present as obvious opacities or lucencies.

Pathology

It is obvious that the clinician carrying out the interpretation must be aware of the pathoses which may arise or manifest in the region included in the radiologic study. Many imaging techniques in dentistry capture a large proportion of the orofacial structures. Examples include panoramic and cephalometric projections. To varying extents, cone beam scans include the paranasal sinuses, pharyngeal air spaces, skull base, cervical spine and upper neck. It is imperative that clinicians responsible for the interpretation are familiar with the diseases which are potentially associated with these structures.

As an example, in order to interpret an image or scan of a temporomandibular joint, the clinician must be familiar with the possible diseases which can affect this joint, ranging from internal derangement and degenerative disease to the erosive arthropathies, synovial chondromatosis, chondrocalcinosis, vascular lesions and various tumours. The radiologic presentations of these conditions differ and can often be differentiated radiologically. The application of the correct modality is also critical and it is worth pointing out that cone beam imaging is often not necessarily the optimal technique for these joints. Lack of knowledge of the potential diseases which can affect this joint, especially when the optimal modality is not employed, can lead to suboptimal outcomes (Fig. 1).

image

Figure 1.  Opacities associated with the right TMJ typical of chondrocalcinosis.

Download figure to PowerPoint

Imaging modality

There are numerous imaging techniques which can be applied in the radiologic examination of the orofacial structures, including intraoral and other plain 2-D views, the panoramic tomograph, multislice CT, cone beam imaging, MRI, ultrasound and nuclear medicine. While modern dentistry can no longer rely on intraoral and panoramic radiography alone, relatively new modalities such as cone beam imaging must be applied judiciously. Importantly, there are other advanced techniques which may be more appropriate and should also be considered. The radiations dose levels delivered vary substantially between cone beam machines, some of which can be higher than multislice CT (when appropriate protocols are employed).4

The significant differences of the strengths and limitations of all these modalities are not within the scope of this paper. However, of note are the substantial limitations of the commonly used panoramic radiograph. The orofacial structures depicted in this view are not necessarily sufficiently well demonstrated and can lead to erroneous interpretations. Cone beam imaging has been a relatively new technique which is increasingly employed. Persons engaged in the interpretation of these scans must be thoroughly familiar with the limitations of this modality, including beam hardening, noise, low signal, metal artefact and poor soft tissue contrast resolution and the effect of motion artefact.4 The effects of the various protocols using the same equipment must also be understood.

Viewing conditions

Optimal viewing conditions are necessary to allow identification of all the key features in an image, including normal anatomy. The subtle absence of a normal structure can be a key finding leading to identification of a significant abnormality. Ambient light must be kept to a minimum and extraneous light from a viewing box should be obscured.

The ability to optimally view every aspect of an image is crucial. Digital images viewed on a computer monitor can be easily manipulated, including magnification and windowing. The quality of the monitor is crucial and is not infrequently the weakest link in the dental practice. For traditional analogue images, optical magnification and using a brighter light source for darker regions can be critical.

The interpretation of plain 2-D radiographic digital images printed on paper can be problematic. The quality of the printer and paper are crucial. Even high quality photographic paper generated from a high quality printer does not demonstrate the same optical range as film or high quality monitors. The interpretation of 2-D digital images is optimally performed on high quality monitors or on high quality film. When images are transferred electronically, the level of compression of images must be minimized or avoided so that crucial data is not lost. Volumetric data such as multislice CT is ideally interpreted by interrogating all the data with a computer.

Ultimately, a lesion must be entirely included in a field of view or scan. If this is not the case initially (e.g. in a periapical or panoramic view) preliminary interpretation should still be carried out, which can be useful in deciding which modality is optimal in further evaluation.

Optimal imaging is assumed and the technical aspects of imaging are well discussed in many texts.

Radiologic intepretation

Radiologic interpretation is essentially based upon the understanding of disease processes and the behaviour of diseases in a specific anatomic region. By carefully applying a series of steps, the key features of a lesion can be identified. Combined with knowledge of the specific radiologic characteristics of various lesions, this can contribute substantially to diagnosis.

It is desirable that the diagnostic imaging and interpretation is completed prior to biopsy. Appropriate interpretation can be useful in identifying the optimal and/or safe site(s) for biopsy. It is also critical that some lesions are excluded prior to biopsy or surgical intervention. Of note are the vascular malformations. In addition, surgical procedures and biopsies can substantially alter the radiological appearances of a lesion, usually by introducing inflammatory changes, potentially compromising diagnosis and management.

The ability to perform morphologic analyses and plan surgical procedures with a specific imaging technique is different to the skill set required to evaluate the same data set for the presence of disease and the interpretation of the radiologic features of a lesion.

Recognizing the presence of an abnormality

The practitioner is responsible for the entire volume of information of a scan and all structures in the field of view, not just for the primary focus of the study.

The process of interrogating radiologic images for abnormalities varies, depending on the modality. While the actual sequence in which the evaluation of an image is carried out may vary, the entire image must be methodically and thoroughly studied. Every normal anatomical structure that should be within the field of view must be specifically identified and evaluated, including its normal boundaries and internal appearances. This is critical since not all lesions are obvious. For example the absence of a cortical boundary of a structure in a panoramic radiograph, cone beam or multislice CT can reflect the presence of significant disease (Fig. 2).

image

Figure 2.  Cone beam corrected sagittal image demonstrating the cortical erosions typical of an infiltrative lesion such as lymphoma.

Download figure to PowerPoint

The algorithm and skill set required to thoroughly evaluate volumetric data, including those captured with multislice CT and cone beam techniques, differ to that which is employed for intraoral and panoramic images. Volumetric data should be evaluated in multiple appropriate planes and windows, depending on the structures involved and potential associated diseases. Evaluation of volumetric data in one plane can result in misinterpretation and non-identification of the presence of disease.

In addition, there are commonly known specific radiologic features which suggest the presence of disease. For example, a wide stylo-mandibular notch is a feature which suggests presence of a mass related to the deep lobe of the parotid salivary gland.

Radiologic evaluation of a lesion

The following describes a series of steps which assists in the identification of the important radiologic features, highlighting the behaviour and nature of a lesion. These features identified are also important from a surgical standpoint.

1. Location

Before focusing on the precise location of a lesion, the entire field of view or scan should be evaluated for other possible related lesions. That is, the disease may be multifocal or generalized. If there is more than one lesion, it is important to note if the lesions are monostotic or polyostotic and unilateral or bilateral. For example, Gorlin-Goltz syndrome (nevoid basal cell carcinoma syndrome) needs to be considered if multiple cystic lesions are identified.

A key point of this step is to attempt to identify the point at which the lesion originated. This is often the anatomic centre of a lesion (e.g. the centre of a spherical lesion). However, the origin of a lesion is not always at the anatomic centre. A lesion arising from the maxillary alveolar process expanding into maxillary sinus is an example. Being filled with air, the maxillary sinus allows for much easier expansion than the alveolar bone. Hence, most of the volume of this lesion is within the sinus. In these cases, the anatomic centre is not the origin. This highlights the importance of understanding the involved anatomy.

The location and extent of the lesion can provide useful information about the likely tissues involved. For example, a lesion in the posterior body of the mandible which arises from below the mandibular canal is unlikely odontogenic. Detailed information regarding the location and extent of a lesion is also critical in relation to surgical planning and biopsy.

2. Shape and contour

The shape of the lesion can provide useful information on the lesion. True cysts such as the dentigerous cyst are generally spherical or ovoid. In contrast, an odontogenic cyst often demonstrates a scalloped peripheral morphology. An osteoma usually presents as a smooth convex bony prominence while an osteochondroma tends to present with a more irregular surface. A bone island typically demonstrates an irregular outline (Fig. 3).

image

Figure 3.  Periapical radiograph depicting a bone island.

Download figure to PowerPoint

3. Border

The first part of this step is to identify if the border of a lesion is well defined or poorly defined. Generally, a well-defined border suggests that the lesion is likely slower growing. It should be noted that similarly well-defined lesions generally appear less defined in the maxilla than the mandible, especially in plain 2-D views. This is related to the trabecular architecture and also the thickness of the bone.

If a lesion is well defined, this border has to be further examined and subcategorized. Most well-defined borders fall into one of the following descriptions:

  • 1
     A sharp demarcation between normal and abnormal with no other features. This is often referred to as ‘punched out’. Multiple myeloma is a classical example.
  • 2
     A corticated border. This describes a sharp opaque usually curved line (Fig. 4).
  • 3
     A sclerotic border. This refers to an opaque border which is thicker and less uniform than a corticated border. Most chronic inflammatory bony lesions demonstrate sclerotic margins, which reflect the reaction of the surrounding trabecular bone to the inflammatory lesion (Fig. 5). However, other lesions including cement-osseous dysplasia and some malignant lesions can also demonstrate sclerotic margins.
  • 4
     A surrounding lucent margin. This usually refers to opaque and mixed-density lesions where the lucent margin reflects presence of soft tissue surrounding the lesion (Fig. 6).
image

Figure 4.  A cropped panoramic radiograph depicting the corticated border of a radicular cyst and associated inferior deflection of the mandibular canal.

Download figure to PowerPoint

image

Figure 5.  A periapical radiographic image depicting a periapical inflammatory lesion.

Download figure to PowerPoint

image

Figure 6.  A cropped panoramic view depicting an odontome.

Download figure to PowerPoint

If a lesion has an ill-defined border, further analysis is necessary to decide which of the following two best describes it:

  • 1
     A gradual change from abnormal to normal. Inflammatory lesions, unless extremely chronic, demonstrate these borders (Fig. 5).
  • 2
     An aggressive margin. Malignant lesions classically demonstrate these borders. Aggressive and infiltrative margins include the appearance of extension into the adjacent bone and enlargement of adjacent marrow spaces. Irregular widening of the periodontal ligament space with focal destruction of the lamina dura is also another example of the leading edge of an infiltrative lesion spreading around tooth roots.
4. Internal appearances

The internal appearance of a lesion provides important clues as to the likely nature of the lesion. Initially, it is useful to identify if the lesion is completely opaque or complete lucent.

For completely opaque lesions, the density and degree of homogeneity or heterogeneity should be identified together with any consistent pattern. A classical fibrous dysplasia is internally homogenous with ground glass appearances (Fig. 7). A bone island is usually internally homogenous and isodense with cortical bone. Osteomyelitis can demonstrate heterogenous internal appearances with focal areas of rarefaction and sclerotic regions.

image

Figure 7.  Fibrous dysplasia of the right posterior maxillary alveolar process, better visualized in the axial multislice CT image than in the cropped panoramic radiograph.

Download figure to PowerPoint

A complete lucent lesion seen in plain 2-D imaging could reflect presence of air/gas, fluid or soft tissue. Typically, air/gas and fat will appear more lucent than fluid and soft tissue in plain 2-D radiography. It is important that this is understood in relative terms. For example, in a panoramic radiograph, a fluid-filled cyst within the maxillary sinus will appear relatively opaque while a fluid-filled cyst with the mandible will appear lucent. Cone beam imaging can clearly differentiate between air and soft tissue but cannot differentiate between fluid and soft tissue and different types of soft tissue. Multislice CT has far superior soft tissue contrast resolution and can demonstrate density differences between different types of soft tissues. Multislice CT is often able to differentiate between soft tissue and fluid, although the effects of beam hardening need to be considered when evaluating the internal attenuation characteristics of intrabony lucent lesions. It is important to note that soft tissues are best evaluated with MRI.

There are also lesions which demonstrate both lucent and opaque internal appearances. In these cases, it is important to try to identify the nature of the opacities e.g. bony, odontoid or dystrophic calcification. In addition, the pattern of the opacities is also important as many lesions often demonstrate a certain pattern. For example, an ameloblastoma classically demonstrates coarse curvilinear internal septae while giant cell lesions often reveal much finer septae.

5. Adjacent anatomic structures

The way in which the normal anatomic structures influence the pattern growth and spread of a lesion and the effects that a lesion has on these adjacent anatomic structures as the lesion enlarges are both important features which require detailed analysis.

For example, a radicular cyst would displace the mandibular canals while an inflammatory lesion does not usually display this characteristic (Fig. 4). Tooth displacement is a characteristic of a lesion with mass effect, usually benign in nature. The way in which the periapical bone is effaced provides important clues as to the likely nature of the lesion (Fig. 8). Some lesions, such as an ameloblastoma and a giant cell lesion demonstrate a propensity to resorp tooth roots. The resorption of roots require time, which is not a typical feature with malignant lesions. Malignant lesions tend to destroy the bone surrounding the tooth. While destruction of cortical boundaries can occur with a variety of lesions, including inflammatory lesions, it requires special attention as malignant lesions often demonstrate this characteristic (Fig. 2). Periosteal new bone formation is often seen in more extensive inflammatory lesions affecting the jaws, such as osteomyelitis (Fig. 9).

image

Figure 8.  A periapical view depicting the superior aspect of a keratocystic odontogenic tumour. Note the right angle appearance of the superior border in relation to the tooth roots, which is not usually seen with radicular cysts.

Download figure to PowerPoint

image

Figure 9.  An axial multislice CT image depicting periosteal response related to an inflammatory lesion.

Download figure to PowerPoint

Interpretation of the findings

Many lesions can demonstrate similar features. In addition, many lesions do not always present classically and demonstrate only a few or even just one of its typical features. Therefore, the next critical step is for the observer to weight the various features identified. This requires knowledge of the classical radiologic features of the possible pathoses and experience in the application of this information.

For example, fibrous dysplasia does not always present with classical ground glass internal appearance and can appear quite heterogenous internally. However, in these cases, the nature of the expansion often remains typical which assists the observer in differentiating it from other lesions which can otherwise appear similar, such as the cement-ossifying fibroma.

A lucent lesion located at the apex of a tooth root is often inflammatory in nature. However, malignant lesions at root apices are also lucent. Furthermore, some can indeed demonstrate adjacent sclerosis which can appear similar to the reactive sclerosis often seen in inflammatory lesions. In these cases, the invasive nature of margins and the irregular way in which the adjacent lamina dura and periodontal ligament spaces are destroyed are critical features.

Classification

One of the final steps in radiologic interpretation involves the classification of the lesion. While it is tempting to rush to provide a specific diagnosis, it is important that the thought process focuses initially on classifying the lesion into a broad category, e.g. inflammatory, fibro-osseous, cyst, benign or malignant tumour, vascular etc. This step allows the observer to finally re-evaluate all key features identified and reduces the likelihood that a key feature is not considered, e.g. the serpiginous appearance of the mandibular canal typical of a vascular malformation. It is often useful to also consider the patient’s age and ethnicity (if known) at this stage. Once the lesion has been classified into the broader categories, the observer will then be able to decide upon the most likely nature of the lesion.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References

It is important to reiterate the importance of employing a methodical approach to radiologic interpretation. Sole reliance upon a classical appearance of a lesion and use of the ‘Aunt Minnie’ style of interpretation (looks like the lesion the observer last encountered or saw in a book) is insufficient and can lead to erroneous interpretation and misdiagnosis.

In dentistry, the observer is often also the clinician. In these instances, the observer can easily develop a preconception as to the likely diagnosis and thorough evaluation of the radiologic study is not performed. An example is when there is a clinically suspicious periapical inflammatory lesion where radiologic interpretation ends as soon as a lucent appearance is noted periapically. A thorough radiologic evaluation to identify the features of an inflammatory lesion is then not performed, which can contribute to misdiagnosis.

Like other facets of dentistry and diagnosis, knowledge combined with appropriate training and guided experience is critical in the development of competent radiological interpretive skills. The author hopes that this paper has been effective in introducing an algorithm for radiologic interpretation and raising associated pertinent issues. Equally important, it is also hoped that this paper inspires those involved in radiologic interpretation to continue development of their individual skills and to be cautious where the imaging is beyond their interpretive skill set.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Conclusions
  5. References