Oral and maxillofacial radiology is the dental specialty devoted to producing and interpreting images of the teeth, jaws, skull, and surrounding structures. It covers everything from the familiar bite-wing X-ray your dentist snaps during a checkup to sophisticated three-dimensional scans used to plan jaw surgery or place dental implants. Although every dentist takes and reads basic radiographs, oral and maxillofacial radiologists are the specialists who handle the difficult calls: distinguishing a benign cyst from an aggressive tumor on a panoramic image, reading a cone beam CT before complex surgery, or catching an incidental finding that has nothing to do with the patient’s teeth but everything to do with their health.
The Core Imaging Technologies
The toolkit has grown dramatically over the past half-century. For most of that history, dental imaging meant small film packets pressed against the inside of the mouth, capturing a handful of teeth at a time. Those intraoral radiographs are still the workhorses of everyday dentistry, though the film packets have largely been replaced by digital sensors and phosphor plates that cut radiation exposure and let clinicians adjust contrast on screen.
Panoramic radiography, which produces a single wide image of both jaws in one shot, became practical in the 1960s and remains one of the most commonly ordered dental images. It works by rotating an X-ray source and detector around the patient’s head, capturing a curved slice of anatomy called the focal trough. Structures sitting inside that trough appear sharp and geometrically accurate, while anything outside it blurs and distorts.
The geometry matters because even objects within the trough get magnified unevenly. Vertical magnification tends to be more consistent than horizontal, and only certain positions in the focal trough produce zero distortion.
The technology that reshaped the specialty most profoundly is cone beam computed tomography, or CBCT. Unlike a medical CT scanner, which spins a narrow fan-shaped beam around the patient hundreds of times, a CBCT unit makes a single rotation with a wide, cone-shaped beam and a flat-panel detector. The result is a three-dimensional volume of the jaws and surrounding anatomy at high spatial resolution, with a radiation dose roughly 30 percent lower than a comparable medical CT scan.
Where Panoramic Images Excel and Where They Fall Short
A panoramic radiograph is fast, cheap, and gives the clinician a broad overview. It is usually the first image ordered when a patient complains of jaw pain, when wisdom teeth need evaluating, or when a dentist suspects a cyst or tumor. Panoramic views can reveal the overall shape and borders of jaw lesions and help distinguish between entities like ameloblastomas and odontogenic keratocysts. Ameloblastomas tend to cause obvious jaw swelling, multilocular (soap-bubble) patterns, and resorption of nearby tooth roots, while keratocysts more often grow along the length of the jawbone with less internal division and a lower rate of root resorption.
But panoramic images are two-dimensional projections of three-dimensional anatomy. They cannot show whether a lesion has punched through the outer or inner wall of the jawbone, and they compress overlapping structures into a single flat image. For anything that requires depth information, whether it is the buccolingual extent of a tumor, the position of an impacted tooth relative to a nerve canal, or the thickness of bone available for an implant, clinicians turn to CBCT or medical CT.
Cone Beam CT in Practice
CBCT has become the default three-dimensional imaging tool in dentistry. It excels at visualizing high-contrast structures like bone and teeth, which is exactly what most dental and maxillofacial questions demand. CBCT is designed to provide relatively low-dose, high-spatial-resolution visualization of these high-contrast structures in the head and neck.
Implant planning is one of its signature uses. Before placing an implant, a surgeon needs to know the height and width of bone at the planned site, the density of that bone, and the exact location of structures like the inferior alveolar nerve or the maxillary sinus floor. CBCT-derived density values show a strong positive correlation with actual bone volume measured by micro-CT, making preoperative density estimation from CBCT a reliable way to objectively gauge bone quality before drilling.
That density information feeds directly into surgical guide fabrication. Clinicians can merge a CBCT volume with a digital surface scan of the teeth to virtually plan implant positions, then 3D-print a guide that snaps onto the teeth and directs the drill. Studies show these digitally planned and printed guides achieve higher placement accuracy than older transfer methods that relied on physical jigs.
CBCT is also used to hunt for vertical root fractures in teeth that have already had root canal treatment, a notoriously difficult diagnosis. A systematic review found CBCT accuracy averaged around 78 percent for teeth without posts and about 74 percent for teeth with posts inside the canal.
Diagnosing Jaw and Facial Fractures
When patients arrive with mid-face trauma, conventional X-rays can miss fractures or underestimate their extent. Three-dimensional CT reconstruction gives surgeons a clear picture of where major fracture lines run and how the fragments have shifted. In a study of 100 mid-face fracture cases, 3D CT significantly influenced the final diagnosis and surgical plan in about 28 percent of cases, particularly in severe injuries where multiple bones were involved.
Research comparing standard CT slices, multiplanar reformats, and 3D reconstructions found that 3D and axial CT performed similarly for detecting fractures, and both were markedly better than multiplanar reformats alone. Where 3D reconstructions really shine is in localizing complex fractures that span multiple planes, giving the surgeon a rotatable model of the damage before making an incision.
Temporomandibular Joint Imaging
The temporomandibular joint, or TMJ, sits just in front of the ear and lets you open and close your mouth. When it malfunctions, patients can experience clicking, locking, pain, or limited jaw movement. Imaging plays a central role in sorting out what is going on structurally.
MRI is considered the gold standard for evaluating the soft-tissue components of the TMJ, particularly the articular disc (the small cartilage cushion between the skull base and the jaw condyle). MRI can show whether the disc has slipped forward, sideways, or in a combination pattern, and whether it pops back into place when the mouth opens (displacement with reduction) or stays stuck (displacement without reduction).
One MRI-based study of displaced discs found anterior displacement with reduction in 39 joints, anterior displacement without reduction in 31, compound dislocations in 13, and sideways displacement in 5. Roughly a quarter of affected joints also showed abnormal shape changes in both the disc and the condyle.
An important caveat is that disc displacement frequently shows up on MRI in people who have no symptoms at all. That means a displaced disc alone does not seal the diagnosis of TMJ dysfunction; clinicians have to weigh the imaging against the patient’s actual complaints and clinical exam findings.
Salivary Gland Imaging
Salivary stones and chronic gland inflammation are common reasons patients end up needing maxillofacial imaging. The traditional approach is sialography, where contrast dye is injected into the gland’s duct and X-rays are taken to map the duct system. It remains one of the best ways to visualize the fine branching anatomy of salivary ducts. Digital subtraction sialography and ultrasound are considered the methods of choice for imaging salivary stones specifically.
MRI is increasingly used as a less invasive alternative, and MR sialography, which maps the ducts without injecting contrast, may eventually replace conventional sialography for many indications. Conventional sialography is now mostly reserved for evaluating chronic inflammatory conditions unrelated to stones.
Incidental Findings That Have Nothing to Do With Teeth
One of the more surprising aspects of modern maxillofacial radiology is how often a scan ordered for a dental reason reveals something unrelated and medically significant. CBCT fields of view routinely capture anatomy beyond the teeth and jaws, including the cervical spine, the airway, the base of the skull, and portions of the carotid arteries.
In a review of 400 CBCT scans, 653 incidental findings were identified in about three-quarters of the scans. The most common were soft-tissue calcifications (around 30 percent of findings), followed by intracranial calcifications (27 percent), cervical vertebral changes (20 percent), and airway findings (11 percent). About 31 percent of all incidental findings warranted referral to a physician, and another 18 percent needed monitoring.
A larger analysis of 1,260 CBCT reports found a similar pattern. The most common findings outside the area of dental interest involved the cervical vertebrae (18 percent), sinuses (15 percent), and temporomandibular joints (8 percent). Among findings that triggered referrals, carotid atheroma (calcified plaque in the carotid artery wall) topped the list at about 2.7 percent of all scans.
Carotid calcifications are particularly consequential because they may signal elevated stroke risk. A maxillofacial radiologist spotting these on what was supposed to be an implant-planning scan could be the first clinician to flag a cardiovascular issue the patient did not know they had. This underscores why interpretation of CBCT scans by trained specialists matters: a general dentist focused on the teeth might not think to examine the soft tissues at the edge of the image.
Radiation Safety and Dose Reduction
Any X-ray-based imaging involves ionizing radiation, and the mouth sits close to radiosensitive structures like the thyroid gland, the salivary glands, and the lens of the eye. Keeping doses as low as reasonably achievable (the ALARA principle) is a core concern of the specialty.
CBCT delivers meaningfully less radiation than medical CT for comparable dental imaging tasks. One comparative study found CBCT produced roughly a 30 percent reduction in effective dose compared to medical CT scans of the same region.
Beyond choosing the right modality, clinicians can dial in a range of parameters to trim exposure further. A systematic review of radioprotective strategies in children laid out the key levers:
- Collimation: Narrowing the X-ray beam to cover only the area of interest, which is the single most effective dose-reduction tool across nearly every imaging type.
- Receptor speed: Using faster digital sensors or phosphor plates, which need less radiation to produce a usable image.
- Thyroid shielding: Recommended for intraoral radiographs when the thyroid is near the primary beam and for most CBCT protocols, though it can interfere with image quality in specific scan geometries.
- Exposure settings: Lowering tube voltage, reducing the number of projections, shortening exposure time, or using automatic exposure control to tailor dose to patient size.
- Voxel size: For CBCT, choosing the largest voxel size that still answers the clinical question reduces the amount of radiation needed.
All of these adjustments need to be balanced against diagnostic adequacy. An image so low-dose that it cannot answer the clinical question helps no one. The goal is matching the exposure to the specific indication, not blanket minimization.
Image Artifacts and the Limits of CBCT
CBCT is powerful, but it has real blind spots. Metal objects in or near the scan field, whether dental implants, posts, crowns, or orthodontic hardware, scatter the X-ray beam and produce streaking artifacts that can obscure adjacent anatomy. Metal artifact reduction (MAR) algorithms built into scanner software can improve image quality, but the problem does not disappear entirely. One study found that applying a MAR algorithm significantly improved contrast-to-noise ratios around both titanium and zirconia implants, though zirconia implants still generated more artifacts than titanium ones.
A subtler limitation involves tissue density measurement. Medical CT scanners report density in standardized Hounsfield units (HU), which lets clinicians compare values across scanners and over time. CBCT gray values do not map reliably onto Hounsfield units. The discrepancy varies by region within the same scan: one study found the anterior mandible showed the smallest gap between CBCT gray values and true HU, while posterior mandibular teeth produced the largest error.
This variability means you cannot take a CBCT gray value, treat it as a Hounsfield number, and make clinical decisions the way a physician would with a medical CT scan. Researchers have developed conversion factors for specific regions and scanners, and these can produce reasonably consistent bone-density estimates for tasks like implant planning. But the conversions are machine-specific and region-specific, not universal.
Artificial Intelligence in Dental Imaging
AI, and deep learning in particular, is rapidly entering maxillofacial radiology. The area attracting the most research attention is automated detection of dental caries (cavities) on panoramic radiographs. Cavities can be subtle on two-dimensional images, especially early ones between the teeth, and studies consistently show that dentists miss a meaningful fraction of them.
Deep learning models trained on large sets of annotated panoramic radiographs have achieved overall accuracy in the range of 93 to 94 percent for detecting and segmenting caries across different severity levels. One architecture called CariesNet, trained on over 3,100 annotated caries lesions spanning shallow, moderate, and deep decay, reported a mean Dice coefficient of about 94 percent, meaning the AI’s outlined caries boundaries closely matched expert annotations.
A separate clinical application study found that while overall accuracy and specificity were high (93 to 94 percent), the model’s precision for distinguishing true caries from false positives was more moderate, around 77 percent, particularly in posterior teeth. Recall, meaning the ability to catch actual caries, was much higher at 96 percent. In practical terms, these models are better at not missing cavities than they are at avoiding false alarms, which is arguably the right trade-off for a screening tool.
AI is also being explored for differentiating jaw lesions on panoramic images. Machine learning classifiers trained to distinguish ameloblastomas from odontogenic keratocysts on panoramic radiographs show promise, though the clinical workflow for integrating these tools is still being worked out. The pattern across dental AI research is consistent: the technology performs well enough to serve as a second reader or screening aid, but not yet well enough to replace a trained radiologist’s judgment on ambiguous cases.
Forensic Identification Through Dental Radiographs
Dental radiographs are one of INTERPOL’s three primary identification methods for disaster victim identification, alongside fingerprints and DNA. The logic is straightforward: teeth are among the most durable structures in the body, every person’s dental work is unique, and dental records including radiographs are widely available in countries with established dental care systems.
Forensic dental identification typically involves comparing antemortem images, such as panoramic radiographs, CBCT scans, bitewings, or periapical views, with postmortem radiographs taken during examination of remains. Matches can be made based on the pattern of fillings, missing teeth, root canal treatments, implants, and anatomical features like sinus shape or unique root curvatures.
An interesting development is the use of non-dental medical radiographs for dental identification. Lateral skull films, cervical spine images, and even chest X-rays occasionally capture enough dental anatomy to allow comparison. A Queensland case series documented successful identifications using non-dental antemortem radiographs, including what appears to be the first published forensic dental identification involving comparison of a postmortem CT multiplanar reformat image with a non-dental antemortem radiograph. This expands the pool of usable comparison images, which matters in cases where the deceased had no dental records but did have hospital imaging on file.
Dental radiographs can also help estimate age in cases involving minors, by assessing the stage of tooth development and root formation. In some contexts, they contribute to assessments of sex and ethnic background, though these applications are less precise than direct identification through dental work comparison.