Panoramic radiograph

A panoramic radiograph is a single image of both jaws, made by a source and receptor that rotate around the patient. Only a curved zone called the focal trough is rendered sharply; structures outside it blur, shift or superimpose. It is a survey view — good for questions larger than an intraoral receptor can hold, weak on fine detail.

Synonyms: OPG; orthopantomogram; OPT; DPT; panoramic x-ray

How is the image formed, and what is the focal trough?

Start with what the machine is doing, because almost everything else on this page follows from it.

A panoramic radiograph is not a single projection. The source of the beam and the receptor both move, travelling around the patient's head on opposite sides while the exposure runs. That movement is tomographic, built around one or more centres of rotation, and the image you end up looking at is assembled from the whole sweep rather than captured in one shot.

The point of the movement is to hold one zone of the jaws in focus and smear everything else. That zone is the focal trough. Structures inside it are rendered acceptably sharp. Structures outside it are blurred, displaced, or lost. Crucially, the focal trough is a property of the machine — it is the plane the equipment is built to image.

It is also not a neat slab. A 2021 phantom study mapped the image layers of one multilayer panoramic unit and found every layer curved, running through 39° to 51° across both arches, and varying in both position and thickness. The layers were thicker at the mandible than at the maxilla, and the anterior maxilla sat further forward than the anterior mandible. That was one device with a five-layer program, so read the angles as a description of that machine rather than as a universal geometry. The qualitative finding is the one that travels: the trough is curved, and it is not the same thickness everywhere in the mouth.

What is a panoramic radiograph good for, and what does it miss?

It is the right examination when the question is bigger than an intraoral receptor can hold.

The UK Faculty of General Dental Practice's selection criteria, in the third edition updated in 2018, list four situations: a bony lesion or unerupted tooth too large or too awkwardly placed to show completely on intraoral radiographs, a grossly neglected dentition where multiple extractions are clinically likely, assessment of third molars before planned surgery, and orthodontic assessment where there is a clinical need to know the state of the dentition and which teeth are present. The ADA and FDA patient-selection guidance, revised in 2012, points the same way — a panoramic or periapical examination to assess developing third molars in adolescents, and posterior bitewings with a panoramic examination as one option in children with transitional dentition. For adults, that guidance records panoramic as usually not indicated for monitoring growth and development.

What it misses is detail, and the guidance is unusually direct about it. Intraoral radiographs outperform panoramic radiographs for caries, periodontal disease and periapical disease. They are also more accurate for dentoalveolar trauma, root shape, root resorption and pulp pathosis. A panoramic radiograph is the first-choice view for a mandibular fracture, though poor image quality has been shown to limit that accuracy, and it has limited ability to detect mid-facial fractures. In patients with TMJ symptoms, one study found it contributed little or nothing that changed diagnosis or management.

And it is not a screening tool. Both bodies say so plainly. The FGDP text states that routine radiography without clinical signs or symptoms cannot be justified; the ADA and FDA text states that radiographic screening to detect disease before clinical examination should not be performed. The FGDP's own commentary notes a study in which 63% of general dental practitioners routinely screened new adult patients with a panoramic radiograph — a gap between the guidance and the habit that is worth knowing about before you inherit the habit.

What it's confused with

The useful comparison here is not "which image looks similar" but "which view answers which question".

If you want the reasoning behind choosing between the two intraoral views, that is its own comparison: bitewing versus periapical.

Why can't you diagnose proximal caries on one?

The first answer is not about resolution. It is that the surfaces are not there to read.

In a 2016 clinical comparison across 20 patients, four observers assessed 489 non-restored readable surfaces. A surface counted as non-readable when more than half the enamel was overlapped or the surface simply could not be seen. On standard panoramic images, 51.5% of proximal surfaces were non-readable. On intraoral bitewings, 4.1% were. The same study found open posterior contacts visible on 48.5% of standard panoramic images against 95.9% of intraoral bitewings.

Read that again as a reading problem rather than a statistic. On a panoramic radiograph, roughly half the proximal surfaces are hidden behind the neighbouring tooth. You cannot grade what you cannot see, and you cannot know from the image alone which half you lost. That is the same failure you already know from overlapped contacts on a bitewing — except that on a panoramic it is the normal condition, not a technique error.

Where the surfaces can be read, accuracy is still lower. A 1994 clinical comparison had 18 viewers assess 35 sets of radiographs; the area under the ROC curve for detecting proximal caries was 0.69 ± 0.03 for a standard panoramic projection and 0.68 ± 0.03 for an orthogonal panoramic projection, against 0.79 ± 0.03 for bitewings. That study used film-era equipment and a consensus radiographic reference standard rather than histology, so treat the numbers as historical. But a 2012 ex vivo study of 80 extracted premolars and molars, on a digital panoramic unit and against a histological gold standard, reached the same direction: the highest ROC values were for intraoral bitewing, and the authors concluded it was superior to panoramic radiography.

The FGDP guidance summarises the whole body of work in one sentence — the published studies on diagnostic accuracy, sensitivity and specificity comparing panoramic to intraoral radiography indicate that panoramic is generally inferior for imaging the dental diseases with the greatest prevalence. Its grade A recommendation follows: where radiographs are clinically indicated, intraoral radiography should be considered the optimal examination, because of better detail and lower radiation doses.

So the answer to "can it show caries" is that it will sometimes show a large lesion, and that this is not the same as being able to grade depth or to rule caries out. If the clinical question is proximal caries, the answer comes off a bitewing — and how early you can actually see it there is a separate limit worth knowing.

Why can't you measure or localise reliably on a panoramic?

Because magnification on a panoramic radiograph is both large and uneven.

The FGDP guidance puts it at up to 30%, and adds the part that matters more: it may vary significantly at different locations within the same radiograph, depending on the equipment. So a millimetre near the midline and a millimetre at the ramus are not the same millimetre.

The distortion is also not symmetrical between the two axes. Vertical measurements are affected much less than horizontal ones. A study of 32 implants of known 10 mm length, imaged on 17 postoperative panoramic radiographs from a single unit, found the vertical magnification factor stable at 1.27 across the premolar and molar regions of the posterior mandible — the authors themselves asked for that stability to be confirmed on other units. Horizontal distances behave far worse. In a study of 30 skulls imaged in ideal, shifted and rotated positions, the largest errors were in horizontal measurements, and applying the unit's stated magnification factor was consistently less accurate than calibrating against an object of known size placed in the image.

That same skull study is the clearest demonstration that this is geometry rather than operator error. Panoramic accuracy for detecting mandibular asymmetry, using an internal reference, was 67% with the skull ideally positioned, 70% rotated and 47% shifted. Even at ideal position, a third of the comparisons were wrong.

Two further consequences of the construction:

Superimposition is built in. A panoramic radiograph carries ghost images that intraoral bitewings do not — a real structure reappearing as a second, false one somewhere it is not.

Localisation in depth is not available. Where a mandibular third molar root appears close to the inferior dental canal, the guidance states directly that panoramic radiography does not provide an accurate indication of a close relationship. What settles it is a second view with different projection geometry, or a localised CBCT — and the same guidance notes there is no evidence supporting routine CBCT in these cases, and warns against it becoming defensive practice.

How does the dose compare with intraoral radiographs?

Closer than most people assume, and the ranges matter more than the means.

The FGDP's literature summary gives effective dose for a panoramic radiograph as 2.7 to 38 µSv, and for an intraoral radiograph — bitewing or periapical — as 0.3 to 21.6 µSv. Those ranges overlap heavily. The two modalities sit in the same order of magnitude, and the spread within each one is wider than the gap between them. The same table gives 11 to 214 µSv for a small field-of-view CBCT and 18 to 674 µSv for a medium field of view. The document attaches its own caveat: effective doses are calculated for a reference patient, and individual risk estimates may be higher or lower by a factor of five.

A separate review of dental dosimetry papers published between 2010 and 2020 reports mean effective doses of 17.93 µSv for a panoramic procedure, with a range of 3.47 to 75.00, and 1.32 µSv for a single intraoral radiograph, with a range of 0.60 to 2.56. Those figures come from phantom dosimetry and Monte Carlo simulation rather than patient measurement, and the panoramic mean rests on nine papers. Note what the intraoral figure is: one radiograph, not a set. Any comparison against a bitewing set has to say how many radiographs are in the set, or it is not a comparison at all.

Expressed the way patients tend to ask, the FGDP guidance sets a periapical radiograph at an effective dose of up to about three and a half days of background radiation, and a panoramic radiograph at no more than about six days, against a UK annual background of about 2230 µSv.

The honest summary: the panoramic radiograph's advantage is coverage per exposure, not a dramatic dose saving. The ADA and FDA guidance says panoramic examinations may have the advantage of reduced radiation dose, cost, and imaging of a larger area — and the ordering against CBCT is where the dose argument actually bites, since CBCT delivers roughly seven times the effective dose of a panoramic procedure.

Common errors

Treating a clear panoramic as a clean bill of health. Half the proximal surfaces were unreadable, and the image does not tell you which half.

Measuring off it with the manufacturer's magnification factor. That was consistently less accurate than calibrating against an object of known size in the image — and it under-estimated real dimensions.

Trusting a horizontal measurement the way you would trust a vertical one. Vertical is much less affected. Horizontal is where the error lives.

Taking one because the patient is new. Both the UK and US guidance say a radiograph follows a clinical finding rather than preceding one.

Confusing the extraoral bitewing mode with the panoramic radiograph. Several panoramic machines offer a bitewing-mode program, and it performs much closer to an intraoral bitewing than a standard panoramic does. It is a different acquisition. Do not read results from one as results for the other.