Herringbone artifact

The herringbone artifact is the pattern a dental radiograph carries when the film packet was exposed from the wrong side: a repeating herringbone or tractor-tread texture across the image. What prints it is the embossed lead foil at the back of the packet, which the beam had to cross before it reached the film. It is a film-era finding.

Synonyms: reversed film, backwards receptor, reversed sensor, tractor tread pattern, tyre track pattern, tire track artifact, herringbone pattern

What does the herringbone artifact look like, and why does it appear?

Every receptor has one surface meant to face the tube — film, phosphor plate and wired sensor alike. Film carries an embossed dot to tell you which one it is.[3]

Turn the packet round and the beam meets the back of it first. The back of a film packet is lead foil, and that foil is embossed. The pattern you get on the processed radiograph is that embossing, printed through onto the emulsion. It has been described since at least 1990 as a herringbone or a set of tractor treads, and it is named there as a well-known consequence of backward film positioning in the mouth.[1]

The image is also faint, and the foil is why. One 1992 bench study of five commercial dental film packets measured what the foil does to the beam: on average it attenuated the primary beam by 77% at 60 kVp and 56% at 90 kVp.[2] That foil was put there to stop backscatter reaching the emulsion from behind. Exposed the wrong way round, it stands between the tube and the film instead, and most of the beam never arrives.

So the classic teaching case gives you two findings at once — a pattern that shouldn't be there, over an image that is too light to read comfortably.

Note what has to be true for any of that to happen. There has to be a lead foil. That is what makes this a radiographic artifact of a particular era rather than a permanent feature of the job.

What does a reversed receptor look like on a phosphor plate, where there's no herringbone?

The lead-lining explanation is specific to film. The herringbone clue does not transfer.[3]

This is stated directly in the literature rather than inferred. A 2014 forensic odontology report contrasts the two: the characteristic herringbone appearance on analogue periapical film comes from the lead lining on the reverse of the packet, and that clue is not apparent in phosphor plate receptors. In at least one older plate model, the report notes, a wrong-side exposure leaves no visual clue at all that the plate was reversed.[3]

What you get instead is a laterally inverted image — a mirror of the one you meant to take. In the same paper's demonstration, a right bitewing was deliberately exposed with the wrong side of the plate facing the tube. The only sign of it was the orientation indicator sitting in the wrong corner of the image. Contrast and clarity were not affected.[3] That is one author's demonstration on one plate model, so read it as a warning about what you cannot rely on rather than a rule about every plate; but it points the opposite way from what most teaching implies, and it is the evidence retrieved for this entry.

Two things make it worse rather than better. Newer plates carry an incorporated indicator to help you orient the image — which means the fix and the trap arrive together, because viewing software will happily rotate and laterally invert an image, and an attempt to line the indicator up correctly can invert a correct image into a wrong one. And plates have no attached wire, unlike wired sensors, which is exactly why they handle like film in the first place.[3]

One gap worth naming, because the sources don't close it: what a wired sensor looks like when it is exposed from the wrong side is not described in the literature retrieved for this entry. You will see it asserted that the image comes back blank or white. That may well be right, but it is not sourced here, and reasoning across from film or from a phosphor plate is exactly the move this entry exists to argue against.

How do you recognise a reversed receptor from the image alone?

The orientation marker is the discriminator on both media, and it is the thing to look at first. On film it is the embossed dot. On a phosphor plate it is the incorporated indicator.[3] If the marker is in a corner it should not be in, treat the image as suspect before you interpret anything on it.

This is not a rare slip. A 2019 study of 5,000 periapical images from one dental hospital found operator error to be the largest category of phosphor plate fault at 34.7%, with projection geometry errors the main errors within that group at 31.9%. The second most common group was placement of the plate in the mouth, and within that group the most common single error was an incorrectly positioned dot.[4]

Density will not do this job for you. On film, a light image plus a printed pattern is a reversed packet; a light image with nothing printed on it is underexposure from some other cause. On a phosphor plate you lose even that, because plates have a wide dynamic range that lets under- and over-exposed images be corrected[4] — the same latitude that makes them forgiving makes brightness a poor witness.

The other check costs nothing: read the anatomy. If the tooth shapes, the restorations and the bone pattern don't match the side the image is labelled, believe the anatomy.

What it's confused with

Can the image still be read?

On film, usually not well — the pattern sits over the anatomy and the image is faint for the reason described above.

On a phosphor plate the answer is more uncomfortable, because the image may be perfectly readable and still be wrong. The 2014 report is explicit about the cost: failure to spot the error can lead to a procedure being started on the wrong tooth or site, and the risk is highest in mouths with few or no restorations, where there is nothing distinctive to orient by.[3] A diagnostically clean bitewing of the wrong side is a more dangerous object than an obviously ruined one, because nothing about it asks to be checked.

That is also why the error outlives the appointment. The image goes into the record, and the next person to open it — a colleague, a referral, anyone reading it months later — has only the marker and the anatomy to go on.

For scale on how often digital receptors get re-exposed at all: one 2022 series of 50,000 intraoral phosphor plate radiographs at a single university clinic recorded 3,550 retakes for image artifacts, 7.1% of the total, with operator error the dominant reason at 5%.[5]

This entry doesn't give you a retake threshold. No professional-body guidance for one was retrievable for this page, and a number that merely sounded official would be worse than the gap. What stands is the clinical statement: an image whose orientation you cannot establish is not an image you can safely act on.