Can the Edentulous Mouth Really Be Recorded Digitally?
Suppose a patient has come to us for a complete denture. Instead of a conventional impression, we pick up the intraoral scanner and scan both jaws. A three-dimensional image of the arches appears on the monitor, and we can now continue the design and fabrication of the denture digitally.
But is the information the scanner has recorded really enough to make a complete denture?
In fixed prosthodontics, the main goal of scanning is to record precisely the geometry of the teeth and their surrounding tissues. But in a complete denture the situation is different. The prosthesis rests on a combination of fixed and movable tissues, and its function does not depend only on the fit of its inner surface to the mucosa; appropriate extension of the borders and their relationship to muscle movements also matter.
For example, in a functional impression, we use border molding to try to define the appropriate extent of the denture according to the movements of the surrounding tissues. The aim is for the border of the prosthesis to be neither so long that muscle movement displaces it nor so short that part of the support capacity and the peripheral seal is lost.
In contrast, an intraoral scanner records the shape of the mucosal surface only in its unloaded state; something close to the concept of a mucostatic impression.
The scanner records the shape of an area with adequate accuracy, but it does not give us enough information about how that same tissue behaves during function.
Of course, conventional impressions have their own limitations too. Pressure control, dimensional changes in impression materials and errors in making the stone cast can all affect the result. Going digital eliminates some of these problems, but in exchange we face other limitations.
One of the most important of these is the way the scanner reconstructs the 3D model.
The scanner captures successive images, and the software builds the 3D model by joining them together, a process called stitching.
In a dentate arch, the geometric features of the teeth help the software place the images together correctly. But in a completely edentulous jaw, the mucosal surface is relatively uniform and there are fewer distinct landmarks for joining the images. For this reason, the likelihood of error accumulating during a full-arch scan increases.
Saliva and soft-tissue movement complicate the problem further. For example, when scanning the vestibule we usually have to retract the lip or cheek, yet this very action may change the position of the mucosa.
So even if the captured image is accurate, we still have to ask: in what position was this tissue recorded?
To examine this question, Rasaie et al. published a systematic review in 2021 on the accuracy of intraoral scanners in recording denture-bearing areas.
The researchers examined laboratory and clinical studies, and the results showed that scanning accuracy is not the same across different parts of the arch.
In areas with relatively immobile mucosa, such as parts of the residual ridge and the hard palate, scanners showed acceptable accuracy in many studies. But in mobile areas, such as the vestibule, the soft palate and the sublingual regions, discrepancies and recording problems were greater.
In one of the studies, when the mobile peripheral areas were excluded from the evaluation, the measured error decreased sharply.
This finding shows that part of the problem with scanning the edentulous jaw relates to the characteristics of mobile tissue, and not just to the extent of the arch.
In the same review, recording the mandible was usually more difficult. The presence of the tongue, movement of the floor of the mouth and the harder control of saliva may play a role in this.
However, in some studies, when only the immobile surfaces were evaluated, the accuracy of mandibular scans was even better than that of maxillary scans.
So we cannot say in absolute terms that the scanner is less accurate in the mandible; we have to specify which tissue and which part of the arch we are examining.
The researchers had also examined strategies such as using auxiliary markers, changing the scanning path and choosing a suitable scanner tip. Some of these changes improved accuracy, particularly in laboratory studies; but there is still no standard, reliable technique for all clinical situations.
Another limitation is that many of the studies were carried out on static models. Good accuracy on a laboratory model does not necessarily mean that the same area will be recorded successfully in the patient's mouth. In addition, patients with severely resorbed ridges or severe undercuts were not included in this review's clinical studies of completely edentulous arches.
Now let us turn to the second article.
In 2025, Nguyen et al. published a scoping review that asked a different question: how is functional recording of completely edentulous arches carried out in a digital workflow?
The key finding of this review was that, among the methods examined, there was still no fully digital technique for functional recording of edentulous arches.
This finding does not mean that a digital denture cannot be made. The issue is that in many workflows the functional recording stage is still carried out conventionally or is omitted altogether.
In some methods, an intraoral scan is obtained first, but the final recording of the borders is done with a conventional impression. In other methods, a functional impression is made first and the resulting impression or cast is then digitized.
We call this combination of conventional and digital methods a hybrid workflow.
The advantage of the hybrid approach is that we can record the functional information conventionally but continue the design and fabrication of the prosthesis in the digital environment.
On the other hand, methods have also been reported in which the denture was made directly from an intraoral scan, without a functional impression.
In some case reports, particularly in the maxilla with favorable anatomic conditions, early results in terms of retention and patient satisfaction were favorable.
But this evidence is limited. A case report cannot prove the equivalence or superiority of a method, and in some cases, despite a good early result, a reline was later needed.
There is one more difference: recording the geometry of the mucosa is not the same as recording its behavior under pressure.
In a selective pressure impression, we can control the distribution of pressure across different areas. But the scanner by itself does not record such information.
As a result, when we use a direct scan alone, part of the decision-making about relief, tissue pressure and the functional extension of the borders is shifted to the design stages.
So how does this information help in clinical decision-making?
First, we need to distinguish between using a scan for the preliminary impression and using it as the sole basis for making the definitive denture.
The evidence from the Rasaie review shows that an intraoral scan can be used for the preliminary recording of many edentulous arches. But for the final impression, in addition to the geometry of the ridge and palate, we also need sufficient information about the peripheral areas and the behavior of the tissues.
For a patient with a favorable maxilla, an adequate ridge and stable mucosa, wider use of direct scanning may be worth considering. But in more difficult situations, such as a severely resorbed ridge, mobile mucosa close to the crest or functional problems in the mandible, relying on direct scanning alone lacks sufficient support.
In such situations, a hybrid approach may be a more reasonable choice; that is, we use the scan to record the anatomy and the conventional impression for the information that depends on the movement and behavior of the tissue.
However, we should not conclude from these two reviews that the hybrid method is superior on every clinical measure. The first review focused mainly on the accuracy of surface recording, and the second identified and categorized the existing methods.
Therefore, there is still not enough evidence for a definitive conclusion about the effect of each workflow on denture longevity, the number of post-insertion visits or long-term patient satisfaction.
In the end, the main question is not whether digital scanning or the conventional impression is better.
The question is what information we need to make the prosthesis, and which part of that information each method records better.
The scanner can be a valuable tool for recording tissue geometry, but geometric accuracy alone does not guarantee that functional information is recorded.
Perhaps the reasonable choice for now is, rather than insisting on digitizing every step, to use each technology where the evidence supports its capability.