How Far Can You Take Part in Digital Planning for Clear Aligner Treatment? From Intraoral Scans and 3D Models to Treatment Simulation and Plan Refinement

How Far Can You Take Part in Digital Planning for Clear Aligner Treatment? From Intraoral Scans and 3D Models to Treatment Simulation and Plan Refinement

Direct answer: You can take part at four points — during the intraoral scan, when the 3D model and digital setup are reviewed, when the treatment simulation is examined, and at every refinement; the simulation represents the planned objective rather than a result that is certain to be achieved; a three-dimensional study of 40 adults treated for mild to moderate crowding only measured an overall mean achievement of 50.3% for anterior alignment after the first series of aligners and before any refinement, and the study therefore recommended additional refinement [F1] (that figure comes from that population and cannot be transferred to more severe crowding or different movement types), and a three-dimensional study measured an overall mean achievement rate for anterior alignment of 50.3% after the first series of aligners and before any refinement, which is why that study recommended additional refinement [F1].

Geographic scope: Global. All cited evidence comes from international peer-reviewed literature; no country-specific reimbursement scheme or regulation is involved. The procedures, materials and fees actually available to you should follow your own health system and your dentist's assessment.

TL;DR｜The simulation is a ‘design drawing’, not a preview of the result

One of the most appealing aspects of clear aligner treatment is that, before treatment begins, you can watch an animation of your teeth moving on a screen. Many people ask the same question when they first see it: ‘So is this what I will look like in the end?’

That question deserves a careful answer, because the literature gives a clear one: the simulation represents the planned objective, not a result that is certain to be achieved.

A three-dimensional study of 40 adults and a total of 480 anterior teeth recorded that, after the first series of aligners had been completed and before any refinement, the overall mean achievement rate for anterior alignment was 50.3%. It was 47.6% in the maxilla and 53% in the mandible, and all four types of movement—vertical movement, mesiodistal rotation, mesiodistal crown tipping and buccolingual crown tipping—showed statistically significant differences between the planned and actual values (P < 0.05) [F1]. The study concluded that anterior alignment had not been achieved as planned with the first series of aligners and therefore recommended additional refinement [F1].

This is not bad news; it is something you should know in advance. The value of digital planning lies in making the plan understandable, discussable and adjustable along the way, not in guaranteeing the endpoint.

The four points at which you can participate in practice are set out below.

Point one: intraoral scanning—the first time you can truly ‘see’ your own mouth

With a conventional impression, impression material is placed in the mouth and held there while it sets. Intraoral scanning instead uses a scanner moved along the dental arches to build a three-dimensional model on the screen in real time.

What does the literature say about patients’ experiences?

A systematic review included 10 of 269 articles, published between 2016 and 2024, and compared conventional and digital impressions in terms of accuracy, chairside time and patient experience [F2]:

• All included studies had a cross-sectional observational design, with sample sizes ranging from 5 to 50 participants [F2]
• The scanners assessed included Cerec, Trios, iTero and Primescan [F2]
• All 10 studies assessed accuracy, but only 2 assessed working time and patient comfort [F2]
• Digital impressions were consistently reported to be faster and more comfortable for patients [F2]
• Findings on accuracy were mixed; most studies considered both methods clinically acceptable, while conventional impressions performed better in full-arch situations [F2]

A note on interpretation: the review itself explicitly calls for cautious interpretation because of high methodological heterogeneity, small samples and a participant population consisting mainly of young, fully dentate people [F2]. Assessment with QUADAS-2 found that the overall risk of bias for patient selection was ‘unclear’, while concerns about applicability were also mixed [F2].

In other words, the evidence consistently points towards intraoral scanning being more comfortable and faster. The current literature does not support, however, the claim that an intraoral scan is invariably more accurate.

What can you do at this point?

The model is on the screen as soon as the scan is complete. This is the first time you can view your bite from the outside—where the teeth are crowded, where spaces exist and how far a tooth has rotated. This is the best time to ask questions, because you and the dentist are looking at the same thing.

Point two: 3D models and the digital setup—how much is the computer actually doing?

Between the scan file and the aligners is a stage called the ‘digital setup’: each tooth is segmented and aligned, and its target position is arranged for every stage.

A scoping review of artificial intelligence applications in clear aligner treatment included 41 of 708 studies and also evaluated orthodontic software commercially available as of March 2024 [F3]:

• Of the 41 included studies, 16 concerned tooth segmentation, 4 concerned digital model alignment, 13 concerned digital setup and 8 concerned remote monitoring [F3]
• Artificial intelligence achieved 98% accuracy in tooth segmentation [F3]
• A total of 13 aligner software packages were identified, of which only 1 demonstrated full automation across every step of the digital orthodontic workflow [F3]

There is a very important and easily overlooked detail here: the review explicitly states that none of the 13 identified aligner software packages had been evaluated in any of the 41 included studies [F3].

A note on interpretation: the figure of ‘98% accuracy in AI segmentation’ therefore describes algorithm performance in a research setting; it is not equivalent to the performance of the commercial software used in your clinic. The literature currently treats these as separate matters, and one should not be substituted for the other.

What can you do at this point?

You can ask: Who created this plan? After the software arranged the teeth, what did the dentist change? A digital setup is not completed merely by pressing a button. The dentist’s modifications are often crucial. You are entitled to know where human judgement enters your plan.

Point three: treatment simulation—which movements are more predictable, and which are not?

This is the section on which you should spend the most time.

Predictability differs greatly between types of tooth movement

A systematic review searched English-language literature published from 1 January 2020 to 31 March 2025 and evaluated the accuracy of common tooth movements in clear aligner treatment [F4]. It concluded that:

• Distalisation and rotation can be expressed more predictably [F4]
• Transverse expansion, tipping/torque, deep-bite correction and especially intrusion are less reliable [F4]
• The review therefore recommended that treatment planning should anticipate underperformance, incorporate overcorrection, use attachments and auxiliaries judiciously, and ensure adequate patient education to improve adherence and satisfaction [F4]

A note on interpretation: the abstract of this review provides only its objective, methods and conclusions, without pooled statistical values [F4]. It is therefore suitable for understanding the direction of the findings, but not for citing a specific success-rate figure.

What does the actual discrepancy look like in anterior teeth?

Returning to the three-dimensional comparison study introduced at the outset, the figures are broken down in more detail [F1]. The study included 40 adults with mild to moderate crowding (mean age 27.8 years) and compared initial and final intraoral scans with the planned final digital model:

• Maxilla: distal tipping had the highest achievement rate (64.85%), followed by mesial rotation (63.32%); the predictability of intrusion was low (23.1%) [F1]
• Mandible: lingual tipping was highest (75.52%), followed by labial tipping (66.51%) [F1]
• The study specifically highlighted that intrusion of the maxillary incisors remains an area requiring attention [F1]

A note on interpretation: this was a study of 40 people, and it measured the state after the first series of aligners had been completed but before refinement [F1]. It does not say that ‘clear aligner treatment is only half effective’; it says that ‘after the first round, another round of refinement is generally required’—which is precisely why refinement exists.

What can you do at this point?

When viewing the simulation, replace ‘What will I look like in the end?’ with three more useful questions:

1. Which movements does my case rely on most? If it relies heavily on intrusion or transverse expansion, you should know that these categories are less reliable in the literature [F4].
2. Does this plan include overcorrection? This is an approach explicitly recommended in the literature [F4].
3. How many rounds are anticipated? Ask for refinement to be explained in advance as part of the process, rather than appearing later as an unexpected event.

Point four: refinement—it is part of the design, not an error

The severity of your crowding directly affects the discrepancy

A retrospective cohort analysis conducted in Saudi Arabia included 44 participants between May and September 2023 (months 5 to 9) (63% were women; mean age 32 years). They had Class I cases ranging from mild to severe crowding, and the treatment strategy was proclination followed by interproximal enamel reduction [F5]:

• All mandibular anterior teeth showed a significant discrepancy between ‘planned movement’ and ‘actual movement’ (3.64 ± 3.25) [F5]
• For each unit increase in crowding severity, the rotational discrepancy increased by 0.49 for the mandibular right canine and by 0.405 for the mandibular right lateral incisor [F5]
• The study concluded that dental crowding reduces the accuracy of clear aligner treatment; overcorrection should be routine in clinical practice, and additional auxiliaries should be used in markedly crowded cases to improve predictability [F5]

(In its background, the study cited that Invisalign could resolve at most 91.4% of mandibular incisor crowding, while also noting that evidence supporting the effectiveness of clear aligner treatment for different tooth movements remains insufficient [F5].)

The need for refinement varies greatly by tooth position

A retrospective study used data from 116 patients and 696 anterior teeth in a logistic regression analysis of factors associated with ‘the need for refinement’ [F6]:

• For tooth 11 (the maxillary right central incisor), greater planned rotation was associated with a small but statistically significant reduction in the likelihood of requiring refinement (odds ratio 0.92, p < .05) [F6]
• The effect of an attachment was in opposite directions at different tooth positions: it reduced refinement risk at tooth 21 (odds ratio 0.05, p < .01), but increased refinement risk at tooth 12 (odds ratio 4.95, p < .05) [F6]
• At tooth 22, greater planned extrusion was associated with a lower probability of refinement (odds ratio 0.34, p < .05), while greater rotation and inclination were associated with higher refinement rates (odds ratios 1.13 and 1.52, respectively; p < .05) [F6]
• No statistically significant predictors were found for the canines (teeth 13 and 23) [F6]

A note on interpretation: pay particular attention to the fact that ‘the same attachment reduced risk at tooth 21 but increased it at tooth 12’ [F6]. These opposing findings do not mean that attachments are either ‘good’ or ‘bad’; they indicate that biomechanical responses differ substantially between anterior teeth. The study’s own conclusion is that these findings support a personalised, tooth-specific treatment-planning strategy [F6]. Any claim that ‘this method applies to every tooth’ is therefore inconsistent with these data.

What can you do at this point?

When the dentist says, ‘We need to scan again and perform another round of refinement’, you can understand it this way: the actual result of the first round has now been measured, and a new plan will be made from the real positions. This is the feedback mechanism designed into the digital workflow from the outset.

Your contribution is very concrete: report whether you actually wore the aligners for the prescribed number of hours, whether any particular aligner clearly failed to fit, and which area seemed not to move. These are things the dentist cannot see on the screen; only you know them.

Bringing the four points together: what is your role?

One sentence runs through all four points: digital planning makes ‘discussion’ possible, but it does not remove biological uncertainty for you.

Risk factors: what to know before treatment

Before deciding on clear aligner treatment, these limitations are worth knowing:

• Predictability differs greatly between types of tooth movement: distalization and rotation can be expressed more predictably, whereas transverse expansion, tipping/torque and deep bite correction—intrusion in particular—are less reliable [F4]; in the three-dimensional comparison study, the achievement rate for maxillary intrusion was 23.1% [F1].
• Finishing the first series usually does not mean finishing treatment: after the first series of aligners and before any refinement, the overall mean achievement rate for anterior alignment was 50.3% (47.6% in the maxilla, 53% in the mandible), and all four types of movement showed statistically significant differences between planned and actual values, which is why the study recommended additional refinement [F1]. This is a population-level average, not a prediction for you.
• The more marked the crowding, the larger the discrepancy may be: retrospective data showed a significant discrepancy between planned and actual movement for all mandibular anterior teeth (3.64 ± 3.25); for each unit increase in crowding severity, the rotational discrepancy increased by 0.49 for the mandibular right canine and by 0.405 for the mandibular right lateral incisor [F5]. That study had only 44 participants and was carried out in a single country, so it is suitable as a guide to direction rather than as an individual prediction [F5].
• Whether refinement is needed varies by tooth, and attachments can even act in opposite directions: an attachment lowered the risk of refinement at tooth 21 (odds ratio 0.05) but raised it at tooth 12 (odds ratio 4.95); for the canines (teeth 13 and 23) no statistically significant predictor was found [F6]. Any claim that a single approach suits every tooth is therefore inconsistent with these data.
• Software performance in a research setting cannot be transferred directly to commercial products: the scoping review records 98% accuracy for artificial-intelligence tooth segmentation, but the same paper notes that none of the 13 aligner software packages identified had been evaluated in any of the 41 included studies [F3]. The systematic review recommends ensuring adequate patient education during planning, in order to improve compliance and satisfaction [F4].

Conclusion｜Make ‘understanding’ the starting point for your participation

What the digital workflow for clear aligner treatment truly gives patients is not an attractive animation, but four points at which they can speak up: during scanning, when viewing the 3D model, when examining the simulation and at every refinement.

The literature’s position on this is actually quite consistent. Digital tools make plans visible, discussable and measurable in terms of discrepancy. A three-dimensional study could calculate the figure of 50.3% precisely because the plan had been recorded and could be compared retrospectively [F1]. Yet the same body of literature repeatedly reminds us that predictability varies by type of movement [F4], by severity of crowding [F5] and by tooth position [F6], while software performance in a research setting cannot be directly inferred for commercial products [F3].

The most practical mindset is therefore this: treat the simulation as a shared language between you and your dentist, not as a guarantee of the outcome.

If you are considering clear aligner treatment, take these questions to your dentist: Which movements does my case rely on most? Does the plan incorporate overcorrection? How many rounds of refinement are anticipated? And—when the actual result differs from the screen, how will we adjust?

There is no standard answer to these questions, but they will help you move from ‘the person watching the animation’ to ‘the person participating in the plan’.

*This article is a review of the literature, not personalised medical advice. The cited studies differ in sample size, follow-up period and applicable population. Please discuss your individual circumstances with your dentist.*

Every clinical statement in this article is mapped line by line to its cited source (see the sources and evidence chain below). It has not been clinically reviewed by a licensed practitioner. This is health information, not individual advice; assessment by a clinician is required.