AI is changing the face of orthodontics

English

简体中文 繁體中文 English Filipino हिन्दी Hmong Bahasa Indonesia 日本語 Basa Jawa 한국어 ພາສາລາວ Bahasa Melayu മലയാളം ဗမာစာ سنڌي தமிழ் ไทย Tiếng Việt Afrikaans Shqip العربية አማርኛ Հայերեն Azərbaycan dili Euskara Беларуская мова বাংলা Bosanski Български Català Cebuano Chichewa Corsu Italiano Hrvatski Čeština‎ Dansk Nederlands Esperanto Eesti Suomi Frysk Galego ქართული Ελληνικά ગુજરાતી Kreyol ayisyen Deutsch Français Harshen Hausa Ōlelo Hawaiʻi עִבְרִית Magyar Íslenska Igbo Gaeilge ಕನ್ನಡ Қазақ тілі ភាសាខ្មែរ كوردی‎ Кыргызча Latin Latviešu valoda Lietuvių kalba Lëtzebuergesch Македонски јазик Malagasy Maltese Te Reo Māori मराठी Монгол नेपाली Norsk bokmål پښتو فارسی Polski Português ਪੰਜਾਬੀ Română Русский Samoan Gàidhlig Српски језик Sesotho Shona සිංහල Slovenčina Slovenščina Afsoomaali Español Basa Sunda Kiswahili Svenska Тоҷикӣ తెలుగు Türkçe Українська اردو O‘zbekcha Cymraeg isiXhosa יידיש Yorùbá Zulu

Artificial intelligence (AI) is being applied across a wide range of dental fields. Orthodontics is no exception, with artificial intelligence algorithms and machine learning apps used for diagnosis, AI treatment planning and monitoring.

Many studies and innovations have been proposing – and bringing to light – the ways in which Artificial Intelligence can play a role in the world of orthodontics ever since Invisalign’s cloud-based software ClinCheck software was introduced to market in 1998.

The ClinCheck algorithm has enabled clinicians to map out and calculate the minute forces required to move the patient’s teeth to exacting precision. With this Dental AI innovation, the doctor is able to remotely view patient records on their computer screens while providing them with real time feedback about how well they are progressing.

AI applications for orthodontics

AI technology applied in orthodontics has evolved over time, but its basic premise remains unchanged. It generally uses digital data collected by sensors attached to patients’ braces or other devices, a machine learning algorithm creates customised 3D models of their mouths based on these measurements. These virtual images are then used as reference points for planning individualised treatments.

Orthodontists are using AI to combine 3D imaging and machine learning algorithms to analyse digital data that contribute to what will become a comprehensive portrait of the patient’s teeth, jaw and facial characteristics.

As you will see in this article, we can even use this information to predict future changes in your teeth and jaw — and the benefits of having such information could be enormous. let’s look at some practical examples.

AI treatment planning

Multilayer perceptron artificial neural networks are being used to aid orthodontists to formulate their treatment plans. It can be used to predict how much tooth movement is needed for each patient based on their unique set of dental features. This helps doctors make more informed decisions about which patients need braces or other treatments. The technology also allows them to identify potential problems before they occur.

​​To achieve satisfactory orthodontic outcomes, AI treatment planning must be carried out in a systematic manner and the process should include an evaluation of all available data including: Patient history, clinical examination, radiographs, photographs, models, study casts, cephalometric analysis, and 3D imaging such as cone beam computed tomography scans. Orthognathic surgery may be required if there are significant discrepancies between facial appearance and occlusion.

AI treatment planning for orthodontics is becoming more common due to its ability to provide accurate information about patient’s anatomy and function. The use of this technology has been shown to improve accuracy when compared with traditional methods. It also provides clinicians with additional tools that can help guide their decision making during diagnosis and planning stages.

An example is the AI treatment planning software Studio Pro 4 from Henry Schein Orthodontics. The artificial intelligence here is designed to help the patient find an ideal smile design based on their face shape, age, gender, ethnicity and other factors. It also helps them choose from among hundreds of available tooth aligners or braces options. In so doing, the use of AI helps to reduce the dentist’s assessment workload while improving diagnostic accuracy.

Artificial Neural Network (ANN) in orthodontics

The Artificial Neural Network (ANN) has the advantage of being able to learn from experience without requiring programming or training by human experts. It is also capable of learning complex relationships among variables that would otherwise require extensive manual coding.

As such, this makes it possible for ANNs to perform tasks which were previously considered impossible using traditional methods. Applied to planning an orthodontic procedure, the ANNs can predict how much tooth movement will occur in each stage of a planned course of action. The system then suggests alternative courses of action based on its predictions. If one of these alternatives results in less than optimal outcomes, the clinician can modify the plan accordingly.

AI predicts extraction decision

For the longest time, the question of whether or not to extract has been a contentious issue in orthodontics – with significant disparity noted in the decisions of different orthodontic practitioners.

In a bid to minimise the relative subjectivity of making the extraction decision, the artificial neural network (ANNs) was adopted as part of a study by The Department of Orthodontics at West China Hospital of Stomatology to develop systems able to make predictions based on extraction patterns.

According to the research article, these ANNs were able to predict the extraction decision with an accuracy of 94 per cent. Additionally, the system scored 84.2% and 92.8%, respectively for extraction pattern determination and anchorage pattern determination.

Studies suggest that an ANN can output the feasibilities of multiple extraction options thus allowing orthodontists greater flexibility. By highlighting important factors such as anchorage, it can even predict the amount of resistance to undesirable tooth movement.

The Department of Orthodontics at West China Hospital of Stomatology trained three neural networks with 302 cases to determine whether a patient needs tooth extraction.

Results from the studies conducted by the Department’s neural network models produced predictive accuracy of extraction reaching 94 per cent – higher previous prediction models. Notably, it was said to be one of the first studies, if not the first, to make a prediction of the use of maximum anchorage by an ANN. At 92.8 per cent, the level of accuracy shows that ANN can potentially assist orthodontists in making highly detailed treatment plans.

Conclusion

In addition to being used as an accurate ancillary tool for diagnostics, treatment planning and predicting procedural results, we have seen in this article the power of fully automated AI systems. They have the added advantage of being highly efficient as well as a huge time saver. They do not require human intervention or supervision during their operation. There is no need to spend hours manually inputting data into computers which would otherwise take up valuable clinical time.

These AI systems may be useful for clinical decision support, thereby improving the accuracy of diagnosis. Orthodontists can use Artificial Intelligence systems to help them make decisions about patient care such as whether they should extract teeth before placing braces on patients.

That said, clinicians should always rely on their own experienced judgement. The best way to ensure that a system does not replace human judgment is by using it alongside other tools and methods. For example, orthodontic treatment plans are often created with both computer-aided design software and traditional hand drawing techniques. This allows an orthodontist to have access to all available information at once while still relying on his or her experience when making important decisions.