Edited by Kimberley Chee.
“All right girl, we're gonna to pull out one tooth today, so I’ll apply some strawberry ointment on the extraction site.”
8 years ago, I started to prepare my mouth cavity for my future orthodontic treatment. At the age of 12, my primary teeth won over the permanent ones in terms of number. I did not know why. After several rounds of painful extractions came the beginning of my braces journey, and the deep yearning for this torturing dental treatment to end within 2 years, as is the case for most people.
Did my dream come true? Sadly, it didn’t. What followed in the next 7 years were countless X-ray inspections, follow-up treatments, and — operations. For my family and I, the news of undergoing an operation just for orthodontic treatment was indeed a bombshell, as it was not a common procedure for braces patients to have to go through, to our knowledge.
However, I now know at least one of the reasons behind this atypical event: diagnostic error.
When I was instructed to undergo yet another surgery recently, I wondered, can Artificial Intelligence (AI) drive dentistry to greater heights? In ways where patients can know the progress of their treatment with certainty, and exactly how long their treatment is supposed to take; where orthodontists know which medical approach should be taken to reach optimal treatment outcome, in the most ideal conditions? I believe we have yet to step into that era, but we are definitely taking our first steps towards it.
To everyone undergoing orthodontic treatment, we are all too familiar with that first step: scanning our teeth for reference using X-ray or traditional plaster models. Well, are you familiar with the fact that X-rays can harm our cells, and even cause mutations in the worst case scenario? And what of those innocent-looking moulds? News flash! Using traditional moulds can cause unbearable pain. Hence, we are looking for alternatives to augment the “scanning skill” in orthodontics.
Lo and behold, we now have digital impressions, which are essentially digital oral scans that provide dentists with a complete illustration of the patient's gums and teeth.1 3D Visualization allows orthodontists and patients themselves to view teeth from different angles so orthodontists can come out with better plans for patients’ treatment. Additionally, it also aids orthodontists in placing brackets, determining the 3D position of dental implants and fabricating bars and bridges.
Aside from that, digital impressions have not only given more comfort to patients, especially those who have gag reflexes, but also reduced the probability of getting errors and distortions. Recent research has revealed that scanners manufactured in the same batch but by varied companies produce similar if not indistinguishable results in terms of precision. Therefore, experimentations to come should emphasise on innovative applications for digital impressions in order to create novel but effective orthodontic practices.2 When the use of digital impressions starts to become prevalent in all orthodontics clinics, patients will feel more secure and engaged when they themselves are able to watch their dental problem being displayed live on screen. What’s more, patients gain the chance to question orthodontists when something is amiss for them, compared to the traditional way where the final decision always rests with the orthodontists, and patients have no say in their treatment.
It was suggested that the mean orthodontic treatment time was about 23.5 months, although it is still affected by certain factors that differ for each patient, such as distinct choices made by orthodontists, and even their individual traits.3 In other words, we need something to help orthodontists make more accurate decisions. That being said, this is where the significance of Augmented Reality (AR) comes into play. As we all know, AR helps to enhance information of the real, physical world and overcome mankind’s shortcomings that happen when we can only analyse things using our eyes.
Admittedly, the efficiency of orthodontic treatment also depends largely on the ideal bracket position. Some ways for ideal bracket positioning have been proposed, but the AR-assisted bracket navigation system demonstrates more strengths over the others, such as increased safety. The overall bracket positioning is always determined by our orthodontists who are professional in this matter, but there is also the possibility of wrong judgement and decisions — we are all humans after all. Nevertheless, it is undoubtedly difficult to replace orthodontists with robots for treatments and even surgeries, but the unquestionable statement is that we can certainly make good use of the latest technologies to reduce mistakes.
In 2012, Aichert used a Computed Tomography (CT) scan to track teeth in a video image, which served as a guide in a monocular AR system.4 A few years later, Cone Beam Computed Tomography (CBCT) was also suggested to provide high definition photographs for orthodontic evaluations.5 Nevertheless, both of them can expose patients to high doses of radiation. On the other hand, the AR-assisted bracket navigation system is safer to be used as radiation is not emitted. The system is also equipped with a wireless intraoral camera to take oral images that are later sent for data-processing. Real-time natural feature registration is utilised due to its high accuracy to match the ideal bracket position on the tooth surface. Thus, the precise bracket position can be determined using digital modelling with an intra-oral scanner. However, frankly speaking, such a system should be investigated further, in order to overcome the potential underlying problems before it is open to the public, as people might question the feasibility of AR for orthodontics. Moreover, any advantages should be strengthened to further benefit patients in the long run.
Due to the rapid development of computers, the applications of AI in a wide array of sectors has garnered worldwide attention, including the medical field. In 2019, a Dental Monitoring (DM) app, which was originally a selfie tool, was introduced in orthodontists’ clinics.6 Patients are responsible for taking smiling pictures using the ScanBox. After that, orthodontists will communicate with the patients using the app upon receiving their pictures. The app assists orthodontists in tracking the progress of patients between appointments, thus discovering recessive problems as soon as possible to avoid waste of time and labour. Besides, patients can utilise the photo time-lapse technology, which depicts the overall progress of their treatment. Although it is a huge step to apply 3D models and AI in orthodontics, I believe it should be utilised in both government and private orthodontic clinics for the benefit of people. In fact, the appliance of AI is not limited to these kinds of apps, there are still countless other mediums that we can creatively use to help dental patients.
Nonetheless, to further utilise technology in dentistry, in hopes of avoiding pain and suffering of patients like myself, further studies should emphasise on the combination of the AR-assisted bracket navigation system with AI models. Developments in this technology can enhance teeth detection and image extraction to locate the growing teeth precisely7, find out hidden teeth, and integrate other optimised algorithms related to orthodontic treatment, thereby increasing effectiveness and reliability.
In conclusion, although artificial intelligence technology has not yet been widely used in clinical practices, with the continuous development and advancement of technology, it is believed that AI will bring a new era of technological innovation to orthodontic treatment in the near future. I hope that, by increasing applications of AI and technologies in orthodontic treatment across the globe, the duration for treatment can be greatly reduced, resulting in less pain and suffering for those who wish to have healthier, and/or tidier teeth. AI also has the potential to decrease the possibilities of diagnostic errors, clinical revisits, and money wasted.
But most importantly, we can smile with confidence, knowing that our teeth are in optimum health, thanks to the help of computer science.
References:
Inoue, A. (2019, December 5). Digital Impressions for Braces and Orthodontics. Dental Associates. Retrieved June, 2022, from https://www.dentalassociates.com/blog/digital-impressions-braces-and-orthodontics
Jedliński, M., Mazur, M., Grocholewicz, K., & Janiszewska, J. (2021, 1 27). 3D Scanners in Orthodontics—Current Knowledge and Future Perspectives—A Systematic Review. MDPI. Retrieved June, 2022, from https://www.mdpi.com/1660-4601/18/3/1121
Skidmore, K. J., Brook, K. J., Thomson, W. M., & Harding, W. J. (2005, 10 3). Factors influencing treatment time in orthodontic patients. PubMed. Retrieved June, 2022, from https://pubmed.ncbi.nlm.nih.gov/16473715/
Aichert, A. (2012, - -). Image-Based Tracking of the Teeth for Orthodontic Augmented Reality. Springer Link. https://link.springer.com/chapter/10.1007/978-3-642-33418-4_74#
Coşkun, İ., & Kaya, B. (2018, 6 31). Cone Beam Computed Tomography in Orthodontics - PMC. NCBI. Retrieved June, 2022, from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6046632/
Hurwitz, S. (2019, Summer Summer). Artificial Intelligence at the Orthodontist's Office. ARCH Orthodontics. Retrieved 2022, from https://archorthodontics.com/artificial-intelligence-at-the-orthodontists-office/
Hu, C., Zhang, K., Lyu, P., Li, H., Zhang, L., Wu, J., & Lee, C. H. (2019, 3 7). A deep learning approach to automatic teeth detection and numbering based on object detection in dental periapical films. Nature. https://www.nature.com/articles/s41598-019-40414-y
Comments