AI For Improved Orthodontic Outcomes: Evidence And Outcomes

Warum Verantwortung und Feedback für deinen Praxisalltag so wichtig sind…

Orthodontics is a complex field that requires precise measurements, deep analysis, and accurate treatment. Just like a clockwork, you get optimal performance only when all the clogs work perfectly. This is why orthodontists need a lot of expertise to get the desired outcomes.

Research shows that the knowledge level of general dentists (GDs) regarding orthodontic treatment needs is moderate. Major problems faced with GDs taking on orthodontic cases include inaccurate and postponed treatments. Therefore, the type and timing of treatment are of crucial value. Training dentists can be a hectic and laborious task but with the advent of artificial intelligence (AI), we have seen a revolution in the field of orthodontics.

This article sheds some light on how AI transforms orthodontic practice and improves orthodontic outcomes.

How AI Improves Orthodontic Outcomes?

AI incorporation in multiple steps of the treatment allows quicker, and more accurate and superior results. There are plenty of ways in which AI can pleasantly change orthodontic results.

Diagnosis

Needless to say that the primary factor in the success of any medical/dental treatment is an accurate diagnosis. Orthodontic treatment requires a lot of data analysis and the identification of diagnostic factors.

Research shows that accurate diagnosis is the key to treatment success. It further concludes that accurate imaging (3D scans and 2D cephalograms) and analysis is a must to reveal the “anatomic truth” of the patient’s features. Landmark identification on cephalometric images is an integral part of orthodontic diagnosis. By performing highly accurate and efficient landmark identification, AI creates chances for better treatment planning and consequently superior results.

Studies have also revealed that clinicians spend an average of 7 to 15 minutes manually placing landmarks in each image. This makes the diagnostic procedure time-consuming. With greater time and effort involvement comes an inherent risk of human error. On the other hand, the mean computational time taken per image by YOLOv3 and SSD was 0.05 seconds and 2.89 seconds for SSD (astonishing! Isn’t it?).

Moreover, the automated (AI-based) method identified 80 landmarks on cephalometric images. A 2021 study revealed that AI programs like YOLOv3 (You Only Look Once, Version 3) are based on deep learning convolutional neural networks (CNNs). These programs are not only fast but also highly efficient. In some cases, the automated landmark identification was more precise than humans.

Another aspect of orthodontic diagnosis enhanced with AI is better visualization of outcomes. Evidence suggests that AI has the capability to accurately visualize post-treatment outcomes of orthodontic treatment. This way, orthodontists can plan and alter treatment strategies to obtain the maximum output.

Age Determination

A fundamental facet of good orthodontic treatment is the determination of skeletal age. Studies reveal that the age of the patient plays a significant role in tooth movement (during orthodontic treatment). Younger patients have better tooth movement than older ones. For excellent results, an orthodontist needs to determine the skeletal maturation stage of the patient. It serves as a diagnostic tool for the prepubertal growth period.

Improper determination of skeletal age may delay orthodontic treatment. Numerous studies have concluded that as the age advances, there is a delay in orthodontic tooth movement. This can potentially lead to under-treatment, incomplete treatment, and/or reduced patient satisfaction.

Fortunately, AI does an exceptional job in the age determination department too. Be it wrist X-rays or Cervical Vertebral Maturation assessment, AI performs tremendously in deciding the skeletal stage of the patient.

Modern reports show that with AI, orthodontists can spare complicated intermediary steps for skeletal diagnosis and rely on CNN-incorporated AI systems. Moreover, AI-based systems consider skeletal maturity indicators which help in deciding the ideal timing and method of treatment. This increases the treatment efficiency and serves to be a clinically reliable tool.

Artificial intelligence-powered systems are also capable of carrying out analysis of the temporomandibular joint (TMJ). Disorders of TMJ that may negatively impact the treatment results are pre-emptively pointed out for transparent planning.

Treatment Planning

Orthodontists are required to extract premolar teeth in a large number of cases. A well-developed plan is usually the result of meticulous planning and close monitoring of diagnostic data. Even then, treatment outcomes can be insufficient because of the consultant’s bias.

The deep learning system learns from multiple examples and the neural networks do an exceptional job at identifying patterns. So, AI programs accurately instruct about the need for orthodontic extractions (based on cephalometric data and model analysis). Clinical studies revealed that the accuracy of AI in predicting teeth extraction patterns was 93.9%.

The high accuracy of AI models means that AI can be used as a reliable tool. After careful analysis, AI guides you to provide the right treatment at the right time. Dentists perplexed about the decision of extraction, can take maximum benefit from deep learning AI models.

Moreover, it improves the collaboration between orthognathic surgeons and orthodontists. At times, camouflage treatment is paired with orthognathic surgery. Identifying which patient will get the maximum benefit from surgery is a complex decision. If any of the strategies is missed or underdone, there is a poor treatment outcome. According to a study, orthognathic surgery treatment plans offered by AI have a high accuracy (above 90%) and can significantly improve treatment outcomes.

Patient Monitoring

A large number of orthodontists and patients now prefer treating mild-to-moderate malocclusions with clear aligners. Patients are given different sets of aligners which bring about tooth movements. Such cases do not require extensive dental visits. Thus, AI-based telehealth software allows orthodontists to monitor treatment progress. The latest research suggests that AI enhances clinical performance and treatment outcomes.

Another study concluded that AI dental monitoring system improves patient-doctor interaction, improves patient compliance which eventually leads to enhanced treatment outcomes.

Final Word

Artificial intelligence improves orthodontic outcomes by quickly (within seconds) and accurately identifying cephalometric landmarks (for which clinicians take minutes to days). It accurately determines the skeletal age and provides the ideal treatment plan. AI also aids the dentist in deciding the extraction pattern and predicting orthognathic surgery. With optimal patient monitoring AI also improves treatment outcomes.

References

  1. Madani, S. M., Eslamipour, F., Sadeghian, S., & Tahani, B. (2023). General dentists’ awareness of orthodontic treatment needs of patients and their referral practices. Dental Research Journal, 20(1), 93.
  2. Harrell Jr, W. E. (2009, March). 3D diagnosis and treatment planning in orthodontics. In Seminars in Orthodontics (Vol. 15, No. 1, pp. 35-41). WB Saunders.
  3. Wang, C. W., Huang, C. T., Hsieh, M. C., Li, C. H., Chang, S. W., Li, W. C., … & Ibragimov, B. (2015). Evaluation and comparison of anatomical landmark detection methods for cephalometric x-ray images: a grand challenge. IEEE transactions on medical imaging, 34(9), 1890-1900.
  4. Park, J. H., Hwang, H. W., Moon, J. H., Yu, Y., Kim, H., Her, S. B., … & Lee, S. J. (2019). Automated identification of cephalometric landmarks: Part 1—Comparisons between the latest deep-learning methods YOLOV3 and SSD. The Angle Orthodontist, 89(6), 903-909.
  5. Bulatova, G., Kusnoto, B., Grace, V., Tsay, T. P., Avenetti, D. M., & Sanchez, F. J. C. (2021). Assessment of automatic cephalometric landmark identification using artificial intelligence. Orthodontics & Craniofacial Research, 24, 37-42.
  6. Ghowsi, A., Hatcher, D., Suh, H., Wile, D., Castro, W., Krueger, J., … & Oh, H. (2022). Automated landmark identification on cone-beam computed tomography: accuracy and reliability. The Angle Orthodontist, 92(5), 642-654.
  7. Jourieh, A., Khan, H., Mheissen, S., Assali, M., & Alam, M. K. (2021). The correlation between dental stages and skeletal maturity stages. BioMed Research International, 2021(1), 9986498.
  8. Schubert, A., Jäger, F., Maltha, J. C., & Bartzela, T. N. (2020). Age effect on orthodontic tooth movement rate and the composition of gingival crevicular fluid. J Orofac Orthop, 81(2), 113-125.
  9. Shekar, Suma; Chandrashekar, B R1. Is Age Related to Orthodontic Treatment?. D Y Patil Journal of Health Sciences 9(3):p 113-115, Jul–Sep 2021. | DOI: 10.4103/dypj.dypj_15_21
  10. Yu, H. J., Cho, S. R., Kim, M. J., Kim, W. H., Kim, J. W., & Choi, J. (2020). Automated skeletal classification with lateral cephalometry based on artificial intelligence. Journal of dental research, 99(3), 249-256.
  11. Kim, H., Kim, C. S., Lee, J. M., Lee, J. J., Lee, J., Kim, J. S., & Choi, S. H. (2023). Prediction of Fishman’s skeletal maturity indicators using artificial intelligence. Scientific reports, 13(1), 5870.
  12. Del Real, A., Del Real, O., Sardina, S., & Oyonarte, R. (2022). Use of automated artificial intelligence to predict the need for orthodontic extractions. Korean Journal of Orthodontics, 52(2), 102-111.
  13. Shin, W., Yeom, H. G., Lee, G. H., Yun, J. P., Jeong, S. H., Lee, J. H., … & Kim, B. C. (2021). Deep learning based prediction of necessity for orthognathic surgery of skeletal malocclusion using cephalogram in Korean individuals. BMC Oral Health, 21, 1-7.
  14. Thurzo, A., Kurilová, V., & Varga, I. (2021, December). Artificial intelligence in orthodontic smart application for treatment coaching and its impact on clinical performance of patients monitored with AI-TeleHealth system. In Healthcare (Vol. 9, No. 12, p. 1695). MDPI.
  15. Caruso, S., Caruso, S., Pellegrino, M., Skafi, R., Nota, A., & Tecco, S. (2021). A knowledge-based algorithm for automatic monitoring of orthodontic treatment: the dental monitoring system. Two cases. Sensors, 21(5), 1856.

You'll find more articles in my blog:

Read more