Orthodontic Implementation of Machine Learning Algorithms for Predicting Some Linear Dental Arch Measurements and Preventing Anterior Segment Malocclusion: A Prospective Study
<i>Background and Objectives</i>: Orthodontics is a field that has seen significant advancements in recent years, with technology playing a crucial role in improving diagnosis and treatment planning. The study aimed to implement artificial intelligence to predict the arch width as a prev...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-11-01
|
| Series: | Medicina |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1648-9144/59/11/1973 |
| Summary: | <i>Background and Objectives</i>: Orthodontics is a field that has seen significant advancements in recent years, with technology playing a crucial role in improving diagnosis and treatment planning. The study aimed to implement artificial intelligence to predict the arch width as a preventive measure to avoid future crowding in growing patients or even in adult patients seeking orthodontic treatment as a tool for orthodontic diagnosis. <i>Materials and Methods</i>: Four hundred and fifty intraoral scan (IOS) images were selected from orthodontic patients seeking treatment in private orthodontic centers. Real inter-canine, inter-premolar, and inter-molar widths were measured digitally. Two of the main machine learning models were used: the Python programming language and machine learning algorithms, implementing the data on k-nearest neighbor and linear regression. <i>Results</i>: After the dataset had been implemented on the two ML algorithms, linear regression and k-nearest neighbor, the evaluation metric shows that KNN gives better prediction accuracy than LR does. The resulting accuracy was around 99%. <i>Conclusions</i>: it is possible to leverage machine learning to enhance orthodontic diagnosis and treatment planning by predicting linear dental arch measurements and preventing anterior segment malocclusion. |
|---|---|
| ISSN: | 1010-660X 1648-9144 |