Automated Dental Aesthetics with Machine Learning

ECU Author/Contributor (non-ECU co-authors, if there are any, appear on document)

Ashwinee Mehta (Creator)

Institution

East Carolina University (ECU )

Web Site: http://www.ecu.edu/lib/

Abstract: While dentures contribute an important role in restoring the dental and facial structure, facial aesthetics are an equally important consideration during the restoration of a patient with missing upper, lower, or both upper and lower teeth for elevating the treatment outcomes. Each denture is tailor-made for every patient, meaning the dental technician takes the impressions and measurements of the patient to make a perfect functional fitting denture. The current denture design workflow does not systematically include the aesthetic factors, patient's pre-treatment facial shape and in-progress denture design visualizations, instead relying on discussing mockups with the patients during appointments. This results into waiting for the final denture fitting on the patient to evaluate the final denture aesthetics. In this research, we plan to develop and validate some facial aesthetic proportion techniques that are used in the current denture design workflow. Given a frontal image of a person with missing teeth or a collapsed face, the proposed method will automatically generate an image of the patient, with teeth, while restoring the patient's face shape. This will assist the dental technicians to choose the best aesthetically fitting denture model, its size and position based on the current state of the face. Towards this goal, the method will automatically identify several facial landmarks, classify the patient's facial shape for easy selection of the denture model, and automatically create a three-dimensional denture design by using the patient's frontal and side-view images, as well as images captured from inside of the mouth. The goal of the research is to streamline the denture design process by considering the facial and teeth aesthetics, to enable denture-in-progress visualizations that avoid the end-moment denture refinements, with a simple graphical user interface that is easy to use by dental technicians.

Additional Information

Publication

Thesis

Language: English

Date: 2023

Subjects

Vertical Canon, One Thirds, Facial, Dental Reconstruction, Anthropometric Landmarks, Machine Learning, Similar, Non-collapsed Face, Face Recognition, Classification, Collapsed Face, Reconstruction

Email this document to