Evaluation of an Automated Tooth Segmentation Tool

In dentistry, communication between dentists to plan treatment is very important. Originally, treatment planning involved communicating information by speaking or writing, often resulting in communication errors and misaligned expectations. The partner organization is involved in the development and provision of solutions for treatment planning and the development of software that can be used in dental specialties. Their ultimate goal is to allow dentists to communicate with each other through a web application that displays 3D results and can also produce treatment tools based on the created treatment plan.
The preparation of data for use in orthodontic dentistry is complex and has a high probability of error. It also takes time to prepare the data, including learning how to use the program. The developers, therefore, saw that if the complexity and time required to prepare the data could be reduced, it would be beneficial both for treatment and for learning. To address this need, the partner organization has developed an AI for tooth segmentation to work on complex parts and reduce the working time. The purpose of this research project is to evaluate the tooth segmentation algorithm through the use of 3D imaging collected at the School of Dentistry at the University of Alberta. The supervisors are experienced in developing similar ML algorithms for image segmentation tasks. They will supervise the task for a graduate student who will benefit from the learning experience in validating an external and commercially developed algorithm.

Faculty Supervisor:

Hollis Lai

Student:

Partner:

Tomorrow Smile Digital Healthcare

Discipline:

Life Sciences

Sector:

Professional, scientific and technical services

University:

University of Alberta

Program: