|M.Sc Student||Halperin Gustavo|
|Subject||Segmentation of Orthodontic Models and Recognition of|
Free Form Features
|Department||Department of Mechanical Engineering||Supervisor||Professor Anath Fischer|
One of the main challenges in dentistry, especially in orthodontics, is to improve the processes of measuring and analyzing a patient's dentition.
Currently, the analyzing process starts with intra-oral measurements of both the upper and the lower teeth using elastomeric materials, which are unpleasant and uncomfortable for the patients especially for children. This mold only serves as the basis for making another plaster mold, known as the dental cast. The casts that represent the dental relationship are trimmed and then analyzed manually from several aspects.
This process is time-consuming and may be accompanied by accumulated errors. Advanced technology that can directly take oral measurements of the patient's dentition by 3D scanning and then process it, without the need for both an intermediate mold and manual cast analysis, would make the entire process significantly more efficient and accurate.
The new measurement technique should be quick, more comfortable and less expensive. In cases where measurements and analysis are needed, they can be performed in real time using a computerized system for dentists. Another advantage to such a system is that a number of specialists will be able to consult via the Internet in real time, as the system will support distance visualization.
For more complex processing and production (i.e., in the case of crowns and implants), the information can be sent via the Internet for VC consulting or to a specialist lab.
In this work, algorithms for reconstruction and feature recognition were developed as part of this future dental system. The reconstruction algorithm is applied on the sampled teeth and creates a 3D medical model in the form of a mesh. The recognition algorithm is applied on the resulting medical model, where teeth and jaws are segmented and recognized. For efficiency, the data is filtered and reduced as part of a pre-process.
In this work the feasibility of the algorithm was tested and demonstrated on plaster molds of mouth cavities. A performance analysis was applied on these algorithms to show the potential of the automatic dental system.