LEDAS is working on an ambitious project dedicated to next-generation orthodontic appliances for a major US orthodontic company. This research and development project requires a very high level of qualification in applied mathematics and 3D computational geometry algorithms, with a focus on processing 3D meshes produced from scans.
Thanks to this project, our developers have got a deep expertise in a number of third-party libraries for basic operations on 3D meshes.
This project is implemented as a Web browser application using our LEDAS Cloud Platform, which provides advanced visualization and navigation out-of-the-box. These are crucial for the doctor review process. Specific functions are implemented on top of LCP.
Kinematic joints is an area of expertise for LEDAS due to our know-how in geometric constraint solving, and is a vital part of planning bone implantations.
Based on our Cloud Platform, this browser application is focused on the treatment planning process. It provides a multi-user environment with access rights assignable to different user roles. User requests are dispatched between back-end servers for better scalability.
On the back-end side, the mold preparation processes are fully automated, thanks to efficient new methods and algorithms developed by our programmers for this project.
An algorithmic problem was solved by LEDAS for a US-based medical company, thanks to our expertise in optimization, performance tuning, and automatic alignment and mapping of similar-looking models (see LGC). In this case, 3D meshes had to be mapped to specific parts of 3D images, which are not explicitly identified. We developed a number of algorithms that focus on 3D image processing, global orientation, landmark detection, and energy calculation and reduction.
The computational library we implemented was developed on the basis of data structures on existing desktop software, and then was integrated into this client’s software. Testing on real data sets showed excellent results, even for cases with significant differences in the shapes of the 3D images and 3D polygonal meshes.