The 53
rd
Annual Scientific Congress of the Korean Association of Orthodontists
PASSION FOR A LONG JOURNEY :
Nature vs Nurture
Online
November 5(Thu) - 6(Fri), 2020
Precongress: November 1(Sun) / VOD: November 7(Sat) - 8(Sun)
2020 Young Investigator Award
Automated tooth segmentation of 3D scan data using generative adversarial networks
Dr. Yoon-Ji Kim
Assistant professor
Department of Orthodontics,
University of Ulsan College of Medicine
Asan Medical Center, Seoul, Korea
Digital setup provides the orthodontist with insights into the possibilities and limitations of treatment. Orthodontic diagnosis and anchorage preparation can be planned by aligning the teeth in the virtual space. The first step of the digital setup is the tooth segmentation of the patient¡¯s intraoral scan data. Although full-arch scan data that have been acquired using intraoral scanners have high dimensional accuracy, there are limitations. Intraoral scanners use a stereo-vision system, which has difficulties in scanning narrow interdental spaces. These areas, which lack accurate scan data, are called areas of occlusion. Owing to such occlusions, intraoral scanners often fail to acquire data in the interdental areas. This presentation proposes a method of reconstructing occluded areas and automatically segment the teeth from the intraoral scan data using a generative adversarial network.