医工連携の成功のためには，医学のニーズを的確にとらえて工学のシーズ技術により問題解決を考え，プロト機を開発し何度も運用性を検討することが重要である．良い製品を生み出すには工学的シーズが必要となり，例えば，魚ロボット等の生物運動を応用したロボット技術，人工知能等の機械学習をベースとするIoT技術，さらには，ロボットとIoTの統合型システム技術が今後の医工連携の発展のキーとなる．

Purpose: This study aimed to verify the thickness detectability of our evaluation system using a dental scanner (D900) for the occlusal facets on sleep splints.

 Materials and Methods: First, the accuracy of the D900 was investigated via a multi sphere test. Surface roughness, diameter, and distances among centers in three SUJ2 stainless spheres were investigated. Then, the thickness detectability of our evaluation system was investigated using gauge blocks. The ceramic gauge blocks were located on a flat stone board, and then were scanned and superimposed on a flat-surface CAD model used as a reference plane. The distance between the scanned surface of each gauge block and the reference plane were calculated. Finally, to evaluate the performance of our system in a clinical setting, a 35-year-old female volunteer who was aware of sleep bruxism was recruited. The differences in vertical heights between images of two models (before and after applying the splint) were investigated.

 Result: The relative errors of diameter (0.154%) were significantly bigger than that of distance among three spheres (0.01%) because of powder coating. The D900 can produce high accuracy 3D scanning if powder coating is not required, completely distinguish the thickness difference of 10 μm of gauge block. As our evaluation system detects changes in the splint surface by comparing surface profile between two stone model before and after splint wearing, it is considered that the minimum detectable thickness of our system is approximately 50 μm, because it is affected by the surface roughness of both two stone model. Not only the wear facet on splint but also the deformation of splint was caused by excessive occlusal force after 14days applying. It was considered that the palate region was an appropriate region of interest (ROI) for registration to evaluate the whole deformation and fit of the splint. To evaluate the occlusal facets, the surface of the splint region was an appropriate ROI. We were able to detect the occlusal facets using our evaluation system.

 Conclusion: It was revealed that our system can detect differences of approximately 50 μm in thickness. It was suggested the appropriate ROI for registration should be selected depending on analytical parameters. Our system can detect the occlusal facets on the surface of splint after 14days applying.