Surface Reconstruction And Occlusal Adjustment For Restoration In Dental CAD System

Although dental CAD systems have been widely used in clinic,number of interactive operations is always required in these systems to design restoration.So it is important to reduce the interactive operation and improve the efficiency of design restoration.The methods of surface reconstruction and shape optimization are studied in this thesis to reduce the interactive operation in restoration design.Combined with dental CAD,the research work includes standard tooth positioning,reconstruction surface detail of standard tooth,extraction of tooth biological characteristic curve and adaptive interference deformation.Standard tooth positioning is accomplished through complex interaction in most dental CAD.In this thesis,it will be regarded as a model deformation,which will facilitate to realize automatic standard tooth positioning.Firstly,occlusal information and anatomical characteristics are introduced into standard tooth library.Then the position of the paired tooth in repairing dentition is regarded as the target one of its same name tooth in standard tooth library.And the target coordinates of the anatomical feature points of this standard tooth in repairing dentition are obtained by using a radial basis function.The position algorithm of standard tooth is also studied based on the anatomical feature points and Laplacian deformation algorithm.Next,the surface details of standard tooth are required to reconstruct because standard tooth has more abundant and distinct surface detail compared with natural tooth.The details attenuation of standard tooth can be obtained by the change of Laplacian coordinate.So in this thesis,the improved Morse function is used to extract salient feature points on standard tooth.Then the attenuation details are obtained based on the salient points on standard tooth and the mapping points on natural tooth.The attenuation details are spread to the same name tooth of the lost tooth in standard tooth library based on the radial basis function.Furthermore the Laplacian deformation algorithm of attenuation detail is proposed to reconstruct the surface detail of standard tooth.The shape of reconstructed standard tooth is closer to the ideal restoration.Tooth biological feature curve is important constraint for standard tooth to generate initial restoration.Therefore,extraction accuracy directly determines the quality of the final restoration.Tooth preparation should be prepared before extracting,which is always manually grinded by dentist.It is easy to cause fuzzy regions.So the outermost feature line should be extracted to prevent food residues in this situation.The feature line extraction is transformed to an optimization path extraction.And a graph is established based on the extremality coefficients and mesh topology.Then the outward coefficient is proposed to extract the feature line on the fuzzy region.The curvature and outward coefficient are unified into the characteristic attribute to form the real cost function of the improved A* algorithm.The tabu table of ant colony algorithm is introduced into the improved A* algorithm to prevent the repeated extraction of feature lines,which can realize the extraction of the tooth biological feature curves.Interference deformation in occlusion adjustment is the key step for initial restoration to generate final one.At present,adaptive adjustment is provided in most commercial dental CAD,but it is easy to cause serious occlusion surface subsidence or shape loss.In order to reduce the post-adjustment problem,it is necessary to minimize the overall shape change of the restoration before and after occlusion adjustment,which is conductive to retain the pre-modeling work on restoration,and prevent the repeated adjustment for a certain area.In this thesis,a method of interference deformation is proposed,which makes the interference areas to lose a small amount of details and the occlusion surface to subside slightly.The interference information is obtained based on AABB(Axially-aligned bounding box)hierarchical bounding box.Then,through fixing some parameters in deformation,the problem is transformed into a constraint point selection problem,which is a generalized traveling salesman problem.The problem is solved by improving ant colony algorithm to realize minimum overall deformation for intervention adjustment.Finally,the dental CAD platform is also built to design single crown restoration with fixed denture.The experimental results show that the proposed method is efficient to design qualified restorations.