Font Size: a A A

Research On The Rapid OCT Imaging Model Of Dental Tissu

Posted on:2022-03-02Degree:MasterType:Thesis
Country:ChinaCandidate:X Y ZhouFull Text:PDF
GTID:2554307055454364Subject:Electronic and communication engineering
Abstract/Summary:PDF Full Text Request
Optical Coherence Tomography(OCT)is a new non-invasive three-dimensional medical imaging technology,and its applications in dentistry have attracted wide attention in recent years.OCT technology can be used for the detection and prevention of early dental caries,the detection of other oral diseases such as dental clefdoms,oral mucosal lesions,periodontitis,etc.,the guidance of tooth repair process,and the evaluation of the treatment effect of dental hard tissue diseases,etc.OCT signals and images of tooth tissues are important sources of diagnostic information for clinicians and researchers in hard tissue diseases.Deep learning has been proven to be effective in clinical decision making and information extraction.For supervised deep learning methods,large amounts of data with accurate labels are essential.However,it is impractical to obtain a large amount of accurately measured data in dental applications.The establishment of OCT imaging models of tooth tissues is an important way to generate OCT signals of different tooth tissues.At present,OCT imaging models for teeth with irregular structures are not universal,and the simulation speed is slow.Therefore,this paper proposes a fast OCT imaging model suitable for any tooth structure.The specific research contents are as follows:1.Based on Monte Carlo simulation,combined with the principle of time-domain OCT imaging and the optical characteristics of tooth tissue,this paper establishes an OCT imaging model suitable for smooth tooth structure;2.On the basis of the OCT imaging model of the tooth tissue,the importance sampling technology is adopted to improve the photon utilization rate,accelerate the convergence of the model,and reduce the simulation operation time;use the tetrahedral grid to segment the tooth tissue,refine the tissue structure,and realize the photon The precise positioning of the transmission path in the tooth tissue;use CUDA language to realize program parallel operation.The final improved model runs only1/10 the time required to simulate the original model.3.The OCT signal of human healthy enamel and dentin was simulated,and compared with the real OCT experimental results to verify the correctness of the established model,and to analyze the accuracy of the model and the influence of different photon numbers;the OCT signal of enamel caries was simulated,The effects of different degrees of enamel caries on OCT signals were analyzed;OCT images of enamel with different structures were simulated;and the simulation time of OCT imaging models before and after enamel and dentin improvement were compared and analyzed.
Keywords/Search Tags:Optical coherence tomography, Monte Carlo simulation, Tetrahedral mesh segmentation, GPU acceleration
PDF Full Text Request
Related items