Study On The Algorithm For The Pulp Cavity Image By CBCT

The up-to-date medical research found that the pulp cavity deposition can be used to predict the age of human in forensic medicine.Based on the current popular oral cone beam CT imaging technology,how to accurately realize the three-dimensional image segmentation of pulp cavity is the prerequisite for the application of the conjecture method.There are many difficulties in the accurate segmentation of the CBCT dental pulp cavity due to influence of the noise interference,the blurring of tooth boundaries,similar gray value between the tooth and the alveolar bone and other factors.The Pulse Coupled Neural Network has a biological background to extract effective information from complex background,with synchronization pulse release and global coupling and other characteristics,the signal form and processing mechanism more in line with physiological basis of the human visual nervous system.In this paper,an improved PCNN model based on the deep research of PCNN theory and application is proposed to achieve the accurate segmentation of three-dimensional fault sequence image of CBCT pulp cavity.The main work and innovation of the thesis are as follows:(1)In order to solve the problem that the traditional PCNN model is too complicated,a lot of artificial parameters are set and the threshold attenuation is unstable,this paper presents an improved PCNN model by adjusting the PCNN to accept part of the network structure and ensuring its biological characteristics,which optimizes the external input of neuron,the weight of connected input L and the mode of threshold attenuation.The experimental results show that the proposed model can effectively reduce the complexity of the algorithm and improve the description ability of pixel spatial information.(2)Aiming at the problem that the ambiguity of tooth image leads to the difficulty of determining the number of PCNN model iteration and the need for manual setting,this paper gives an optimal criterion for iterative times based on minimum cross entropy by combining the spatial information of image pixels,analyzing the influence of iterative times on PCNN model segmentation algorithm and using information entropy optimization criterion.The optimal criterion realizes the accurate segmentation of CBCT tooth cavity image by improved PCNN model segmentation algorithm.