Application Of Edge Detection Based On SPCNN In Ultrasonic Image Of Cholelithiasis

With economic development and improvement of people's living standards,the gallbladder incidence increase significantly and is second only to appendicitis in northwest China.Since the ultrasound image has some meaningful advantages,such as non-invasive monitoring,non-electromagnetic radiation,fast imaging etc,it has been playing a major role for clinical diagnosis of gallbladder disease.Moreover,the edge detection,which is a significant technique for image processing,provides more valuable edge information for physicians' work.This is because pixel gradient values of extracted edge regions have a great difference in ultrasound images.For edge detection,PCNN(Pulse-Couple Neural Networks)has a good performance compared with the state-of-the-art neural networks because of its capture behavior and synchronous dynamics in neuronal activity.What is more,the relation between image pixels and the network neurons is a one-to-one correspondence.SPCNN(Simplified Pulse-Coupled Neural Networks)retains most properties of basic PCNN model.It has been proved that the SPCNN has lower computational complexity and less parameters number when compared with other PCNN algorithms.So,based on SPCNN,the edge detection algorithm that is used to quickly extract edge information of ultrasound image assists effectively physicians' clinical diagnosis for gallbladder and gallstone.In this paper,we firstly introduce several main imaging techniques in medical research and illustrate basic principles of ultrasound imaging system,working approaches of ultra-sonographic examination and attractive features of ultrasound image.Based on the above-mentioned imaging knowledges,we analyze imaging characteristics of diagnostic ultrasound from gallbladder and gallstone,and explain morphology and anisotropic diffusion algorithms.Secondly,we elaborate basic knowledges of edge detection theory,the specific conception of edge and several prevalent edge detection algorithms including Canny,Sobel and Roberts.Meanwhile,we also analyze experimental indices of edge detection containing subjective and objective indices.Thirdly,we give a brief description and the parameter setting method of the PCNN models involving a basic PCNN model and the SPCNN model.Further,experimental data from subjective and objective evaluations are acquired by the edge detection algorithm based on SPCNN.Finally,the experimental results using the edge detection algorithm of SPCNN have a competitive performance for ultrasound images in gallbladder and gallstone.In contrast to other state-of-the-art algorithms,our algorithm obtains more precisely edge information,which proves validity and efficiency of our proposed edge detection algorithm of SPCNN.