Research And Analysis Of Automatic Sketching Technology Based On Eclipse Planning System

In recent years,the incidence of cancer has become higher and higher,and radiation therapy is an important part of cancer treatment.The patient's target area and endangered organ delineation are the most important part of the radiotherapy process,and also a part of the most time and effort.Therefore,the target area of the radiotherapy planning system and the automatic delineation function of the endangered organ can effectively serve the clinical.At present,there are many automatic planning functions of the planning system on the market,but there are few studies on the automatic delineation of this function,and there are few clinical applications.At present,it is rarely used in hospitals to automatically sketch this function.Is it unclear whether this function is really suitable for clinical use,is it really for clinical services?This paper studies whether the automatic delineation function of the radiotherapy planning system is suitable for clinical problems.The Eclipse planning system,which is the most widely used on the market,is selected as the research object,and its automatic sketching function is researched and analyzed.The dangerous parts of the four parts of the human body are automatically outlined by the Eclipse automatic delineation function.The results of the automatic delineation are compared with the results of the manual delineation,and the delineation effect of the organs in various parts of the human body is obtained,thereby obtaining the automatic Eclipse planning system.Delineate the clinical value of the function.Main tasks as follows:(1)Analyze and study the automatic sketching function of the Eclipse planning system.The automatic sketching function is divided into two parts.Part of it is done directly by image segmentation,and the other part is done by image registration.Image registration includes rigid registration and non-rigid registration of the image.The image segmentation algorithm is mainly flood filling method and K-NN algorithm.The image rigid registration algorithm is mainly bilinear interpolation,mutual information and normalization correlation coefficient,gradient descent method,downhill simplex method and so on.The main algorithm for non-rigid registration is the Demons algorithm.(2)Select 30 patients with head and neck,chest,abdomen and pelvis.One is to let the three-level doctors in different parts manually delineate the organs,and the other is to automatically delineate the organs by calling the template and image segmentation through image registration.Then compare the two methods of automatically delineating the endangered organ and manual delineation.The effect of automatic sketching was evaluated by Dice Similarity Coefficient(DSC)and Overlap index(OI).(3)The experimental results are obtained.The results show that the coincidence degree between automatic and manual delineation of left and right optic-nerve,optic chiasm,left and right crystals,right and left kidney,liver,bladder,left and right femoral head,rectum and pelvis are poor.It needs to be manually modified and adjusted before they can be used in clinical practice.The coincidence degree between automatic and manual delineation of left and right eye,brain stem right and left parotid and heart are good.It can be used for clinical purposes with a little modification.Only the automatic delineation and manual delineation of the lungs and spinal cord are almost identical to meet clinical needs.