Establishment And Validation Of Two-stage CT Prediction Model For Cervical Cancer

Objective In the course of radiotherapy for cervical cancer,some patients need to revise the target area and the limit of dangerous organs.The purpose of this study is to explore the influence parameters of re-planning?R?and to establish a multivariate prediction model and make a preliminary verification,so as to provide a theoretical basis for guiding the treatment of cervical cancer.Methods The clinical data of 132 cases of cervical cancer before and during radiotherapy?18 times?in our hospital were collected retrospectively,including:carcinoembryonic antigen?CEA?,carbohydrate antigen 125?CA-125?,squamous cell carcinoma?SCC?,tumor diameter,tumor volume,rectal volume,bladder volume,target area and dangerous organ dose?volume receiving 30-50gy,?V30,V40,V50?,maximum dose received by 1-2cc volumes?d1cc,d2cc??.Based on the principal component analysis,a multivariate model for forecasting the secondary plan was established,and the effectiveness of the model was verified by internal data.Results The prediction model of principal component was R=0.12P₁+0.21P₂+0.31 P₃?P₁ was the dose factor,mainly the?planned target volume average dose?,P₂ was the tumor factor,mainly the?initial tumor short path?,P₃ was the volume of 30Gy?V30?,mainly the?rectum V30?%???,when R?1?when the patient needed to make a re-planning,R was 1?,The model was verified by test set and test set,and its accuracy?ACC?is 0.826.Conclusion The model proposed in this paper can guide the implementation of staged radiotherapy plan for patients with cervical squamous cell carcinoma.The experimental results show that the model in this paper accords with the tumor changes measured by cone beam computed tomography?CBCT?,which saves more medical resources.At the same time,it proves that CEA,CA-125,SCC values before treatment are not closely related to the radiotherapy reprogramming of cervical squamous cell carcinoma...