Prognostic Evaluation Of Non-small Cell Lung Cancer Based On CT Radiomics

In all cancers,the incidence and mortality of lung cancer are in the first place,so it is called the "first killer" of cancer.Among them,non-small cell lung cancer(NSCLC)accounts for 80% to 85% of the total number of patients with lung cancer,including squamous cell carcinoma,adenocarcinoma and large cell carcinoma,which is the main pathological type of lung cancer.Non-small cell lung cancer(NSCLC)develops quickly,and the early symptoms are not obvious.Patients are often diagnosed in the middle and late stage,thus missing the best time for treatment.The five-year survival rate for patients with advanced non-small cell lung cancer is reported to be less than 15%.Due to the existence of tumor heterogeneity,the disease development trend of different patients with NSCLC is quite different,and tumors with similar phenotypes often show completely different sensitivities in radiotherapy and chemotherapy.How to develop more personalized treatment and reexamination plans according to the characteristics of patients with individual tumors has become the focus and difficulty of the prognosis of NSCLC.Radiomics is an emerging field in clinical medicine.It is the product of the fusion of clinical image diagnosis and big data information mining technology.It through the extract liquor from medical imaging to quantify the characteristics of a tumor,found it difficult to discern subtle differences in the same phenotype tumor to the naked eye,can effectively solve the problem of tumor heterogeneity is difficult to quantitative assessment,at the same time in methodology,the extracted the characteristics of the data by machine learning methods or statistics analysis method to establish the disease diagnosis and prognosis of tumor model,so as to assist doctors treatment in patients with tumor and review scheme planning,has the important value of clinical research and application.CT image has been widely used as an important modality in the study of pulmonary disease imaging due to its fast,high resolution and high contrast.Therefore,based on CT imaging,this paper studied the prognosis survival analysis of non-small cell lung cancer,mainly in the following two aspects:(1)In order to explore the prognostic and survival analysis model of NSCLC,the following work is carried out in this paper.Firstly,according to different tumor types,different algorithms are selected to segment the tumor area in CT image semi automatically to remove the interference of background area;secondly,the radiomics features of tumor area,including gray,shape,texture and medical signs,are extracted to quantify tumor information;secondly,the extracted feature data is optimized to achieve feature dimensionality reduction;thirdly,the machine is used Machine learning method is used to train the optimized feature data to obtain the prognostic survival analysis model and predict the range of the prognostic survival time of patients with non-small cell lung cancer.Finally,the survival analysis is carried out by using the cut-off survival time,the cut-off survival status and the predicted results to verify the experimental results.(2)In order to explore the prognostic factors of NSCLC,the following work was carried out.Firstly,according to different tumor types,different algorithms are selected to segment the tumor area in CT image semi automatically,and then the interference of background area is removed;secondly,the radiomics features of tumor area including gray level,shape,texture and medical signs are extracted to quantify the tumor information;secondly,the extracted feature data are processed Finally,the cut-off survival time and the cut-off survival state of the patients were correlated with the screened features and the predicted results respectively Analysis and survival analysis to verify the experimental results.The experimental results show that CT radiomics has great potential in evaluating the prognosis of NSCLC.It can quantify the tumor information,effectively assist doctors to evaluate the prognosis survival of NSCLC patients more accurately,and then help doctors to select and formulate treatment and reexamination programs for patients,so as to prolong the survival time of patients.