TY - GEN
T1 - SVM in Classification of stage 0~II and III~IV with Breast Cancer
T2 - 2023 International Conference on Computer, Vision and Intelligent Technology, ICCVIT 2023
AU - Guo, Yeang
AU - Tao, Tan
AU - Ronglin, Ronglin
AU - Xiao, Liangfen
AU - Ding, Lijuan
AU - Li, Qing
AU - Xie, Hui
N1 - Publisher Copyright:
© 2023 ACM.
PY - 2023/8/25
Y1 - 2023/8/25
N2 - Objective: The objective is to develop a predictive model utilizing Support Vector Machines (SVM) for the purpose of classifying the clinical stage of breast cancer. Materials and Methods: Accurate determination of the clinical stage of breast cancer patients holds significant importance in selecting suitable treatment options and minimizing avoidable complications. In this study, we present the application of radiomics and SVM for breast cancer computed tomography (CT) to anticipate the preoperative clinical stage in breast cancer patients. The training dataset encompasses 166 cases obtained from the Affiliated Hospital of Xiangnan University, while the test dataset comprises 91 cases from Chenzhou Third People's Hospital. The integration of clinical parameters with radiomics exhibits the most superior diagnostic efficacy in forecasting the clinical stage of breast cancer. As part of the evaluation, various metrics were calculated, including the area under the curve (AUC), the accuracy (ACC), sensitivity (Sen), specificity (Spe), positive predictive value (PPV) and negative predictive value (NPV). To differentiate between the radiomics model, clinical data model, and fusion model, the Delong test was utilized. The precision of the prediction model was evaluated by generating a calibration curve using 1,000 bootstrap weight samples. Furthermore, the decision curve analysis (DCA) was conducted to assess the model's practicality. Results: The fusion model exhibits superior predictive performance compared to both the single radiomics model and clinical model. The fusion model's test sets of AUC, ACC, Sen, Spe, PPV, and NPV are 0.824, 0.780, 0.932, 0.652, 0.707, and 0.909, respectively. Conclusion: The fusion model exhibits greater efficacy than both the single radiomics model and clinical model, and thus holds significant potential for facilitating the diagnosis of breast cancer stage and the development of individualized treatment plans.CCS
AB - Objective: The objective is to develop a predictive model utilizing Support Vector Machines (SVM) for the purpose of classifying the clinical stage of breast cancer. Materials and Methods: Accurate determination of the clinical stage of breast cancer patients holds significant importance in selecting suitable treatment options and minimizing avoidable complications. In this study, we present the application of radiomics and SVM for breast cancer computed tomography (CT) to anticipate the preoperative clinical stage in breast cancer patients. The training dataset encompasses 166 cases obtained from the Affiliated Hospital of Xiangnan University, while the test dataset comprises 91 cases from Chenzhou Third People's Hospital. The integration of clinical parameters with radiomics exhibits the most superior diagnostic efficacy in forecasting the clinical stage of breast cancer. As part of the evaluation, various metrics were calculated, including the area under the curve (AUC), the accuracy (ACC), sensitivity (Sen), specificity (Spe), positive predictive value (PPV) and negative predictive value (NPV). To differentiate between the radiomics model, clinical data model, and fusion model, the Delong test was utilized. The precision of the prediction model was evaluated by generating a calibration curve using 1,000 bootstrap weight samples. Furthermore, the decision curve analysis (DCA) was conducted to assess the model's practicality. Results: The fusion model exhibits superior predictive performance compared to both the single radiomics model and clinical model. The fusion model's test sets of AUC, ACC, Sen, Spe, PPV, and NPV are 0.824, 0.780, 0.932, 0.652, 0.707, and 0.909, respectively. Conclusion: The fusion model exhibits greater efficacy than both the single radiomics model and clinical model, and thus holds significant potential for facilitating the diagnosis of breast cancer stage and the development of individualized treatment plans.CCS
KW - Breast Cancer
KW - Classification
KW - Clinical Staging
KW - data mining
KW - healthcare data
KW - healthcare information technology
KW - Omics
UR - http://www.scopus.com/inward/record.url?scp=85181401069&partnerID=8YFLogxK
U2 - 10.1145/3627341.3630410
DO - 10.1145/3627341.3630410
M3 - Conference contribution
AN - SCOPUS:85181401069
T3 - ACM International Conference Proceeding Series
BT - Proceedings of 2023 International Conference on Computer, Vision and Intelligent Technology, ICCVIT 2023
PB - Association for Computing Machinery
Y2 - 25 August 2023 through 28 August 2023
ER -
