Abstract—This study applied advanced data mining techniques for recurrent cervical cancer in survival analysis. The medical records and pathology were accessible by the Chung Shan Medical University Hospital Tumor Registry. Following a literature review, expert consultation, and collection of patients’ data, twelve variables studied included age, cell type, tumor grade, tumor size, pT, pStage, surgical margin involvement, LNM, Number of Fractions of Other RT, RT target Summary, Sequence of Locoregional Therapy and Systemic Therapy, LVSI. Two data mining approaches were considered where individuals are expected to experience repeated events, along with concomitant variables. After correcting for the four most important prognostic factors: pStage, Pathologic T, cell type and RT target Summary. Finally, clinical trials should randomize patients stratified by these prognostic factors, and precise assessment of recurrent status could improve outcome.
Index Terms—Recurrent event, cervical cancer, data mining technique.
Chi-Chang Chang is with the School of Medical Informatics, Chung Shan Medical University (e-mail: firstname.lastname@example.org).
Sun-Long Cheng is with the Chung Shan Medical University Hospital,Chi-Jie Lu is with the Department of Industrial Management at Chien Hsin University of Science and Technology.
Kuo-Hsiung Liao is with the Department of Information Management, Yuanpei University.
Cite:Chi-Chang Chang, Sun-Long Cheng, Chi-Jie Lu, and Kuo-Hsiung Liao, "Prediction of Recurrence in Patients with Cervical Cancer Using MARS and Classification," International Journal of Machine Learning and Computing vol. 3, no. 1, pp. 75-78, 2013.