MIXTURE OF EXPERTS AND MODEL APPROACH TO TUMOR CLASSIFICATION BASED ON PERFUSION MRI MEASURES IN LOCALLY ADVANCED RECTAL CANCER

摘要：

　　Mixture of Experts model(ME) provides a modular approach for supervised learning problem. This research studied the application of modified ME variants to repeated measure data of clinical decision problems which are known to be difficult to be dealt with. The purpose of this study is to categorize the disease severity of tumors in locally advanced rectal cancer through the application of a modified ME model. We applied the modified ME model to a dataset with 3- times perfusion MRI measures(before therapy, end of the 2nd week of therapy, before surgery) by Lim et al(2012). The proposed model is expected to provide a robust method for subject classification without user intervention and bias.

关键词： Mixture of Expert Model Linear Mixed Effect model Classification Clinical Decision Problem Magnetic Resonance Imaging

作者: Sungmin Myoung

作者单位: Department of Medical Information and Administration, Jungwon University, Goesan, Chungbuk,Korea

会议类型: 国际会议

会议名称: The 7th Asia Pacific Association for Medical Informatics Conference(第七届亚太医药信息学大会(APAMI2012))

会议地点: 北京

会议语种:英文

页码: 1-2

在线出版日期: 2012-10-24（万方平台首次上网日期，不代表论文的发表时间）