Constructing XGboost prediction model based on 3.0T diffusion kurtosis imaging improves the diagnostic performance for breast cancer
Han Zhou1, Wan Tang1,2, Tianhong Quan3, Xiaoyan Chen1, Huanian Zhang1, Zijie Fu1, Renhua Wu1, and Yan Lin1
1Radiology Department, Second Affiliated Hospital of Shantou University Medical College, Shantou, China, 2Institute of Health Monitoring,Inspection and Protection,Hubei Provincial Center for Disease Control and Prevention,Hubei Provincial Key Laboratory for Applied Toxicology, Wuhan, China, 3Shantou University，College Of Engineering, Shantou, China
This study demonstrated that MK derived from DKI was performed better than MD, ADC, Ve, Kep and Ktrans for differentiating between benign and malignant BLs. Also, MK has great potentialities in predict histological grades, lymph node status and Ki-67 expression of BCs. Finally, a XGboost model was constructed by combining MD, MK, age, shape and menstrual status, which exhibited superior diagnostic performance for BC characterization and an improved assessment of BLs. The findings of current study will aid the development of a novel noninvasive approach for BC screening and clinical diagnosis, therefore reducing unnecessary biopsies and patient`s anxiety.
This abstract and the presentation materials are available to members only;
a login is required.