Predicting Breast Cancer Responsiveness To Hormone Treatment Using Quantitative Textural Analysis

摘要

Biomarker signatures for predicting breast cancer tumor aggressiveness. The signatures are derived from QTA-based parameters from a first population of low risk scores and a second population of high risk scores. The signatures may be expressed in the form log (RS)=Mx+B for linear modeling, or for logistic modeling the signatures may be expressed as either p=ex/[1+ex] where x=Ay+B, or in the form log it(p)=C(PR)+Ay+B, where y is a QTA based parameter.