An Efficient Method for Detection of Outliers in Tracer Curves Derived from Dynamic Contrast-Enhanced Imaging

摘要: Presence of outliers in tracer concentration-time curves derived from dynamic contrast-enhanced imaging can adversely affect the analysis by model-fitting. A computationally efficient method for detecting is presented this study. The proposed based on a piecewise linear model and implemented using robust clustering algorithm. noniterative all parameters are automatically estimated. To compare with existing Gaussian regression-based methods, simulation studies were performed simulating generalized Tofts kinetic different tissue types. Results show that achieve better detection performance than method. Compared method, similar much faster computation speed.