Method for determining the reliable prediction(s) of compositions of tissue phantoms

In a previous paper, we showed that mixture laws can be used to predict the compositions of tissue phantoms. Due to the variation in the performance of the mixture laws when they are applied to phantoms, however, it is difficult to determine which predictions of the composition are most reliable. In this paper, we study the causes of the variation in the performances of different sets of mixture laws, and propose a criterion for choosing a reliable predictor. The predictions selected with the proposed criterion agree very well with the known compositions of phantoms. The potential uses of the mixture methodology and the criterion to tissue characterization are discussed.