A Tutorial Introduction to STAPLE

Introduction STAPLE is an expectation-maximization algorithm for Simultaneous Truth And Performance Level Estimation. The algorithm considers a collection of segmentations and computes a probabilistic estimate of the hidden, implicit, true segmentation and a measure of the performance level achieved by each segmentation. The source of each segmentation in the collection may be an appropriately trained human rater or raters, or automated segmentation algorithms. The probabilistic estimate of the true segmentation is formed by estimating an optimal combination of the segmentations, weighting each segmentation depending upon the estimated performance level, together with a prior model that can account for the spatial distribution of structures and spatial homogeneity constraints. STAPLE is straightforward to apply to clinical imaging data, it readily enables assessment of the performance of an automated image segmentation algorithm, and enables direct comparison of human rater and algorithm performance. The probabilistic segmentation that is estimated is a type of consensus segmentation, which is often used as a reference standard segmentation for assessing performance of different algorithms and when an improved segmentation is needed.