VNG's narrative categories are more predictive of segmentation choices in drawn visual narratives than semantic situational changes (Cohn & Bender, 2017), and P600s also differ between sequences in which inferential situational change is held constant but narrative structure differs (Cohn & Kutas, 2015).
In this context, a radiation oncologist (alone or as part of a multidisciplinary team) reviews the subjective decisions made by the prescribing oncologist when implementing technical aspects of care, including volume segmentation or contouring, choice of prescribed dose and fractionation, selection of regional anatomy covered and dosimetry optimisation.
If this hypothesis holds true, then the segmentation algorithm of choice would be the one that is most reproducible.
However, given that it is time-consuming and less reproducible than 40P and FLAB, we do not consider it the segmentation algorithm of choice.
Special attention had to be placed in regions where vessels were close together, since a wrong choice of segmentation parameters could lead to over-segmentation of the space between them.
In doing so, several important technical details such as selection of discriminating features to extract, sensitivity analysis of data segmentation window size, and choice of the classifier to be trained are investigated.
The fast convergence and high breakdown point of HBM make this estimator an outstanding choice for segmentation of multi-structured data.
Example sample size calculations demonstrate a rather large effect of the choice of segmentation method on the required number of subjects.
The shapes of polygons output by multi-resolution segmentation change with the choice of geometric signatures and their weights.
Such an evaluation criterion can be useful for different applications: the comparison of segmentation results, the automatic choice of the best fitted parameters of a segmentation method for a given image, or the definition of new segmentation methods by optimization.
