A new sampling methodology for creating rich, heterogeneous, subsets of samples for training image segmentation algorithms

Creating a dataset for training supervised machine learning algorithms can be a demanding task. This is especially true for medical image segmentation since this task usually requires one or more specialists for image annotation, and creating ground truth labels for just a single image can take up to several hours. In addition, it is paramount that the annotated samples represent well the different conditions that might affect the imaged tissue as well as possible changes in the image acquisition process. This can only be achieved by considering samples that are typical in the dataset as well as atypical, or even outlier, samples. We introduce a new sampling methodology for selecting relevant images from a larger non-annotated dataset in a way that evenly considers both prototypical as well as atypical samples. The methodology involves the generation of a uniform grid from a feature space representing the samples, which is then used for randomly drawing relevant images. The selected images provide a uniform cover of the original dataset, and thus define a heterogeneous set of images that can be annotated and used for training supervised segmentation algorithms. We provide a case example by creating a dataset containing a representative set of blood vessel microscopy images selected from a larger dataset containing thousands of images.