Machine-learning training and validation datasets

<b>Contains datasets used in "Computational cytomorphology reveals cellular archetypes in whole blood slides".</b><br><i>datasets.zip</i><br>ZIP file. Contains segmentation datasets for white blood cells in whole blood slides (blood smears) collected in three different settings:<br>- segmentation_dataset_AC1 - images from Addenbrooke's Hospital digitalized using a Hamamatsu Nanozoomer 2.0- segmentation_dataset_AC2 - images from Addenbrooke's Hospital digitalized using an Aperio AT2- segmentation_dataset_MLLC - images from the Munich Leukaemia Laboratory digitalized using a Hamatsu Nanozoomer<br><i>quality-datasets.zip</i><i><br></i>ZIP file. Contains two folders (train and test) with images annotated as being of good (1) or poor (0) quality. Retrieved from the MLL.<br><i>rbc-filtering-dataset.h5</i><i><br></i>Contains an HDF5 file with morphologically characterised objects annotated as ""RBC" (1) and "not RBC" (0) from images retrieved from whole blood slides as well as features characterising them. Divided into two groups "train" (where the train images are) and "test" (where the testing images).