Morphological Heterogeneity

The "Morphological Heterogeneity" (MH) dataset contains images, binary masks, bounding boxes, morphology labels, and other metrics of red blood cells (RBCs) flowing through a microfluidic device. RBCs are from seven different units of blood at 6, 7, and 8 weeks of storage. Brightfield-microscope images were acquired at a magnification of 64 X and every 5 s. We derived all other data from these images by applying an in-house image analysis pipeline. Overall, this dataset illustrates the heterogeneity among units of blood and their gradual decline towards mass spherocytosis. Obvious use cases are training & testing machine learning models and data analysis for further biological insights into RBC hypothermic storage.