Using a computational approach known as deep learning, scientists have developed a new system to classify the shapes of red blood cells in a patient’s blood. The findings, published in PLOS Computational Biology, could potentially help doctors monitor people with sickle cell disease.