Erythrocyte morphology analysis is a key step in the diagnosis flowchart of anemia. It is performed on a peripheral blood smear after May Grümwald Giemsa staining. In the context of hemolytic anemias for example, it allows the recognition of therapeutic emergencies such as sickle cell disease crisis, malaria-induced hemolysis and thrombotic microangiopathy, the latter being characterized by the presence of schistocytes and justifying an immediate clinical care. However, cytological analysis of erythrocyte morphology requires pre-analytical interventions (smear spreading + staining), the quality of which determines the accuracy of the result. Moreover, it requires a good cytological expertise and may be sometimes subjective. For several years, alternative methods for erythrocyte morphology evaluation have been developed, based on automated hematology machines or automated microscopy. Nevertheless, none of them has yet proven itself in comparison with cytology, especially in the diagnosis of thrombotic microangiopathies. By combining the advantages of flow cytometry and microscopy, flow imaging appears to be a promising technology for the diagnosis of anemias: it does not require any pre-analytical intervention, does not require any spreading and analyzes a large number of events. Moreover, it can be coupled with artificial intelligence via the generation of an apprenticeship by the constitution of a large image data base, which then allows the recognition of the different red blood cells morphologies without human eyes. The objective of this study is to build a data base containing the main red blood cell morphologies relevant in anemia, and to validate it through a comparison in anemic patients of erythrocyte morphological assessment either directly on whole blood by flow imaging or routinely by cytological analysis of peripheral blood smear after by a trained operator.