Field evaluation of the diagnostic performance of EasyScan GO: a digital malaria microscopy device based on machine-learning