Conference Agenda

Giemsa-stained blood smear microscopy is the gold standard for detecting malaria parasites, but it is time-consuming and limited for storage and reference. To address this, we developed PlasmoCount, a deep learning tool for accurate, automated counting of intracellular parasites and digital archiving support. Principally, we have achieved a substantial reduction in PlasmoCount’s processing time allowing for evaluation of a single image in under 3 seconds (reduced from 40). In addition, we have updated the tool so that it can now detect blood-stage infections from multiple species of human-infective and experimental rodent-infective Plasmodium parasites. Combined with a suite of other updates, including advanced cell differentiation and use at different magnifications, these augmentations broaden the distribution of input data our model can accommodate and radically advance its speed whilst maintaining its high classification accuracy (99.8%). Finally, we provide an offline, on-device version of the standardised framework designed for smartphones, including iOS and Android operating systems. By making use of imported images or image capture via a smartphone camera, PlasmoCount 2.0 markedly improves malaria parasite smear-based detection and provides a reproducible means to assess parasite infections either in routine laboratory work or as a future aid in clinical or field diagnosis.