PlasmoCount: Automated detection and staging of malaria parasites from cytological smears using convolutional neural networks
Description
PlasmoCount is an online tool for automated detection of infected red blood cells by Plasmodium (malaria) parasites. The tool was built using a machine learning approach (convolutional neural networks) developed from training data sets of parasite infected smears (Giemsa-smears) and then subsequently tested on blind data sets to validate accuracy and reproducibility. The data sets behind the work (first described in the preprint of our paper on MedRxiv - https://doi.org/10.1101/2021.01.26.21250284) are supplied here to enable others to build improved models or alternative strategies to automated cell detection. More information about the data sets or model can be gained from contacting the Baum laboratory.
Files
Steps to reproduce
Please see the first publication of the work as a preprint on MedRxiv (https://doi.org/10.1101/2021.01.26.21250284) for a full description of methods (protocols etc) and results for reproduction.