2 days left

DAT264x: Identifying Malaria In Blood Imagery
Hosted By Microsoft

 

Identifying Malaria from Blood Smear Slide Images


Malaria is a blood disease caused by parasites transimitted through the bite of female Anopheles mosquitos. If not detected and treated early, severe illness can develop, often leading to death. Despite ongoing efforts to prevent the disease, an estimated 219 million cases of malaria occured worldwide in 2017 alone, according to a 2018 World Health Organization report. More than two thirds of all cases occur in children under 5.

Early detection not only reduces the disease and prevent deaths, it also helps to limit transmission. However, the diagnosis requires expert examination of thick and thin blood smear slide images. This manual process depends on smear quality and expertise in classifying and counting parasitized and uninfected cells. These limitations leave lots of room for error and present numerous difficulties for large-scale diagnoses.

In this challenge, you will use your AI tools to detect malaria in blood smear slide images.

Background

The malaria dataset contains giemsa-stained thin blood smear slides from 200 patients, 150 infected and 50 healthy. The images were acquired using a mobile application that runs on a standard Android smartphone attached to a conventional light microscope.The images were then manually annotated by an expert slide reader. Finally, a segmentation method was applied to detect and segment the red blood cells in the image.

blood_smear_infected
Infected cell
blood_smear_uninfected
Uninfected cell

As you can see, the task is no small feat!

Now, it's time to use your deep learning skills to parse out which cells are infected with malaria!