Evaluating Plasmodium falciparum automatic detection and parasitemia estimation: A comparative study on thin blood smear images

Abstract

Malaria is a deadly disease that is transmitted through mosquito bites. Microscopists use a microscope to examine thin blood smears at high magnification (1000x) to identify parasites in red blood cells (RBCs). Estimating parasitemia is essential in determining the severity of the Plasmodium falciparum infection and guiding treatment. However, this process is time-consuming, labor-intensive, and subject to variation, which can directly affect patient outcomes. In this retrospective study, we compared three methods for measuring parasitemia from a collection of anonymized thin blood smears of patients with Plasmodium falciparum obtained from the Clinical Department of Parasitology-Mycology, National Reference Center (NRC) for Malaria in Paris, France. We first analyzed the impact of the number of field images on parasitemia count using our framework, MALARIS, which features a top-classifier convolutional neural network (CNN). Additionally, we studied the variation between different microscopists using two manual techniques to demonstrate the need for a reliable and reproducible automated system. Finally, we included thin blood smear images from an additional 102 patients to compare the performance and correlation of our system with manual microscopy and flow cytometry. Our results showed strong correlations between the three methods, with a coefficient of determination between 0.87 and 0.92.

Fig 1. Image of a Plasmodium falciparum-infected thin blood smear at 1000x magnification, both without and with the Miller reticle.

At 1000x magnification, one graduation = 1 micrometer. The average size of RBCs is 7μm (ranging from 5 to 10).

Fig 2

(a) Comparison of the relative standard deviation (RSD) among microscopists, ordered by increasing parasitemia, between two manual techniques, including 10 measurements randomly selected from 10 microscopists, resulted in a mean RSD of 26.40% for the Miller reticle method and 38.22% for the standard measurement method. (b) Distribution of RSD among fields, including the MALARIS estimation on 10 field images. Above each histogram, the measured parasitemia value for each patient, arranged in ascending order of parasitemia, is displayed.

Fig 3. Repeatability of the parasitemia automatic count by the MALARIS system was assessed on 14 patients with estimated parasitemia from 0.7% to 12.72% and a median value of 2.01%.

The relative standard error rate (RSE) (%) was represented to evaluate the repeatability of k combinations (1 to 5) using 10 thin blood smear images (a). Additionally, a comparison was made with the Miller reticle and Standard manual count (b). The yellow triangles represent the mean value of each pair of measurements.

Fig 4. Comparison of regression plots for parasitemia estimation.

MALARIS (x-axis) vs. Miller reticle (y-axis) (a), MALARIS (x-axis) vs. flow cytometry (y-axis) (b), and Miller reticle (x-axis) vs. flow cytometry (y-axis) (c). Below the plots, there is a specific focus on the values of parasitemia between 0 and 5%.