Mobile-Aware Deep Learning Algorithms for Malaria Parasites and White Blood Cells Localization in Thick Blood Smears

Effective determination of malaria parasitemia is paramount in aiding clinicians to accurately estimate the severity and guide response for quality treatment. Microscopy by thick smear blood films conventional method determination. Despite its edge over other existing methods determination, it has been critiqued being laborious, time consuming equally requires expert knowledge an efficient manual quantification parasitemia. This pauses a big challenge most low developing countries as they are not only highly endemic but resourced terms technical personnel medical laboratories study presents end-to-end deep learning approach automate localization count P.falciparum parasites White Blood Cells (WBCs) effective The involved building computer vision models on dataset annotated images. These were built based pre-trained including Faster Regional Convolutional Neural Network (Faster R-CNN) Single Shot Multibox Detector (SSD) that help process obtained digital To improve model performance due limited dataset, data augmentation was applied. Results from evaluation our showed reliably detected returned WBCs with good precision recall. A strong correlation observed between model-generated counts done microscopy experts (posting spear man ρ = 0.998 0.987 WBCs). Additionally, proposed SSD quantized deployed mobile smartphone-based inference app detect situ. Our can be applied support diagnostics settings few trained Experts yet constrained large volume patients diagnose.