Development of a High-Throughput Microscope for the Analysis of Peripheral Blood Smears for Anemia Screening

Abstract

The conventional method of screening for anemia requires pathologists to manually examine slides via microscope, a tedious process during health emergencies. This study presents an automated high-throughput optical digital microscope system capable of sequentially scanning and analyzing 10 blood smear slides per batch in under 15 min using a Laplacian-based autofocusing algorithm at 40x magnification. The acquired images are segmented via the YOLOv5 algorithm, and morphological features of red blood cells (RBCs) are classified using a multilayer perceptron (MLP) model. The system achieved 90.6% accuracy, 95% precision, 91% sensitivity, and 94% specificity in classifying anemia subtypes (macrocytic, microcytic, normocytic) and healthy samples. The trained model is integrated into an Android application for real-time geographic mapping of anemic clusters, enabling healthcare workers to prioritize interventions efficiently. This high-throughput approach eliminates the need for immersion oil and manual slide handling, demonstrating significant potential for rapid, scalable anemia screening in resource-limited settings.