Paper Title
Sickle Cell Diagnosis from Peripheral Blood Smear Images Using Open-Source Image Processing Techniques

Abstract
Sickle cell disease is a genetic disorder that affects millions worldwide. Its diagnosis often relies on manual examination of blood smear images by experts. This traditional method can be time-consuming and resource-intensive. In this study, we developed an image analysis method using Fiji, an open-source software, to classify red blood cells from peripheral blood smear images into normal, sickle and other cell types. The classification was based on key morphological features, including circularity, aspect ratio and roundness. Our findings showed that aspect ratio and roundness were the most reliable metrics for differentiating elongated sickle cells from rounder normal cells, achieving an accuracy of 76% with balanced sensitivity (0.73) and specificity (0.77). Circularity, however, was less effective, resulting in a high false positive rate. This open-source method offers a cost-effective solution for sickle cell detection, particularly in resource-limited regions. Further improvement with advanced segmentation techniques and machine learning approach could enhance the method’s accuracy, making it practical for clinical diagnostics. Keywords - Image Analysis, Morphological Features, Peripheral Blood Smear, Sickle Cell Anemia