Statistical Analysis of Hematological Parameters for Prediction of Sickle Cell Disease

About 30 million people worldwide are affected by the monogenic recessive -globin gene abnormality known as sickle cell disease (SCD), which is a significant public health issue. From asymptomatic to severely symptomatic illnesses that might cause patient mortality, pathological features range. The most common presenting symptom of SCD is vasooclussive crisis (VOC). The red cell membrane of the Sickle Red Blood Cells (SRBCs) is damaged by repeated cycles of sickling and desickling processes caused by the formation and aggregation of HbS (sickle hemoglobin) polymers. Cellular dehydration (reduction of ion and water content), increased viscosity (red cell density) and a transient increase in intracellular calcium are all associated with HbS polymerization. As a result, SRBCs become adhesive and inflexible (rigid), resulting in premature destruction. The decreased life span of SRBCs causes chronic hemolytic anemia, and capillary blockage causes tissue hypoxia and subsequent organ damage. So, it is important to monitor patients suffering from sickle cells. Here we have used machine learning to visualize those patients and categorize them according to their hemoglobin level, percentage of reticulocyte count and serum Lactate dehydrogenase (LDH) level which is regarded as a marker of hemolysis. In this article we propose a framework which uses the statistical analysis using Linear Regression technique on a sickle cell patients dataset showing how hemoglobin is depleted in a body by the use of two parameters called LDH and Retics.