Red blood cell and white blood cell classification using double thresholding and BLOB analysis

Abstract

Malaria caused by the invasion of parasite into human blood through the bite of infected certain mosquito. In order to detect malaria parasite in human blood, first important thing to do is to classify the red blood cell and white blood cell. So the detection method will not misclassed the white blood cell as the infected red blood cell. In this study, we propose a combination method of Double Thresholding and Binary Large Objecy (BLOB) analysis to classify white blood cell and red blood cell using data from Department of Health of Central Borneo Province. We use Recursive Grass-Fire Algorithm in our BLOB analysis to find large objects and it centroid. We also use Annular Ring Ratio (ARR) to compute the cell area. From the cell area size information, we can classify the cell into red blood cell or white blood cell. Our study obtain Positive Predictive Value (PPV) and Sensitivity value: 84.43% and 85.5% respectively.