Automated Protocol for Counting Malaria Parasites (P. falciparum) from Digital Microscopic Image Based on L*a*b* Colour Model and K-Means Clustering

摘要

Basis for malaria parasites diagnosis in most hospitals and clinics, especially in developing countries, which is manually done, is strenuous and time-consuming. In this paper, we present an automated protocol for counting malaria parasites (P. falciparum) from digital microscopic red blood cells (RBCs) mages based on L*a*b* colour model and K-Means clustering algorithm using Matlab. This method is device-independent, perceptually uniform and approximates human vision. An image slide of size 300 x 300 x 3 pixels of RBCs with malaria parasites has been counted in less than 10 seconds using a computer with 64-bit Intel (R) Celeron (R) Central Processing Unit and processing speed of 2.20 GHz. The digital counts have a good correlation with the manual counts. This automated protocol has the potential of providing fast, accurate and objective detection information for proper clinical management of patients.