Abstract:
Malaria is an infectious disease caused by a peripheral blood parasite of the genus Plasmodium. Conventional microscopy, which is currently the gold standard for malaria diagnosis has occasionally proved inefficient since it is time consuming, tiresome and subjective. To support this manual method, in this research work, an accurate, rapid and affordable model of malaria diagnosis using Giemsa stained thin blood smear images was developed. The method made use of morphological and color features of healthy and infected red blood cells (RBCs). Images of healthy and infected RBCs were acquired, pre-processed, relevant features extracted from them and eventually diagnosis was made based on the features extracted from the images. A set of features based on morphology and colors have been used and the performances of these features on the RBCs have been evaluated using an artificial neural network (ANN) classifier. The accuracy of classification using morphology and color features were 93.3% and 100%, respectively. The classification performance of the network was best for color features. Another notable contribution made in this work was the development of techniques for counting RBCs and parasitemia estimation. The results of the algorithm were able to detect and count plasmodium with an accuracy of 98%, sensitivity, 88.6% and specificity, 98.6%.
