Published Paper Details:

A main aim of this research work is to develop and validate an automatic tool for parasite detection in thick blood films. Malaria is still a major health problem but the diagnosis has not improved and it is at the level of technicians due to the lack of Medical resources. An automatic tool developed here would be a useful filter for the specialist to test technician diagnosis. This paper presents the development and validation of an algorithm to detect parasites in thick blood films. The algorithm is trained by selecting a number of correctly classified pixels from each tissue, which are then used as input to a statistical classifier based on the k-Nearest Neighbor (k-NN) decision rule in the RGB color space. A few samples of the classes to segment, i.e. red cells (RC), parasites (PA) and background (BG) are picked to obtain a representation of the pattern space. The k-NN rule classifies a given pixel within the category the most heavily represented among its k-Nearest Neighbors. Validation was performed on a group of six images. Parasite was detected in the entire set of test images and the number of misclassified pixels was observed to amount to 11.41 %.