This paper provides an insight into a pioneering interdisciplinary research, amalgamating engineering principles and clinical diagnostic procedures. The research is underway to design and develop an automated parasite diagnostic tool, using machine learning based image processing operations. The preliminary findings of the work in which feature extraction algorithms are applied on clinical images as a base to a neural network model for diagnosis is described in this paper. Maximally Stable Extremal Regions (MSER) was used as a primary feature extractor in the detection of the parasitic eggs. Statistical information regarding the eggs was then collected and compared with background regions in the development of a search algorithm. Using the collected data, a range of values was obtained that could separate the eggs from the background of the images. All the images used are from Kato-katz slides.