Deep learning algorithm to detect presence of malaria parasite in Blood slides

Abstract

Malaria has been a major problem in developing countries like India ( because of favourable temperature for the mosquitos in Tropics), malaria has been around for centuries. From late 1800s to early 1900s almost one fourth of India’s population suffered from malaria, particularly tribnal belts of Chhattisgarh , Orissa and Jharkhand (Citeseerx.ist.psu.edu, 2019)i The economic loss was estimated to be around Rs. 1000 crores per year due to loss in man hours in the year 1935. However, the situation improved drastically from April 1953, when Govt. of India launched the National Malaria Control Program (NMCP) resulting in drastic decrease in number of cases and mortalities. However, the cases of malaria kept resurging in late 70’s and early 80’s. In recent times, going by the World Malaria Report of 2017, close to 6% of the global malaria cases and deaths due to the disease are accounted for by the Indians. Furthermore, it is estimated that more than 50% of the population is at the risk of contracting malaria (Malaria Site, 2019)ii . Machine Learning Techniques like Deep Learning Algorithms could be used to detect presence of malaria parasite in blood slides. In our study, we intend to use deep learning techniques like Convolutional Neural Networks (CNN) to help classify whether a given blood slide has malaria, through image classification