East Coast Fever Early Detection in Kenya using Machine Learning

Abstract

The tick-borne protozoan parasite Theileria parva causes an acute, often fatal disease in cattle known as East Coast Fever (ECF). East Coast Fever is commonplace in Kenya and East Africa at large. This is because of the high temperatures experienced in these countries.
The incident of occurrence of East Coast Fever comes with a lot of financial losses due to less productivity. Late detection of ECF on a cow may lead to the death of the cow or huge amount of money spent in treatment. Vaccination of the cattle has come in handy in the containment of the disease but it has had its own shortcomings. The current conventional vaccine used in the region (Muguga cocktail) cannot be used across all areas in Kenya due to genetic diversity as well as antigenetic diversity of the parasite.
This paper seeks to find a way to detect the early signs of the disease using sensors implanted on the animals’ bodies so that early treatment can be done. . The real time data obtained from the sensors is sent to a machine learning model which then assesses it to determine the occurrence of ECF. The machine learning model is trained with sufficient data to be able to determine the signs of East Coast Fever. This will save the farmers the agony of having to be on constant lookout for presence of ECF which sometimes may not be reliable due to human errors. 
Proper sensor implanting has to be done in order to prevent sensor stress on the animals. With the success of this way of disease detection, then machine learning can be used in livestock farming for detection of other diseases and other animal behaviors such as mating that can make farmers to action which is an area that calls for further research.