Use of artificial intelligence for secure identification of dogs in low- and middle income countries

Artificial intelligence for facial recognition are novel tools currently in development, but little is known on feasibility for dogs in low- and middle income countries (LMICs). This is an urgent knowledge gap, given that microchip technology for dog registration systems in LMICs are not affordable and accurate identification of dogs is of utmost importance for disease control.  

The proposed research project benefits from the ongoing BlockRabies project in Côte d’Ivoire, which implements a blockchain (BC) secured software application (BlockRabies App) with the overarching aim to reduce mortality of rabies exposed human beings. The App combines electronic health records, the veterinary services, the vaccine supply chain and the health information system. However, dog identification is a current bottleneck of our highly secure system. If the biting dog can reliably be diagnosed as rabies negative, expensive human rabies vaccinations can be stopped. 

Aim of the proposed research project is to test Artificial Intelligence (AI) for secure identification of dogs in Switzerland and Côte d’Ivoire. We will program a Machine Learning software and test technical feasibility of AI using dog images, before validating the software in BlockRabies. Specifically for rabies, this would result in hugely increasing cost efficiency of human post-exposure prophylaxis (PEP) by reducing waste of expensive and scarce rabies vaccine. Use of AI for dog registration could become an example of trickle-up innovation with potential for implementation worldwide.