VetFinder

Inspiration

Many Veterinary Doctors go unemployed for several years while many livestock die from curable diseases yearly, either because most livestock farmers in Nigeria don't know about veterinary doctors or because they have no way to reach out to them for their services. So we decided to create an app that use machine learning to solve this problem.

This article www.gamji.com/article8000/NEWS8651.htm also inspired me to build something that helps to create both awareness and employment for this profession neglected by most Nigerians,Veterinary Medicine.

What it does

VetFinder, through either a Q/A system or photo detection, uses machine learning to help livestock farmers diagnose common livestock diseases and recommends a veterinary doctor that can treat them based on the farmer's location/language And if a vet can't be found nearby, the farmer could choose to get virtual assistance from a professional vet anywhere through text or voice chat. Thereby providing employment for veterinary doctors while helping the Livestock farmers prevent their animals from dyeing and improving livestock production.

How we built it

We built a web app that allows veterinary doctors to sign up,create a profile the and their profile data is been saved to Firebase. Then in the Android app we retrieved the veterinary doctors profile and list them in the doctors category of the app. For our diagnosis, we used IBM Watson Assistant to create a simple diagnoses system that uses questions and answers to predict a disease or group of possible diseases affecting a livestock based on the symptoms been specified by the farmer.

Challenges we ran into

Our major challenge when building the App was getting an animal diseases dataset. We were unable to fully implement the photo diagnoses system because there are no currently available dataset including pictures of diseased farm animals so we had to start creating our own dataset which lead to many inconsistency and consumed most of our time.

Accomplishments that we're proud of

Building a Q/A system that can diagnose ANIMAL diseases was a big plus for us and using machine learning to solve veterinary problems is a rear use case of this powerful technology. So we are proud that we were able to break this problem down till it was solvable.

What we learned

Due to the fact that we have to do some Veterinary research on diseases and their diagnosis, we learnt a lot on the veterinary medicine part. But above all we learnt more amazing and fun ways machine learning and expert systems like IBM Watson could be used to diagnose complex Health problems not just in animals but also in humans...

What's next for VetFinder

There's much work still to be done,we haven't achieved some of our initial goals such as being able to connect the app to an external camera that can automatically detect and diagnose diseases in farm animals by just watching over them and sending feedbacks of an unusual behaviour on an animal to the farmer.

We hope to build a business model out of this app and getting more veterinary doctors and farmers to be aware of it's awesome capabilities and how helpful it could be to them.