Project

Prediction of time to disease, e.g., canine hip dysplasia, by combining survival analysis techniques and artificial intelligence

Code
bof/baf/4y/2024/01/152
Duration
01 January 2024 → 31 December 2025
Funding
Regional and community funding: Special Research Fund
Research disciplines
  • Natural sciences
    • Statistics not elsewhere classified
    • Machine learning and decision making
  • Agricultural and food sciences
    • Veterinary medicine not elsewhere classified
Keywords
neural networks Canine Hip Dysplasia survival analysis
 
Project description

For many animal diseases it is essential to predict when an animal will develop the disease. based on current characteristics of the animal.  In the case of canine hip dysplasia, for instance,  it is important to predict at young age (typically based on radiographs) when the dog will develop symptoms of hip dysplasia. This enables the owner to take correct decisions with respect to including the dog in a breeding program and to train the dog to become a working dog, e.g., blind guide dog.

The modelling of time to event is typically based on survival analysis techniques  that have been developed in the field of statistics. The objective of the current project is to combine such survival analysis techniques with artificial intelligence,  mainly neural networks. Neural networks will thus be  extended so that they can also be used to predict time to event.