This research will create a near real-time decision-making system to allow the early detection of an avian influenza outbreak, enhancing the ability of farmers and policymakers to respond and prevent threats to animal health and food security.
Time plays a key role in avian influenza disease management and control. However, disease data access and use can be very time-consuming and costly. Given the rapid decision-making required to respond to an avian flu outbreak, epidemiologists do not have the time to develop models of the outbreak to make the best recommendations for controlling the outbreak.
The project’s main goal is to develop a multi-pronged system to detect and provide an early warning of avian flu outbreaks that can be used under real-world circumstances in Canada.
This system will allow for near real-time analysis to detect an outbreak in a region and predict and evaluate risk factors for an outbreak, allowing for continual improvement of the analysis based on new information. The approach can be expanded to other animal infectious disease outbreaks, allowing for early detection and addressing threats to animal health in the future.
The researchers will use machine learning to detect hidden patterns that indicate the risk of avian flu, focusing on spatial analysis and social media and news analysis.
The estimated losses due to an avian flu outbreak in British Columbia (BC) in 2004 were around $380 million CAD. More recent (2022) outbreaks in Ontario and all other provinces in Canada (except for PEI) have also caused considerable losses. Aside from the great impact of avian flu outbreaks on animal health, some of these viruses impact public health. Early detection and real-time predictive analysis will decrease the spread and severity of future outbreaks in the Canadian poultry industry, increasing animal well-being and human health.
Dr. Rozita Dara, Dr. Lauren Grant, Dr. Zvonimir Poljak, and Fatemeh Haghighi
