Aquaculture plays a crucial role in food security but faces challenges like infectious diseases. Pathogens, including viruses, fungi, bacteria, and parasites, can cause significant losses, especially in closed systems like Recirculating Aquaculture Systems (RAS). Early disease detection is crucial to prevent mortality that can even be up to 100%.
This project aims to integrate machine learning/AI into the current Mobile Responsive Clinical Fish Health Database created at Kentucky State University (KSU). This integration will allow the database to analyze images and videos of fish to identify disease progression in different species, including clinical signs such as scale loss, abnormal skin growth, and the early stages of lesions. In addition, abnormal fish behavior will be documented and compared to healthy fish.
The project initially focuses on the early detection of columnaris disease in largemouth bass, caused by Flavobacterium columnare, using machine learning. This approach would help detect diseases several days earlier than usual, giving fish farmers a head start on treatment and reducing fish mortalities. The project will also expand the database’s capabilities to include other pathogens such as Saprolegnia fungus, Ichthyophthirius multifiliis, and Aeromonas bacteria. Data from U.S. diagnostic labs will enhance a database for proactive fish health management, benefiting aquaculturists worldwide by improving disease management and reducing losses.