Aquaculture 2025

March 6 - 10, 2025

New Orleans, Louisiana USA

Add To Calendar 09/03/2025 13:30:0009/03/2025 13:50:00America/ChicagoAquaculture 2025INCORPORATING MACHINE LEARNING INTO A FISH HEALTH DATABASEBalcony KThe World Aquaculture Societyjohnc@was.orgfalseDD/MM/YYYYanrl65yqlzh3g1q0dme13067

INCORPORATING MACHINE LEARNING INTO A FISH HEALTH DATABASE

Robert Durborow*, Sophia Okoh, Vincent Teye, Gunnar Psurny, Adetola Ogundipe, and Hamid Marvasti

Aquaculture Research Center                                                                       
Kentucky State University                                                                          
Frankfort, KY 40601                                                                                 
robert.durborow@kysu.edu

 



Since 1990, Kentucky State University has had a fish disease diagnostic laboratory, examining about 50 fish disease cases a year, including fish health inspections done to allow fish growers to transport and sell specific pathogen free fish into other states. This service to our stakeholders is regarded as an Extension function and has benefited fish owners including aquaculturists, federal and state fish & wildlife agencies, aquaculture researchers at universities and private businesses, ornamental fish hobbyists, and aquarium owners.

In 2022, KSU completed the development of a mobile responsive clinical fish health database that was used to upload over 1,400 fish disease cases including photographs and video clips. Having 35 years of fish disease cases in a database has enabled KSU Extension specialists, associates and students to search (with the use of a filter) a wide range of disease cases by year, farm owner, type of disease, and other parameters that have been entered into the database from 1990 through 2024.

KSU is now in the process of incorporating machine learning technology into the database to aid in disease recognition (for identification as well as predicting occurrences of diseases). Photographs and videos of diseased fish are being uploaded to the database for use in machine learning, enabling the database to identify these diseases and the likelihood of their occurrence.