AQUA 2024

August 26 - 30, 2024

Copenhagen, Denmark

USING MACHINE LEARNING IN MOBILE RESPONSIVE CLINICAL FISH HEALTH DATABASE TO RECOGNIZE DISEASE DEVELOPMENT

 Robert Durborow*,  John Kelso,  Vincent Teye,  Adetola Ogundipe, Tyler McKay,  and Tifani Watson McKay

 Kentucky State University Aquaculture Research Center                                                      103 Athletic Road                                                                                                                                              Frankfort, KY 40601                                                                                                                       robert.durborow@kysu.edu

 



Kentucky State University’s di gitized Mobile Responsive Clinical Fish Health Database  (for computers and mobile devices) enables more accurate disease case record-keeping, data mining, and timely identification and remediation of fish pathogens by fish health professionals, enhancing the effectiveness of fish health services to the aquaculture industry. The Database also serves as a reference source and teaching tool for fish disease diagnosticians and students. A current goal is to incorporate machine learning capabilities into the Database by uploading photographs of fish during the development phases of a disease (columnaris disease , Flavobacterium columnare, development will be used initially). Use of a monitoring camera and artificial intelligence to identify early stages of this disease in a fish population will be used when comparing the appearance of these fish to the uploaded photographs in the Database. It is anticipated that initial stages of columnaris will be detectable by a.i. before it is noticeable by those working with the fish, thus enabling earlier treatment of the disease, which would increase the chances of successfully treating the disease.