Advances in AI-Driven Fish Disease Detection: A Comprehensive Review of Image-Based and Machine Learning Approaches (2019–2025)

Fish disease outbreaks continue to pose a major challenge to global aquaculture by causing mortality, reduced growth performance, and economic losses. Rapid and accurate disease detection is therefore pivotal for health management in fish farming systems. In recent years, artificial intelligence based methods have drawn interest as tools for automating disease detection and enabling non-invasive monitoring in aquaculture. This review examines eight peer-reviewed studies published between 2019 and 2025, with developing research primarily densely between 2021 and 2024, that applied artificial intelligence techniques for fish disease detection. The studies mainly focused on image-based analysis using machine learning and deep learning models like CNN, along with advancing non-invasive and integrated monitoring frameworks. The convolutional neural network based models were most commonly employed for identifying externally visible disease symptoms from fish images, while some studies investigates hybrid systems that combined visual data with additional environmental information. The studied literature confirms a the potential of AI driven methods for automated fish disease screening under controlled experimental conditions. However, several limitations remain evident, including small and species specific datasets, different image quality, limited field validation, and challenges related to model interpretability and application in real under water environment. The findings of this review indicate that AI-based fish disease detection has progressed significantly, further research is needed to improve reliability, consistency , and practical applicability. Future efforts should prioritize standardized data information, field level validation, and user system design to support the adoption of affordable and reliable AI tools for fish health management.