The shrimp industry faces a formidable enemy: the White Spot Syndrome Virus (WSSV). This highly contagious pathogen can decimate shrimp populations, causing up to 100% mortality in a matter of days. With its devastating impact, finding effective detection and mitigation strategies is crucial for the future of sustainable shrimp farming.
But shrimp farmers now have a powerful tool in the fight against WSSV: deep learning.
A team of scientists from E.G.S. Pillay Engineering College, University College of Engineering, and SASTRA Deemed to Be University set out to classify shrimp infected with WSSV based on a deep learning methodology (Artificial Intelligence).
WSSV: A Stealthy Threat
WSSV belongs to the Nimaviridae family and specifically targets crustaceans, with cultivated shrimp species being the most affected by this virus.
The White Spot Syndrome Virus attacks all age groups and infects key tissues such as lymph nodes, gills, and the outer layer, causing 100% mortality in just 3 to 10 days.
“The complete sequencing of the genome related to WSSV strains from Thailand, China, and Taiwan has identified small genetic variations among them,” reported the scientists; however, the reasons behind its potency remain a mystery.
Scientists continue to debate the role of genome size, protein expression, and even the number of repetitive DNA sequences in determining WSSV virulence.
Deep Learning
This research proposes a new deep-learning approach to accurately classify shrimp infected with WSSV. Here’s how it works:
- Data collection: Shrimp images are collected from farms and online sources, providing a diverse dataset to train the algorithm.
- Preprocessing and segmentation: The Local Binary Patterns (LBP) technique cleans and prepares images, while Voronoi Fuzzy C-Means guided by textures (TGVFCMS) identifies and separates relevant features.
- Feature extraction: Linear Discriminant Analysis (PLDA) extracts key features from segmented images, highlighting differences between healthy and WSSV-infected shrimp.
- Classification and optimization: Finally, the Enhanced Gated Recurrent Unit (EGRU) network, optimized by the Wild Goose Migration Optimization (WGMO) algorithm, classifies each shrimp as healthy or WSSV-infected.
“The accuracy performance indicators have been compared with those of various conventional methods, and the results show that the methodology is capable of accurately identifying shrimp WSSV disease,” reported the study authors.
A Tool to Combat White Spot Syndrome Virus
This deep learning methodology outperforms traditional methods in accurately identifying shrimp infected with WSSV. This breakthrough paves the way for:
- Early detection and quarantine: Infected shrimp can be identified and isolated quickly, preventing the spread of the virus.
- Specific treatment and prevention: Interventions can be developed based on the severity and type of WSSV strain.
- Sustainable aquaculture: By protecting shrimp from WSSV, we can ensure a safe food supply and safeguard the livelihoods of countless people.
Conclusion
The success of this deep-learning approach could change the game for the aquaculture industry. Early detection of WSSV outbreaks allows for rapid intervention, potentially saving farms from catastrophic losses. Additionally, reducing dependence on antibiotics and other traditional methods through the adoption of sustainable solutions like deep learning aligns with the goal of responsible and environmentally friendly aquaculture practices.
The battle against WSSV is far from over, but innovative solutions like this deep-learning approach offer a ray of hope. Harnessing the power of technology, we can safeguard the future of shrimp farming and ensure a safe and sustainable supply of this vital source of seafood.
Contact
L. Ramachandran
Department of Electronics and Communication Engineering
E.G.S. Pillay Engineering College
Nagapattinam, Tamilnadu, 611002, India
Email: fourstar.lr@gmail.com
Reference
Ramachandran, L., Mangaiyarkarasi, S.P., Subramanian, A. et al. Shrimp classification for white spot syndrome detection through enhanced gated recurrent unit-based wild geese migration optimization algorithm. Virus Genes (2024). https://doi.org/10.1007/s11262-023-02049-0
