An artificial intelligence model has outperformed experienced biologists in detecting salmon lice larvae in seawater samples. Researchers trained the system on a dataset of over 120,000 images of the parasites, which feed on the skin and blood of salmonids.
The breakthrough addresses a critical challenge for the aquaculture industry, where early detection of these parasites is essential to prevent outbreaks. Salmon lice can cause significant harm to farmed fish populations and economic losses.
According to the study published in a peer-reviewed journal, the AI models were both faster and more accurate than human experts at identifying the larvae. The researchers noted that the technology could be deployed for real-time monitoring in fish farms.
The system's speed allows for rapid screening of large water samples, potentially enabling earlier interventions. Fish farmers could reduce reliance on chemical treatments and improve fish welfare through more targeted parasite control.
Further field trials will be needed to validate performance under real-world conditions, including varying water clarity and light levels, before commercial deployment.