BioMedaqu · Project

AI Software Detects Skeletal Defects in Farmed Fish Automatically

foodPrototypeTRL 4Thin data (2/5)

gigabiomedaqu.uliege.be

Fish farms have a persistent headache: a significant share of their fish grow up with bent spines or deformed skeletons, which hurts animal welfare and makes the product unsellable. Meanwhile, medical researchers studying bone diseases like osteoporosis use tiny zebrafish as stand-ins for humans — but the two worlds rarely talk to each other. BioMedaqu brought aquaculture scientists and biomedical researchers together across 9 European labs to share methods, train young scientists, and build tools — including AI software that can automatically spot and classify skeletal problems in fish from images.

By the numbers

research partners across consortium

European countries involved

EUR 3,747,204

total EU contribution

total project deliverables produced

European universities in the network

industry partner ratio in consortium

The business problem

What needed solving

Skeletal deformities in farmed fish are a persistent quality and welfare problem that causes product rejection and financial losses. Current detection relies heavily on manual visual inspection, which is slow, inconsistent, and catches problems too late in the production cycle. Fish feed companies also lack fast, standardized ways to measure how new feed formulations affect bone development during R&D trials.

The solution

What was built

The main commercially relevant output is an integrated web-based software for automatic classification of skeletal defects in fish, using AI to identify, annotate, and quantify skeletal elements from different image types (fluorescent, X-ray). The project also produced 33 deliverables spanning molecular, histological, biochemical, and cell culture methodologies relevant to skeletal biology.

Audience

Who needs this

Fish hatcheries and marine farms screening for skeletal deformitiesAquaculture feed companies testing bone-health effects of new ingredientsVeterinary diagnostics companies building aquaculture health platformsPharmaceutical companies using zebrafish models for bone disease drug screeningAquaculture certification and quality assurance service providers

Business applications

Who can put this to work

Aquaculture & Fish Farming

any

Target: Fish hatcheries and marine farms dealing with skeletal deformity losses

If you are a fish hatchery operator dealing with skeletal deformities that reduce product quality and cause batch rejection — this project developed an integrated web-based software that uses AI to automatically classify skeletal defects from fish images. Instead of relying on manual visual inspection by trained staff, you could screen larvae and juveniles faster and earlier, allowing you to cull or adjust feeding protocols before losses compound across a full production cycle.

Animal Feed & Aquaculture Nutrition

mid-size

Target: Fish feed manufacturers developing specialized nutrition formulas

If you are a fish feed company testing how new ingredients affect fish bone development — this project investigated how nutrition impacts skeletal formation using zebrafish and medaka models. The AI-based classification software can quantify skeletal defects across feeding trials automatically, replacing slow manual counting. This means faster R&D cycles when formulating feeds that reduce deformity rates in commercial species.

Veterinary Diagnostics & Imaging

SME

Target: Companies building diagnostic tools for aquaculture health monitoring

If you are a veterinary diagnostics company looking to expand into aquaculture health screening — this project built software calibrated for multiple image types including fluorescent and X-ray images across different fish species. The tool could be integrated into your existing imaging platforms to offer automated skeletal health scoring as a service to fish farms, adding a new product line without building the AI classification engine from scratch.

Frequently asked

Quick answers

What would it cost to license or use this skeletal detection software?

Based on available project data, no pricing or licensing model is described. The software was developed within a publicly funded MSCA training network (EUR 3,747,204 total EU contribution across 9 partners), so the tool may be available as open-source or through academic licensing. You would need to contact the coordinator at Université de Liège to discuss terms.

Can this work at industrial scale for a large fish farm?

The software was first calibrated for 9-day-old zebrafish larvae, then designed to extend to other species and image types (fluorescent, X-ray). Based on the deliverable description, the tool handles automatic identification and quantification, which suggests batch processing capability. However, validation at commercial aquaculture scale is not documented in the project data.

Who owns the intellectual property for this software?

The project was funded under MSCA-ITN with 9 partners across 6 countries, all academic or research institutions. IP ownership typically follows the EU grant agreement and institutional policies. Since the consortium has 0 industrial partners, commercial licensing would need to be negotiated directly with the coordinating university.

Does this software work for species other than zebrafish?

According to the deliverable description, the tool was first calibrated for zebrafish larvae, with explicit plans to extend to other types of skeletal images (fluorescent, X-ray) and other species. The project studied both zebrafish (Danio rerio) and medaka (Oryzias latipes), so at minimum two species were covered. Extension to commercial aquaculture species would require additional calibration.

How mature is this technology — is it ready for commercial deployment?

This was a Marie Skłodowska-Curie training network, primarily designed to train early-stage researchers rather than deliver market-ready products. The web-based software for skeletal defect classification represents a working prototype. No commercial deployment or industrial pilot testing is documented.

Are there regulatory requirements for using AI diagnostics in aquaculture?

Based on available project data, regulatory compliance is not addressed. Aquaculture health monitoring tools generally must meet national veterinary and food safety standards. Any commercial deployment of the AI classification software would need to be validated against existing inspection protocols in your jurisdiction.

Consortium

Who built it

The BioMedaqu consortium is entirely academic — 7 universities and 2 research institutes across 6 countries (Belgium, Germany, Spain, France, Italy, Portugal), with zero industrial partners and zero SMEs. For a business looking to adopt this technology, this means there is no established commercial pathway or industry champion within the project. The coordinator, Université de Liège in Belgium, would be the single point of contact. The project objective mentions "commercial interests represented by two Economy departments, one aquaculture and a major fish feed production company," but these are not listed as formal consortium partners, suggesting advisory rather than development roles. A business interested in the AI classification software would be dealing exclusively with academic institutions for licensing and technology transfer.

UNIVERSITE DE LIEGECoordinator · BE
UNIVERSITA DEGLI STUDI DI ROMA TOR VERGATAparticipant · IT
UNIVERSITAET MUENSTERparticipant · DE
INSTITUT NATIONAL DE LA SANTE ET DE LA RECHERCHE MEDICALEparticipant · FR
INSTITUTO PORTUGUES DO MAR E DA ATMOSFERA,IPparticipant · PT
UNIVERSIDAD DE LAS PALMAS DE GRAN CANARIAparticipant · ES
UNIVERSITEIT GENTparticipant · BE
UNIVERSIDADE DO ALGARVEparticipant · PT
UNIVERSITA POLITECNICA DELLE MARCHEparticipant · IT

How to reach the team

Université de Liège, Belgium — reach out through their technology transfer office or the project website contact page

Order a report on this consortium See UNIVERSITE DE LIEGE profile →

Next steps

Talk to the team behind this work.

Want to explore whether BioMedaqu's AI skeletal screening tool fits your aquaculture quality control process? SciTransfer can arrange a direct introduction to the research team and help you evaluate licensing options.

Order a report on this topic More in Food & Agriculture