Zoetis, the animal health company, agreed this week to acquire Texas-based VitalRADS, a network that provides around-the-clock review of animal radiology images. Zoetis plans to use VitalRADS’ image repository to train AI models aimed at automating parts of veterinary imaging workflows.
What the acquisition provides and why it matters
VitalRADS’ database contains more than 30 million images, which Zoetis intends to use as training data for machine learning models. Abhay Nayak, who leads US commercial operations at Zoetis, told Semafor that the acquisition will accelerate development of tools that can eventually take over substantial portions of technicians’ routine work.
Nayak said building a comparable image database internally would have taken a long time and also required assembling a team of radiologists, factors that made buying VitalRADS a straightforward decision for Zoetis. The company did not disclose the purchase price for VitalRADS.
Related deals and data strategy
Earlier this year Zoetis agreed to buy Neogen’s animal genomics business for $160 million. That transaction included a large collection of genetic testing records from customers in 120 countries; Jamie Brannan, Zoetis’s chief commercial officer, said at the time that the data would complement the herd-management software the company provides to farmers.
The VitalRADS deal is Zoetis’s second acquisition this year focused on data assets, reflecting a broader strategy in which non-technology companies acquire proprietary datasets to accelerate AI development.
Market context and technical challenges
Zoetis has faced pressure from investors as US pet owners have reduced spending: purchases of premium name-brand products have declined and veterinary visits have become less frequent. Zoetis’s share price dropped about 30% after its most recent earnings announcement.
Radiology has been one of AI’s early success areas in human healthcare, where it has helped improve radiologists’ workflows without markedly sacrificing diagnostic accuracy. Developing AI for animal radiology is more complicated, however, because of the diversity of species and the lower degree of standardization across scans, Nayak noted.
