Vetology Overhauls AI Diagnostic Suite to Mimic Human Radiology

The project, conducted between June and August, marks a strategic pivot toward higher diagnostic precision. According to Data Science Manager Anthony Lutz, the new architecture integrates information both within individual images and across linked study sets. This design change, guided by ACVR diplomates Dr. John Mattoon and Dr. Jennifer Gambino, aims to address performance gaps identified in recent veterinary industry literature.

This update follows a March 2025 position statement from the ACVR and ECVDI, which called for greater transparency and third-party validation in veterinary AI. While a 2026 pilot study published in JAVMA found mixed performance results among commercial diagnostic tools, Vetology is positioning its new suite as a response to those professional benchmarks. The company continues to invite independent researchers to audit its performance data, which it began publishing at the condition level in January 2026.

President Eric Goldman emphasized that the rebuild was driven by a need for clinical accountability. The platform now features continuous feedback loops, allowing practicing veterinarians to flag concerns that are subsequently reviewed by board-certified specialists. Despite the increased automation, VP of Sales Pierre D'Amours noted that the AI remains a consultative tool rather than a replacement for clinical judgment. The technology serves as a secondary input for the veterinarian, who retains full responsibility for the final diagnosis and owner communication.