Our Advisory Board
Michael Bailey
DVM, DACVR
Bill Blevins
DVM, MS, DACVR
Hans Castelijns
DVM, Farrier
Sonya Gordon
DVM, DVSc, DACVIM
Ayman Mostafa
DVM, MS, PhD
Kurt Selberg
DVM, PhD, DACVR
Tracy Turner
DVM, MS, DACVS, DACVSMR
Our Mission Statement
Build an intelligent assistant to aid the veterinarian or radiologist in the interpretation of veterinary diagnostic images.
Philosophy & Approach
Metron-IQ is a virtual assistant which knows how to do certain tasks with veterinary images at human expert level. In particular, Metron-IQ is skilled at making measurements as described in veterinary literature. Additionally, it can point our certain anomalies, and automatically generates fantastic reports for your records and/or your clients.
Metron-IQ reports are PDF files that contain live links to the relevant veterinary literature regarding the various measurements that it performs. Metron-IQ does not diagnose, but can provide an evidence base for your diagnoses and aid in communications with your clients. Results from Metron-IQ are initially water-marked with "Ready to Review" so that you are kept in the loop and can override your assistant if you do not agree with the results.
We are working with world experts to help guide us so that Metron-IQ will become and ever-more valuable assistant, saving you time and money.
How to Use an AI Assistant
It is important to understand two features of any AI implementation: Scope and Performance.
All AI systems work within a limited scope. Within their scope, they might perform as well or better than a human expert, but outside of their scope, they have no common sense.
We feel it is important to keep the human in-the-loop. We are building an assistant which, like any assistant, needs a boss to keep an eye on it. For this reason, our results are always returned with a "Ready to Review" watermark, so that you are reminded to do a quick check of your assistant's work. You can approve of the results with a couple of clicks, or you can edit the report as you desire, changing mark-up, adding comments, etc.
AI tools will grow and improve in both Scope and Performance as the on-going build out continues. This is true of virtually all AI efforts occurring in the world, whether or not the developers state it this way. For example, our algorithms do not (currently) look for fractures - they are outside of the current scope of our tools. So, if you use our Canine Elbow Tool and it does not mention that the radius is fractured, do not be upset or think less of your assistant - fractures are simply outside of the current scope. (fracture detection will be added in the future!)
As a user of any AI system, you can avoid frustration and best leverage the tool if you understand its scope and performance limitations. That is why we try to document each of our tools as far as what it does and what its performance is. For example, have a look at our validation page for our VHS/VLAS tool here.
Our scope, or coverage of various anatomies and modalities will grow as we introduce more tools. Within the workings of each tool, its performance will likely increase as we periodically re-train as we gather more data and more wisdom.
“Doctors will not be replaced by AI. Doctors who use AI will replace doctors who don’t.”
— Dr. Curt Langlotz, Professor of Radiology and Biomedical Informatics, Stanford University.
MetronMind Team
Team CAlifornia
Team Wisconsin
Team Nairobi
Listen to our Dr. John Craig interviewed on the Veterinary Innovation Podcast…