How to enhance veterinary radiology with AI on this episode of The Vet Blast Podcast featuring Seth Wallack, DVM, DACVR
The applications of artificial intelligence (AI) in veterinary radiology are the subject of this episode of The Vet Blast Podcast with Seth Wallack, DVM, DACVR. Wallack explains how advances in AI are changing the game in radiology by improving efficiency with no changes in the clinician’s workflow.
Below is a partial transcript. Listen to the full podcast for more.
Paul Bloom, DVM, DACVD, DABVP (Canine and Feline): The study for the cardiogenic pulmonary edema is a study done at AMC, New York. And we have this other study that’s also being published now on pleural fluid that was done at Tufts, again, helping with this clinical decision support system for that. So the study out of the AMC of New York is, what it showed was, of those 500 cases, 46 of them were in heart failure. The ground truth was the radiologist diagnosing heart failure. Our system agreed with the radiologist in 42 out of 46, or 92% agreement to the radiologist.
[The] study at Tufts for pleural fluid had a very similar number as well in terms of radiologist agreement, but that actually . . . the Tufts study . . . [the] ground truth is actually surgical or path or ultrasound confirmed. So that’s a different ground truth. That’s where they confirm that there was pleural fluid. And we had an 88% agreement in the AI versus background truth. Both the studies, we’re really happy. The performance is fantastic.
This episode is sponsored by Vetology.