A five-year-old terrier mix named Yogi walks in jaundiced, tachycardic at 160 bpm, vomiting for three days. The CBC flags elevated retic count, moderate neutrophilic leukocytosis, anemia, and possible thrombocytopenia. Bilirubin and liver enzymes are significantly elevated. You're looking at immune-mediated hemolytic anemia, hepatic disease, or both, and the low platelet count could be real or it could be clumping that's fooling the analyzer. In the traditional workflow, this is where you prep a smear, read it if you're confident, or FedEx it to the reference lab and wait.
Yogi's clinic had Vetscan Imagyst, Zoetis's point-of-care seven-in-one analyzer with an AI Blood Smear capability. The smear was scanned. Within minutes, the system returned annotated images, quantified findings, and highlighted abnormalities, spherocytes, polychromasia, ghost cells, anisocytosis, and platelet clumping. The clinical picture that had been a broad differential converged toward IMHA. The team then used the add-on Expert Review option, submitting the digital smear to a Zoetis board-certified clinical pathologist. Within hours, the specialist confirmed the AI-identified findings. Yogi's entire hematologic work-up, CBC, chemistry, AI blood assessment, and expert confirmation, was completed the same day. Intensive therapy for IMHA started that afternoon.
The workflow change, compressed
Vetscan Imagyst is a point-of-care analyzer that uses deep-learning algorithms trained on datasets of normal and abnormal hematology samples to evaluate red blood cells, white blood cells, and platelets. The system analyzes cell morphology, distribution, and hematologic features across the digital smear, then generates annotated images and a structured assessment. The company says the platform reduces the variability inherent in manual slide review.
The AI read happens in minutes. The add-on expert review by a board-certified clinical pathologist, which costs extra, returns within hours. Compare that to the reference-lab turnaround most practices are used to: prep the smear, courier it out, wait overnight or longer, get the report the next day or the day after. For cases where you suspect thrombocytopenia, regenerative anemia, or immune-mediated disease and time matters, the speed difference is the whole point.
The liability question nobody at Zoetis wants to answer in the brochure
The company frames this as a hybrid model: AI for speed and consistency, expert oversight for validation. That framing is careful. It sidesteps the question every practice owner is actually asking: if I run the AI read, skip the add-on expert review to save the cost, and the AI misses something or flags a false positive that sends me down the wrong treatment path, who owns that?
Zoetis says the AI is trained on extensive datasets and provides a high degree of consistency. The case study they published shows the expert review confirming what the AI found. What it doesn't show is a case where the AI and the expert disagreed, or where the AI missed a morphologic feature the pathologist caught. The add-on review exists for a reason, it's the safety net. But if you're paying for the analyzer and the per-sample AI read, then paying again for the expert review every time you need confidence, the cost structure starts to add up. And if you're NOT paying for the expert review on routine cases, you're betting the AI is bulletproof.
We write one of these every week for people running the business of veterinary medicine. Subscribe to VeterinaryPracticeNow.
For the associate DVM:
If your practice has this analyzer, the workflow is faster and the smear interpretation is more consistent than what you'd get from whoever on staff feels most comfortable at the microscope. That's real. The annotated images also make client communication easier, you can show the owner exactly what spherocytes or platelet clumping look like, which turns a confusing CBC into something they can see. The risk is that you start trusting the AI read without understanding its limitations, or your practice starts skipping the expert add-on to save money and you're the one who has to defend the diagnosis if it's wrong. Ask your practice owner what the protocol is: does every AI smear get expert review, or only the ones that look weird? That answer tells you how much liability you're carrying.
For the owner-DVM:
The per-sample cost structure has two tiers: the AI read, and the optional expert review. If you're paying for expert review on every sample that matters, your total cost per case includes both, and your gain over the reference lab is turnaround time. If you're NOT paying for expert review routinely, you're self-insuring on the AI's accuracy. The other operational question: who preps and loads the smear? If it's the DVM, you've added time to the appointment. If it's a tech, you need a tech who's trained on sample prep and comfortable with the workflow, and those techs are already overbooked.
What changes Monday
The cases where this matters most are the ones where you're staring at a CBC that doesn't quite add up and you need to know whether the thrombocytopenia is real, whether the anemia is regenerative, or whether you're looking at immune-mediated destruction. In those cases, same-day clarity changes what you can start and when. The risk is that the speed makes it easy to skip the expert review, and then you're diagnosing off an algorithm you can't cross-examine. The smarter play: run the AI, pay for the expert add-on on anything you're about to treat aggressively, and see how the system performs in your hands before you change your protocol.