AI heart murmur detection tool for dogs reaches veterinary clinics: full analysis

Version 2

Boehringer Ingelheim and Eko Health have formally launched Eko Vet+ | CANINEBEAT AI, a new canine cardiac screening tool that brings AI-assisted murmur detection into routine veterinary exams. The product is designed to detect, visualize, and grade heart murmurs in dogs, with the companies reporting more than 95% sensitivity and specificity for murmurs associated with structural heart disease. It pairs Eko’s digital auscultation hardware and app with a canine-specific algorithm from Boehringer Ingelheim, aiming to help clinicians catch abnormalities that might otherwise go unnoticed during standard auscultation. (globenewswire.com)

The launch builds on a broader push to translate digital auscultation and murmur-detection AI from human medicine into veterinary workflows. Eko previously developed FDA-cleared murmur-detection tools for human care, and the canine version appears to use a similar technical foundation adapted for dogs. A recent Cambridge-linked paper described a transfer-learning approach that adapted a human murmur-detection algorithm to canine heart sound data, then fine-tuned it for veterinary use. That study said automated murmur grading could help identify dogs for earlier echocardiography referral and, in preclinical MMVD cases, potentially support earlier treatment decisions while reducing unnecessary referrals. (prnewswire.com)

According to the launch materials, CANINEBEAT AI was trained and validated on more than 4,000 annotated canine heart sound recordings. Eko says the algorithm is deployed on a fixed dataset without open updates, which is intended to keep outputs consistent across patients and visits. The hardware side includes the Eko CORE digital attachment, which the company says amplifies heart sounds up to 40 times and uses noise cancellation to help overcome common exam-room barriers, including ambient noise, movement, and subtle low-frequency murmurs. The app then packages the findings into murmur images, audio files, and reports that can be shared with pet parents. (globenewswire.com)

There’s also an MMVD-specific layer to the story. On the CANINEBEAT site, Boehringer Ingelheim says results from its LISTEN study are used within the app to provide insight into the likelihood of MMVD stage, while stressing that the output is not diagnostic and doesn’t replace echocardiography. The company says dogs in the study were prospectively enrolled and underwent digital auscultation and echocardiography during the same visit. However, the study is still listed as “in the process of submission,” so the supporting evidence for that staging-related feature has not yet been fully vetted in the peer-reviewed literature. (caninebeat.com)

Industry and expert messaging around the launch has centered on support for general practice. In dvm360 coverage tied to the announcement, veterinary cardiologist Gerhard Wess said the tool could help general practitioners detect and grade murmurs with a high level of consistency. In a separate dvm360 article, cardiologist Brian Scansen argued that subtle murmurs can be missed in practice, particularly in noisy settings or in diseases where outward signs may not be obvious early on, framing AI-assisted auscultation as a way to narrow the gap between specialist and frontline detection. (dvm360.com)

Why it matters: For veterinary professionals, this is less about replacing auscultation than about standardizing it. Interobserver variability, clinic noise, patient behavior, and hearing limitations all affect murmur detection, especially in primary care. A tool that improves consistency at the first point of contact could change which dogs get flagged for follow-up imaging, cardiology referral, serial monitoring, or treatment discussions. That may be particularly meaningful in MMVD, where timing matters and where murmur intensity can correlate with disease stage, though definitive staging still requires imaging. The practical value will depend on how well the tool performs across breeds, body sizes, coat types, arrhythmias, and real-world primary care settings outside company-supported studies. (ekohealth.com)

There are still important caveats. The headline performance claims come from company and product materials, and while they’re directionally supported by published academic work in this area, the exact commercial algorithm’s full validation dataset and methods don’t appear to be laid out yet in a peer-reviewed paper. The LISTEN study is also not yet published. For clinics considering adoption, that means the tool may be most useful today as an adjunct for screening, documentation, and client communication, rather than as a standalone basis for diagnosis or staging. (globenewswire.com)

What to watch: The next key developments will be peer-reviewed publication of the LISTEN study, any independent validation of the commercial algorithm in general practice, and early signals on whether AI-assisted auscultation changes referral patterns, treatment timing, or outcomes in dogs with suspected heart disease. (caninebeat.com)