Study adds breed-specific heart size ranges for eight dog breeds

Breed-specific cardiac radiography benchmarks continue to expand in small animal practice, with a new AJVR paper establishing vertebral heart size and vertebral left atrial size reference intervals for eight additional canine breeds: Australian Cattle Dogs, Shih Tzus, Maltese, Dachshunds, Greyhounds, French Bulldogs, Norwich Terriers, and Brittany Spaniels. The study was published July 10, 2026, and adds fresh data to a field that’s been moving steadily away from a one-size-fits-all interpretation of canine thoracic radiographs. (pubmed.ncbi.nlm.nih.gov)

That shift has been building for years. A recent literature review on VHS and VLAS concluded that when breed-specific reference ranges are available, clinicians should use them instead of multibreed reference values. The same review summarizes the traditional multibreed VHS benchmark from Buchanan and Bucheler as 9.7 ± 0.5 vertebrae, with a range of 8.5 to 10.6, but also catalogs substantial breed-to-breed variation, including higher expected values in breeds such as Boxers, Labradors, Whippets, Greyhounds, and Brittany Spaniels. (pmc.ncbi.nlm.nih.gov)

The new study’s design was straightforward and clinically relevant. According to the indexed abstract, dogs were included if they had two- or three-view thoracic radiographs and no cardiopulmonary or systemic disease, and all VHS and VLAS measurements were performed by the same board-certified cardiologist. That consistency matters because observer-dependent variation remains a known limitation of manual radiographic heart measurements. A separate 2026 PLOS One validation study on the AI tool RadAnalyzer notes that human VHS assessment can vary by as much as one vertebra between observers, and that VLAS also shows observer variability linked in part to clinician experience. (pubmed.ncbi.nlm.nih.gov)

The AJVR paper also fits into a larger publication trend from the same research group and others to build out breed-level intervals rather than relying on generic cutoffs. Earlier AJVR studies have already reported breed-specific VHS and VLAS ranges for Miniature Schnauzers, Shetland Sheepdogs, Cairn Terriers, Italian Greyhounds, English Springer Spaniels, and, in 2026, Rhodesian Ridgebacks, Borzois, Afghan Hounds, Salukis, and Basenjis. That accumulation of narrower reference data suggests the field is moving toward more tailored radiographic interpretation across both toy and athletic breeds, as well as brachycephalic dogs. (pubmed.ncbi.nlm.nih.gov)

There doesn’t appear to be a separate press release or broad industry statement tied to this AJVR paper, but the broader expert commentary is consistent. The recent review states that breed-specific ranges should take priority when available, and notes that ACVIM consensus guidance for dogs with myxomatous mitral valve disease uses a VHS threshold of at least 11.5 vertebrae, or a breed-adjusted equivalent, when echocardiography can’t be obtained. Clinician education materials also continue to frame VLAS as a practical adjunct for assessing left atrial enlargement, while acknowledging that historically published breed-specific VLAS data have been limited. (pmc.ncbi.nlm.nih.gov)

Why it matters: For general practitioners, emergency clinicians, and cardiology teams, this is less about a single new number and more about improving signal-to-noise in everyday radiograph interpretation. Dogs with suspected MMVD, occult cardiomegaly, or respiratory signs often get staged first with thoracic imaging, especially where echo access is limited. More precise breed-specific intervals can help reduce false positives in breeds with naturally larger silhouettes and sharpen suspicion in breeds whose normal values sit lower. That’s particularly relevant as radiographic measures continue to support screening, serial monitoring, and decisions about referral or next diagnostics. (pmc.ncbi.nlm.nih.gov)

The study may also have downstream value for veterinary AI. The PLOS One RadAnalyzer validation study, published in 2026, found strong correlation between automated and human measurements for both VHS and VLAS, and the review literature explicitly suggests that breed-specific reference ranges could be used to build more accurate AI models. If that happens, the practical endpoint for clinics could be faster, more standardized radiographic measurements paired with breed-aware interpretation rather than generic cutoffs alone. (journals.plos.org)

What to watch: The next step is whether these eight-breed intervals are incorporated into clinical reference tools, teaching materials, and software, and whether future studies validate how breed-specific VHS and VLAS cutoffs change diagnostic accuracy, staging, or referral decisions in real-world practice. (pmc.ncbi.nlm.nih.gov)