Fine needle aspiration, often abbreviated FNA, is one of the most powerful yet underappreciated tools in veterinary medicine. The technique involves inserting a thin needle into a mass and drawing out cells, which are then smeared onto a slide, stained, and examined under a microscope. Compared with surgical biopsy, it is minimally invasive, carries low risk, costs far less, and can produce results within minutes to days. When a dog or cat presents with a persistent lump measuring at least one centimeter, roughly the size of a pea, that has been present for a month or longer, established guidelines from the American Animal Hospital Association recommend FNA or biopsy to obtain a diagnosis before any treatment decisions are made. The rationale is straightforward: knowing whether a mass is benign or malignant, and what type of tumor it might be, fundamentally changes how a veterinarian plans surgery, staging, and follow-up care.

The stakes are considerable. Roughly 30 percent of all tumors in dogs and cats arise in the skin or the tissue just beneath it, and malignancy rates in these lesions vary dramatically between species. In dogs, an estimated 20 to 40 percent of skin tumors are malignant; in cats, the figure climbs to between 70 and 82 percent. Because a lump that feels benign to the examining hand may in fact be an aggressive cancer, the guidelines emphasize that no definitive diagnosis can ever be made on palpation alone. Yet the new survey suggests that in everyday practice, clinical impression frequently wins out over cytology.

The research team, led by Gillian Dank of Onco Pet Veterinary Services in Israel along with colleagues at HT Vet and in private consulting, distributed a 16-question survey by email to 10,000 small-animal veterinary professionals across the United States, including veterinarians, veterinary technicians, and practice managers. A total of 240 respondents completed the questionnaire, a response rate of 2.4 percent, with 235 providing fully answered forms. Most respondents were veterinarians, about 62 percent of practices were independently owned, and the vast majority worked in general practice. The survey asked about daily caseloads of mass patients, the proportion of cases in which FNA was performed, whether samples were evaluated in-house or sent to reference laboratories, the perceived rate of inconclusive results, and the barriers that discouraged the test.

The statistical analysis, conducted in Python using standard tests of association, revealed several striking patterns. Clinicians who evaluated four or more patients with dermal or subcutaneous masses per day were significantly less likely to perform FNA than those seeing fewer than four such patients daily, a difference that reached statistical significance. The mean FNA rate fell from roughly 41 to 42 percent among lower-volume respondents to 33 percent among those with the heaviest caseloads. The authors suggest that the cumulative cognitive load of making dozens of clinical decisions each day may push busy veterinarians toward a quick wait-and-see discharge rather than an additional procedure, particularly when a lump is an incidental finding during a visit for something else.

Experience, surprisingly, worked in the opposite direction. Clinicians with more than 20 years in practice were significantly over-represented among the high users of FNA, performing the test on more than half of masses at rates of 40 to 44 percent, compared with only 19 to 24 percent among less experienced colleagues. This contradicts the intuitive assumption that seasoned veterinarians feel comfortable judging a lump by look and feel alone. The authors speculate that decades of practice bring greater familiarity with the aspiration technique and a deeper appreciation of its clinical utility, gained through years of watching palpation-based guesses go wrong.

The barriers veterinarians cited for skipping the test were revealing. The most common primary reason, selected by 35 percent of respondents, was the perception that the mass was benign and did not require further investigation. Cost concerns came second at 32 percent, and client declination for reasons other than cost accounted for 19 percent. Taken together, the cost and refusal categories suggest that in roughly half of cases, the veterinarian would have performed the aspiration if the owner had agreed. Only 8 percent cited limited confidence in the diagnostic value of FNA itself, and just 4 percent pointed to the time the procedure takes. Notably, when the same respondents were asked what percentage of lumps could be accurately diagnosed from palpation and history alone, half estimated that this was reliable in fewer than 10 percent of cases, a striking contradiction with their stated reliance on clinical impression. The authors interpret this as a gap between theoretical knowledge and the practical, split-second decision-making of a busy appointment.

One of the most consequential findings concerned surgery. Respondents reported performing excisional biopsies, in which the entire mass is surgically removed, without any prior FNA in an average of 30 percent of cases. Clinicians who expressed confidence in diagnosing lumps by palpation alone were two and a half times more likely to skip the preoperative aspiration. This matters because knowing the tumor type before surgery provides critical information about local invasiveness, metastatic potential, and whether the cancer spreads through lymphatic or blood vessels, all of which influence the width of surgical margins, the need for preoperative staging tests, and the overall surgical strategy. Since the first operation often offers the best chance of complete tumor removal, proceeding blind may, in some cases, compromise the outcome. The authors caution, however, that there are legitimate reasons to excise directly, such as very small masses that appear benign, or palliative procedures where tumor typing would not change the plan.

The survey also examined the role of digital cytology, a rapidly growing technology that allows slide images to be transmitted to remote board-certified clinical pathologists for interpretation. About 35 percent of clinics reported using such services, with medium and large practices significantly more likely to adopt them than small ones. Counterintuitively, clinics with digital cytology performed fewer aspirations overall, 35 percent versus 42 percent, evaluated fewer samples in-house, and reported higher rates of inconclusive results. The authors offer several possible explanations: clinicians may expect definitive answers from submitted samples and grow frustrated when findings are non-diagnostic; the higher cost of digital services may discourage routine use; slide preparation and upload take time during busy consultations; or the association may simply be confounded by the fact that larger, high-volume clinics, which already perform fewer aspirations, are also the ones most likely to have the technology.

The study has limitations that the authors acknowledge candidly. The response rate was low, the survey reached only US-based subscribers of a single professional publication, and many answers relied on recollection rather than verified clinical records. The questionnaire also did not capture the clinical reasoning behind every decision to skip aspiration. Even so, the overall picture is consistent: a fast, cheap, minimally invasive test that veterinary schools teach and guidelines endorse is being underused, especially where caseload pressure is highest. The authors argue that better integration of FNA into routine workflows, along with hands-on training for veterinary students and practicing clinicians in both aspiration technique and slide assessment, could close the gap between what the guidelines recommend and what actually happens when a worried owner points to a lump on their pet. Early diagnosis, they emphasize, remains one of the most powerful predictors of a good outcome in veterinary oncology, and the pea-sized lump dismissed today may be the aggressive tumor that a two-minute needle test could have caught in time.

Subject of Research: Underuse of fine needle aspiration for diagnosing skin masses in dogs and cats in veterinary practice

Article Title: Underutilization of fine needle aspiration for canine and feline superficial masses-survey results

Article References: Dank, G., Buber, T., Levy-Hirsch, L., Fink, A., Marks, N., Greenberg, U., & Hanael, E. (2026). Underutilization of fine needle aspiration for canine and feline superficial masses-survey results. Veterinary Oncology, 3(1), Article 25. https://doi.org/10.1186/s44356-026-00077-0

Image Credits: AI Generated

DOI: 10.1186/s44356-026-00077-0

Keywords: fine needle aspiration, veterinary oncology, cytology, dogs, cats, skin masses, biopsy, digital cytology, diagnostic guidelines, survey research, cancer diagnosis, veterinary practice

Lymphoma is one of the most common malignancies in dogs, and its diagnosis typically begins with fine needle aspiration of an enlarged peripheral lymph node. Cytology, the microscopic examination of cells harvested in this minimally invasive way, is fast and generally reliable, but it is not always definitive. Confirming a diagnosis of lymphoma and characterizing it as B-cell or T-cell disease usually requires histopathology with immunophenotyping through PCR for antigen receptor rearrangement, flow cytometry, or immunohistochemistry. Those ancillary tests add cost, time, and logistical complexity, and they are not uniformly available across veterinary practices. Because microscopic evaluation demands trained pathologists, high-quality optics, and consistent staining reagents, diagnostic delays can affect clinical decisions and outcomes for canine patients.

The Florida team asked whether a convolutional neural network, the class of deep learning architecture that has transformed medical image classification, could shoulder part of this diagnostic burden. They selected ResNet-50, a fifty-layer residual network pretrained on the ImageNet database, and fine-tuned it through transfer learning on cytological photomicrographs captured at 1,000-fold magnification with immersion oil. The researchers assembled their dataset from cases submitted between 2017 and 2024 to the University of Florida Small Animal Hospital, drawing on the archive of the Clinical Pathology Service and cross-checking electronic medical records to verify each diagnosis.

Strict inclusion and exclusion criteria shaped the final cohort. Lymphoma cases required a cytological diagnosis of intermediate or large-cell lymphoma confirmed by PARR or flow cytometry, while reactive lymphoid hyperplasia cases required a predominance of small mature lymphocytes with plasma cells and no concern for a neoplastic population. Three board-certified clinical pathologists independently reviewed every selected slide to ensure consensus. The resulting dataset comprised 260 lymph node slides from 184 dogs, from which 2,600 images were captured: 1,070 images of B-cell lymphoma from 72 dogs, 490 images of T-cell lymphoma from 35 dogs, and 1,040 images of reactive lymphoid hyperplasia from 77 dogs. Most slides were stained with Wright-Giemsa, with a small number stained with Diff-Quick to reflect real-world staining variation.

A critical design decision was that the dataset was partitioned at the patient level rather than the image level. This meant that no images from a dog in the test set ever appeared in the training or validation pools, preventing information leakage from inflating performance. Approximately 20 percent of patients were held out for final testing, while the remaining 80 percent were used for 10-fold cross-validation, with hyperparameter tuning and early stopping applied within each fold. Images were standardized to 1920 by 1080 pixels, normalized using ImageNet channel statistics computed without reference to the validation or test data, and augmented on the fly with random horizontal and vertical flips during training.

The first task was binary classification: lymphoma versus reactive lymphoid hyperplasia. Across ten independently trained models, test accuracy ranged from 85.4 percent to 95.4 percent, with a mean of 90.3 percent, and the area under the receiver operating characteristic curve ranged from 0.943 to 0.988. The best-performing model achieved an accuracy of 0.954 with a Matthews correlation coefficient of 0.904, sensitivity of 0.950, specificity of 0.958, and an F1 score of 0.957. The authors note that this level of sensitivity matches or slightly exceeds the roughly 93 percent sensitivity reported for clinical pathologists distinguishing lymphoma from non-lymphoma in cytology, and it clears the 94 percent threshold proposed for decision-rule validation in this domain. Training, validation, and test performance remained closely aligned across folds, suggesting minimal overfitting.

The second task was far harder: distinguishing B-cell from T-cell lymphoma using cytological appearance alone. This is notoriously difficult even for experienced pathologists, because both immunophenotypes often present as morphologically similar populations of large or intermediate lymphocytes, with only subtle clues such as occasional uropods in some T-cell cases. Unsurprisingly, the models performed more modestly here, with test accuracies between 66.2 percent and 79.8 percent, a mean of 74.3 percent, and area under the curve values between 0.653 and 0.748. The representative model achieved an accuracy of 0.743, though confidence intervals were wide given the limited number of T-cell patients. The authors point out that some discordance may stem from the reference standards themselves, since PARR and flow cytometry do not always agree, especially in samples with low tumor burden.

The study also carries important caveats that the authors address directly. The entire dataset came from a single veterinary teaching hospital, raising questions about geographic and equipment-related bias. Images were manually captured from representative fields chosen by pathologists rather than from whole-slide imaging, confirming high image quality but simplifying the task relative to real diagnostic workflows. Cases of small-cell lymphoma, which account for fewer than two percent of canine lymphomas and are extremely difficult to separate from reactive hyperplasia on cytology alone, were deliberately excluded, meaning the model’s scope is limited to the most cytologically distinct presentations. Class imbalance and the predominance of one imaging system may also have influenced performance, and the researchers emphasize that robust external validation on prospectively collected, multi-institutional data is essential before any clinical deployment.

Transparency was a stated priority throughout. Unlike conventional practice, where a hold-out test set is evaluated once, the team reported results from all ten cross-validation folds and identified the best-performing model, providing a detailed suite of metrics including confidence intervals derived from patient-level bootstrapping with 1,000 iterations. The source code for preprocessing, training, and evaluation will be released publicly on GitHub, and the anonymized image dataset is available from the corresponding author upon reasonable request under a data use agreement. The authors argue that this openness stands in contrast to the proprietary opacity that often surrounds commercial AI diagnostic tools, such as the cytology-triage platforms already entering veterinary practice.

Looking forward, the researchers frame their models as triage and decision-support tools rather than replacements for pathologists, insisting that a board-certified clinical pathologist must retain final diagnostic responsibility and interpret model outputs alongside clinical history and ancillary tests. Future work will explore whole-slide imaging pipelines that could mirror the way pathologists actually examine slides, explainable AI techniques such as Gradient-weighted Class Activation Mapping to reveal which cellular features drive predictions, other CNN architectures, and multi-institutional datasets spanning diverse stains, scanners, and patient populations. The moderate success of B-cell versus T-cell classification, despite the well-known morphological overlap between these immunophenotypes, hints that deep networks may be extracting genuinely subtle cytomorphological patterns. For now, the study establishes a careful, reproducible performance baseline and an explicit methodological roadmap for bringing AI-assisted cytology from the research bench toward the veterinary clinic, where faster answers could mean earlier treatment for dogs with lymphoma.

Subject of Research: Deep learning classification of canine lymphoma in lymph node cytology images

Article Title: Deep learning-assisted cytological image analysis for canine lymphoma

Article References: Deep learning-assisted cytological image analysis for canine lymphoma. (n.d.). https://doi.org/10.1186/s44356-026-00056-5

Image Credits: AI Generated

DOI: 10.1186/s44356-026-00056-5

Keywords: canine lymphoma, deep learning, cytology, ResNet-50, veterinary pathology, convolutional neural networks, B-cell lymphoma, T-cell lymphoma, reactive lymphoid hyperplasia, image classification, veterinary oncology, artificial intelligence