Which species, stains, and sample types can be analyzed?

Projects commonly involve companion animals, livestock, poultry, and laboratory species, using formalin-fixed paraffin-embedded, frozen, or cytology preparations. Haematoxylin and eosin, special stains, chromogenic immunohistochemistry, and immunofluorescence are workable when staining is consistent across the slide set. We first review a representative subset to confirm that tissue quality and scan resolution support the planned measurements.

How do you decide what to annotate and measure?

Annotation follows the study question, not the software default. If the endpoint is overall tumour burden, the rule may define the entire tumour outline; if the endpoint is proliferative activity, a hotspot rule is applied consistently to every section. Exclusion rules for necrosis, folds, and edge artefacts are written down at the same time, so the rule set extends to follow-up studies.

Can you quantify IHC markers such as Ki-67, CD3, or IBA-1?

Yes. Nuclear markers such as Ki-67 are usually reported as positive nuclei per square millimetre, labelling index, or H-score, while membrane and cytoplasmic markers such as CD3 or IBA-1 are reported as cell counts and densities per region. Intensity thresholds are fixed across the study slide set rather than adjusted per image, so group comparisons are not distorted.

How is the quality of automated results controlled?

Detection settings are tuned on representative slides and then locked for the study. Overlay images are reviewed so that missed nuclei, false detections, and incorrect tissue masks are seen rather than assumed, and outlier slides are inspected individually. Slides failing scan quality or staining requirements receive a QC flag.

What do you need from me to start an image analysis project?

Useful inputs include the species and tissue, the stain or marker and its target compartment, the study groups and sample size, the endpoint you want to test, the number of slides or archive location, and any reporting format your team already uses. If the marker signal is not yet established on the tissue, we can start with assay development and continue into quantification on the same material.

Digital Pathology Image Analysis for Veterinary Research

When Pathology Scoring Stops Scaling With the Study

Two pathologists reading the same animal tissue section can reach different scores, and manual grading rarely keeps pace with a study that generates hundreds of slides.

BioVenic delivers veterinary digital pathology support that replaces single-observer impressions with measured image data. We digitize sections, fix annotation rules before scoring begins, quantify lesions and markers, and report animal pathology image analysis results that hold up when a reviewer asks how a number was produced.

Digital Pathology Image Analysis Service Scope

The service is modular and study-driven. A project can begin with one marker on a single cohort and extend to a full slide archive, provided annotation rules and endpoints are agreed before scoring starts.

01

Slide Digitization and Image QC

Slides are scanned at an agreed resolution and checked for focus, tissue coverage, and artefacts before measurement.

02

Annotation and ROI Strategy

Tumour, lesion, or tissue compartments are outlined and non-target regions excluded under a documented rule set.

03

Lesion and Tissue Quantification

Lesion area, affected tissue fraction, necrosis, and inflammatory infiltrate are segmented and measured across the section.

04

IHC Marker Quantification

Nuclear and cytoplasmic IHC signals are classified by intensity and reported as positive-cell counts, densities, labelling indices, or H-scores.

05

Multiplex and Object-Level Analysis

Additional channels, double stains, or cell-level phenotyping are quantified with the same documented pipeline.

06

Image-Derived Endpoint Reporting

Per-animal metrics, group summaries, QC flags, and analysis rules are compiled into statistical-analysis-ready tables.

Quantification Targets and Typical Image-Derived Outputs

Analysis Target	Typical Endpoints	Study Question Supported
IHC marker of interest	Positive nuclei per mm², labelling index, H-score	Whether marker expression separates groups or timepoints
Lesion and tissue area	Lesion area fraction, affected-to-total tissue ratio	How far tissue injury or inflammation has progressed
Cellular infiltrate	Cells per region, density per mm²	Immune recruitment in infection or treatment arms
Special stains and matrix	Stained area percentage, collagen or matrix fraction	Fibrosis and chronic change in animal models
Slide-level quality	Focus and artefact flags, quantifiable tissue area	Which slides must be re-scanned or re-stained

From Glass Slide to Image-Derived Endpoint

The flow keeps pathologist decisions and automated processing separate, so every number traces back to a defined step or rule.

Digital Pathology Analysis Flow

Automated step Pathologist step

01
Slide Digitization and Scan QC

Slides are scanned at the agreed resolution and checked for focus, tissue coverage, and artefacts.

Output: whole slide images
02
Annotation and ROI Definition

A pathologist outlines tumour, lesion, or tissue compartments under a written rule set.

Output: ROI definitions
03
Automated Detection and Measurement

Validated algorithms classify nuclei or objects and compute marker, area, and density metrics.

Output: per-slide data
04
Pathologist Review Gate

Overlays and outlier slides are reviewed; ROIs are corrected or slides re-scanned.

Output: verified quantification
05
Image-Derived Endpoint Report

Metrics are aggregated per animal and group with the rules that produced them.

Output: endpoint tables

Documented return path: when the review gate rejects a slide for poor focus, missing tissue, or unusable staining, it returns to scanning or annotation instead of contributing a number.

What You Receive

✓Scan QC summary flagging slides that cannot be quantified
✓Written annotation rules and ROI definitions for the study
✓Per-slide and per-animal quantification tables in a reusable format
✓Overlay images showing what was detected and measured

Decision Guidance

Whole-section measurement and hotspot scoring answer different questions: whole-section analysis describes the lesion as a whole, while hotspot scoring concentrates on the most active region. BioVenic reports both where the study requires it, and records which rule produced each endpoint.

Published Data on Automated IHC Quantification in Canine Tumour Tissue

The figure shows a complete digital quantification chain for tumour tissue sections. It begins with slide preparation and whole slide imaging, continues with a pathologist outline of the tumour used as the region of interest, automated exclusion of non-tumour tissue and artefacts, and algorithm-based detection and classification of Ki-67-positive and Ki-67-negative nuclei by staining intensity. Automated output is then validated by a pathologist before the quantitative results enter statistical comparison, so a slide-level interpretation becomes a documented measurement chain rather than a single observer score.

The study assessed 309 canine mast cell tumours and reported the practical balance of deep-learning-assisted quantification: Ki-67 quantification itself was fully automated and rapid, whereas region-of-interest definition required frequent manual adjustment by a pathologist. Global digital metrics were compared with manual hotspot counts, and positive-cell density averaged roughly twice as high in hotspots as in whole sections, with differences up to 38-fold; exploratory survival analyses were also presented for the digitally derived parameters. That mix of automated measurement, written annotation rules, and pathologist quality control is the structure BioVenic applies when planning animal pathology image analysis, and it connects naturally to our veterinary immunohistochemistry assay development work when a marker must first be established on animal tissue.

Digital pathology analysis chain for animal tumour sections, from whole slide imaging and tumour region-of-interest annotation to automated nuclear Ki-67 classification with pathologist review. (OA Literature) — Fig.1 Comprehensive workflow for digital Ki-67 quantification in tumour tissue sections: slide preparation and whole slide imaging, pathologist-defined tumour region of interest, automated nuclei detection and intensity classification, and pathologist validation of the automated quantification. ^1,2

Our Advantages in Veterinary Digital Pathology Image Analysis

Image analysis delivers value only when the measurement rules are as carefully designed as the staining.

Study-Aware Analysis Design

Digitization, annotation, and endpoints are planned around your study groups.

Marker-Ready Quantification

IHC and immunofluorescence staining can be developed or optimized before measurement.

Transparent Annotation Rules

Every region-of-interest decision is documented and reused across the slide set.

Reviewable, Reusable Outputs

Quantitative tables, QC flags, and rule sets support independent review.

Connected Work

Related Veterinary Diagnostic and Model Development Services

Quantification is only as good as the tissue and staining behind it. These services usually run alongside an image analysis project.

Staining and Marker Detection

Veterinary Immunohistochemistry (IHC) Assay Development Establish and validate marker staining before quantification. Veterinary Immunofluorescence Assay (IFA) Development Add fluorescence readouts and multiplex channels to the imaging plan.

Animal Models and Study Material

Animal Models for Veterinary Infection Diseases Generate infection model tissue with defined groups and timepoints. Animal Model Development Build the animal model whose endpoints the image data will support.

Frequently Asked Questions About Veterinary Digital Pathology

It means digitizing animal tissue slides and measuring their structures with software instead of by eye. The whole slide image is annotated under written rules, algorithms detect and classify the objects of interest, and the results become quantitative endpoints such as positive-cell density, labelling index, lesion area fraction, or H-score.