Related Experiment Video
Updated: Jan 18, 2026
Label-Free Identification of Lymphocyte Subtypes Using Three-Dimensional Quantitative Phase Imaging and Machine Learning
Published on: November 19, 2018
Deep-learning based morphological segmentation of canine diffuse large B-cell lymphoma
Kenneth Ancheta1, Androniki Psifidi2, Andrew D Yale2
1Pathobiology and Population Science, Royal Veterinary College, Hatfield, United Kingdom.
Insights
A new AI tool, HawksheadNet, accurately distinguishes canine diffuse large B-cell lymphoma (cDLBCL) from benign conditions using whole slide images. This convolutional neural network (CNN) approach aids veterinary diagnostics, improving accuracy and efficiency in lymphoma detection.
Area of Science:
- Veterinary Pathology
- Computational Pathology
- Artificial Intelligence in Medicine
Background:
- Diffuse large B-cell lymphoma (DLBCL) is a common cancer in humans and dogs.
- Canine DLBCL (cDLBCL) is aggressive, and current diagnosis relies on time-consuming histopathology.
- There's a need for faster, more accurate diagnostic tools in veterinary medicine.
Purpose of the Study:
- To develop and evaluate a convolutional neural network (CNN) for differentiating cDLBCL from reactive lymphoid hyperplasia (RLH) in canine lymph node biopsies.
- To introduce HawksheadNet, a novel CNN architecture for cancer image classification.
- To assess the impact of stain normalization on CNN performance.
Main Methods:
- Whole slide images (WSIs) of H&E stained canine lymph nodes were digitized.
- A modified Aachen protocol was used for image pre-processing.
- HawksheadNet, a lightweight CNN, was trained and fine-tuned on a dataset split into training, validation, and testing sets.
- Stain normalization using StainNet was applied.
Main Results:
- HawksheadNet achieved a high area under the receiver operating characteristic (AUROC) of 0.9691 for differentiating cDLBCL from RLH on StainNet-normalized images.
- The CNN outperformed other pre-trained models like EfficientNet, Inception, and MobileNet.
- WSI segmentation using tile-wise predictions provided visual diagnostic aids aligned with pathologist interpretations.
Conclusions:
- Convolutional neural networks, particularly HawksheadNet, show significant potential for accurate cDLBCL diagnosis from WSIs.
- Stain normalization is crucial for optimizing CNN performance in veterinary cancer image analysis.
- This AI-driven approach can enhance veterinary diagnostic workflows, potentially improving patient care and prognostication.
Abstract:
Diffuse large B-cell lymphoma is the most common type of non-Hodgkin lymphoma (NHL) in humans, accounting for about 30-40% of NHL cases worldwide. Canine diffuse large B-cell lymphoma (cDLBCL) is the most common lymphoma subtype in dogs and demonstrates an aggressive biologic behaviour. For tissue biopsies, current confirmatory diagnostic approaches for enlarged lymph nodes rely on expert histopathological assessment, which is time-consuming and requires specialist expertise. Therefore, there is an urgent need to develop tools to support and improve veterinary diagnostic workflows. Advances in molecular and computational approaches have opened new avenues for morphological analysis. This study explores the use of convolutional neural networks (CNNs) to differentiate cDLBCL from non-neoplastic lymph nodes, specifically reactive lymphoid hyperplasia (RLH). Whole slide images (WSIs) of haematoxylin-eosin stained lymph node slides were digitised at 20 × magnification and pre-processed using a modified Aachen protocol. Extracted images were split at the patient level into training (60%), validation (30%), and testing (10%) datasets. Here, we introduce HawksheadNet, a novel lightweight CNN architecture for cancer image classification and highlight the critical role of stain normalisation in CNN training. Once fine-tuned, HawksheadNet demonstrated strong generalisation performance in differentiating cDLBCL from RLH, achieving an area under the receiver operating characteristic (AUROC) of up to 0.9691 using fine-tuned parameters on StainNet-normalised images, outperforming pre-trained CNNs such as EfficientNet (up to 0.9492), Inception (up to 0.9311), and MobileNet (up to 0.9498). Additionally, WSI segmentation was achieved by overlaying the tile-wise predictions onto the original slide, providing a visual representation of the diagnosis that closely aligned with pathologist interpretation. Overall, this study highlights the potential of CNNs in cancer image analysis, offering promising advancements for clinical pathology workflows, patient care, and prognostication.
