CellPrior-Net: Prior-Guided Nuclei Detection and Classification for H&E Whole-Slide Images
2026-07-01 • Multimedia
Multimedia
AI summaryⓘ
The authors developed CellPriorNet (CP Net), a fast and efficient method to find and identify cell nuclei in large pathology images used to study tumors. Their approach uses a simple neural network and focuses on a specific stain channel to better learn features of nuclei. They tested CP Net on various datasets with millions of nuclei and found it performed as well as other methods but much faster. Additionally, they created CellQuantNet, which adds a quality check to remove bad image areas before detecting and classifying cells. Both tools aim to help researchers analyze large pathology images more quickly and are available publicly.
nuclei detectionclassificationhematoxylin and eosin (H and E)whole-slide images (WSIs)convolutional neural networktumor microenvironmentartifact removalquality assessmentcomputational pathologyimage segmentation
Authors
Falah Jabar, Pasquale Lombardi, Aria Torkpour, Masoud Tafavvoghi, Per Niklas Benzler Waaler, Sigve Andersen, Erna-Elise Paulsen, Mette Pøhl, Lill-Tove Rasmussen Busund, Tom Donnem, Elin Richardsen, David J. Pinato, Mehrdad Rakaee
Abstract
Accurate nuclei detection and classification in hematoxylin and eosin (H and E) whole-slide images (WSIs) is a key task in computational pathology, particularly for quantitative analysis of the tumor microenvironment. However, this task remains highly challenging due to variations in nuclei morphology, staining procedures, scanners, organs, magnifications, and WSI artifacts. In addition, many existing pipelines rely on computationally demanding architectures and post-processing procedures, making gigapixel WSI analysis time consuming. In this work, CellPriorNet (CP Net) is proposed, an efficient nuclei detection and classification pipeline that utilizes a lightweight convolutional neural network architecture and hematoxylin (H) channel as prior information to enhance nuclei-aware feature learning. Extensive benchmarking was conducted against state of the art pipelines on 8 public and private datasets (total:10.4M nuclei) obtained from different organs, scanners, magnifications, and clinical centers. Experimental results demonstrate that CP Net achieves comparable performance while significantly reducing inference time. Furthermore, CellQuant Net was introduced, an end to end nuclei quantification pipeline, that integrates a quality assessment (QA) model to exclude regions with artifacts, followed by CP-Net cell detection and classification. The pipeline is publicly available on GitHub, and provides a potentially efficient and scalable framework for downstream computational pathology applications.