Nuclei classification is a mandatory process to obtain scoring information for whole slide images. In immunohistochemistry (IHC) staining specifically for estrogen-receptor (ER) biomarker, an Allred score based on the proportion and intensity of cancer nuclear staining is widely used in histopathology practice to predict response to hormonal treatment. This manually exhaustive process can be accelerated with the help of computational intelligence. In this article, we present a thorough analysis of 37 whole slide images of breast cancer cases with over 2.8 million segmented nuclei. ER-stained nuclei were classified into negative, weak, moderate and strong intensities using DenseNet deep learning architecture, contributing to Allred scoring. Seven different models and configurations were exhaustively analysed in six tests to obtain the scoring reaching best concordance of 56.8% and 81.1% with the pathologist’s manual score and suggested hormonal treatment. We also discussed in detail the causes that lead to the non-concordances. This study follows the pathologists’ workflow in obtaining the Allred score but is fully automated. It provides a basis for the development of more complex deep learning models, particularly for nuclei classification and achieving accurate scoring of ER-IHC stained whole slide images.

The HER2-IHC-40x and HER2-IHC-40x-WSI datasets are high-resolution whole slide image (WSI) and patch-extracted region collection for HER2 immunohistochemistry (IHC) scoring in breast cancer pathology. 107 WSIs are scanned at 40 × magnification with Regions of Interest (ROIs) annotated by expert pathologists. Patches of 1024 × 1024 pixels are extracted from the ROIs and classified into four HER2 scores (0, 1+, 2+, 3+), yielding structured data for computational pathology analysis. There were two strategies of splitting: WSI-based split, where data was first split before extracting the patches and named as HER2-IHC-40x for this dataset, the other one is patch-based split, where patches were extracted first and then split, named as HER2-IHC-40x-WSI of this dataset. The filtering method for color histograms was applied to remove the non-tumour regions and artifacts, generating high-quality data. The dataset is applicable to deep learning applications, including HER2 classification and explainable AI. It is freely available on Zenodo, with preprocessing scripts provided via GitHub, enabling reproducibility in digital pathology research.

Background/Objective: The integration of machine learning (ML) and deep learning (DL) has significantly enhanced medical image classification, especially in histopathology, by improving diagnostic accuracy and aiding clinical decision making. However, data privacy concerns and restrictions on sharing patient data limit the development of effective DL models. Federated learning (FL) offers a promising solution by enabling collaborative model training across institutions without exposing sensitive data. This systematic review aims to comprehensively evaluate the current state of FL applications in histopathological image classification by identifying prevailing methodologies, datasets, and performance metrics and highlighting existing challenges and future research directions. Methods: Following PRISMA guidelines, 24 studies published between 2020 and 2025 were analyzed. The literature was retrieved from ScienceDirect, IEEE Xplore, MDPI, Springer Nature Link, PubMed, and arXiv. Eligible studies focused on FL-based deep learning models for histopathology image classification with reported performance metrics. Studies unrelated to FL in histopathology or lacking accessible full texts were excluded. Results: The included studies utilized 10 datasets (8 public, 1 private, and 1 unspecified) and reported classification accuracies ranging from 69.37% to 99.72%. FedAvg was the most commonly used aggregation algorithm (14 studies), followed by FedProx, FedDropoutAvg, and custom approaches. Only two studies reported their FL frameworks (Flower and OpenFL). Frequently employed model architectures included VGG, ResNet, DenseNet, and EfficientNet. Performance was typically evaluated using accuracy, precision, recall, and F1-score. Federated learning demonstrates strong potential for privacy-preserving digital pathology applications. However, key challenges remain, including communication overhead, computational demands, and inconsistent reporting standards. Addressing these issues is essential for broader clinical adoption. Conclusions: Future work should prioritize standardized evaluation protocols, efficient aggregation methods, model personalization, robustness, and interpretability, with validation across multi-institutional clinical environments to fully realize the benefits of FL in histopathological image classification.

Cardiovascular diseases (CVDs), particularly coronary artery disease (CAD), remain the leading cause of global mortality, necessitating advanced diagnostic solutions. Accurate deformable image registration plays a crucial role in enhancing segmentation precision and classification performance in cardiovascular imaging. However, existing registration methods, including VoxelMorph, face limitations in computational efficiency and memory usage, restricting their real-time applicability for high-resolution cardiac imaging. This study proposes the Mamba-Optimized VoxelMorph framework, which leverages GPU-based parallelization and memory optimization to address these challenges. The framework achieves superior registration accuracy, yielding a Dice Similarity Coefficient (DSC) of 0.95 and Normalized Cross-Correlation (NCC) of 0.90, while reducing computational time by 40% and memory usage to 800 MB. These advancements ensure efficient alignment of complex cardiac structures, thereby improving segmentation accuracy and classification reliability. By addressing these critical limitations, the Mamba-Optimized VoxelMorph framework significantly enhances cardiovascular imaging, enabling precise, scalable, and real-time deformable image registration for improved CAD diagnosis and treatment planning.

Background: Breast cancer remains a critical health concern worldwide, with histopathological analysis of tissue biopsies serving as the clinical gold standard for diagnosis. Manual evaluation of histopathology images is time-intensive and requires specialized expertise, often resulting in variability in diagnostic outcomes. In silver in situ hybridization (SISH) images, accurate nuclei detection is essential for precise histo-scoring of HER2 gene expression, directly impacting treatment decisions. Methods: This study presents a scalable and automated deep learning framework for nuclei detection in HER2-SISH whole slide images (WSIs), utilizing a novel dataset of 100 expert-marked regions extracted from 20 WSIs collected at the University of Malaya Medical Center (UMMC). The proposed two-stage approach combines a pretrained Stardist model with image processing-based annotations, followed by fine tuning on our domain-specific dataset to improve generalization. Results: The fine-tuned model achieved substantial improvements over both the pretrained Stardist model and a conventional watershed segmentation baseline. Quantitatively, the proposed method attained an average F1-score of 98.1% for visual assessments and 97.4% for expert-marked nuclei, outperforming baseline methods across all metrics. Additionally, training and validation performance curves demonstrate stable model convergence over 100 epochs. Conclusions: These results highlight the robustness of our approach in handling the complex morphological characteristics of SISH-stained nuclei. Our framework supports pathologists by offering reliable, automated nuclei detection in HER2 scoring workflows, contributing to diagnostic consistency and efficiency in clinical pathology.