Advances in Deep Learning for Computational Pathology and Oncology Diagnostics

Computational pathology has recently emerged as a field that combines digital pathology, medical imaging, and artificial intelligence, offering new opportunities to improve cancer diagnosis. The increasing availability of data and images has enabled the application of deep learning models to large-scale datasets, allowing automated extraction of complex morphological patterns that may be difficult to quantify through conventional visual assessment. In recent years, advances in convolutional neural networks, attention-based architectures, and multi-task learning frameworks have demonstrated promising results in tasks such as tumor detection, cancer subtyping, and prognostic biomarker prediction. At the same time, the incorporation of explainable artificial intelligence techniques is becoming increasingly important to enhance the interpretability and reliability of these models. Nevertheless, important challenges remain, including data variability, model generalization, and clinical validation, highlighting the need for further research in this rapidly evolving field.Recent advances in digital pathology and artificial intelligence have opened new opportunities to improve cancer diagnosis and characterization through automated analysis of the data. Deep learning techniques have demonstrated strong potential in tasks such as tumor detection, tissue segmentation, cancer subtyping, and prediction of clinically relevant biomarkers. However, several challenges still limit the translation of these approaches into routine clinical practice. These include the need for robust models capable of generalizing across institutions, the effective analysis of extremely large sets of data, limited availability of annotated datasets, and the need for interpretable and trustworthy models that can support clinical decision-making.The goal of this Research Topic is to gather recent advances in deep learning methods for computational pathology and oncology diagnostics. In particular, we aim to highlight novel algorithms, methodological developments, and practical applications that improve the accuracy, robustness, and interpretability of AI-driven systems. Contributions addressing data-efficient learning, explainable AI, multimodal integration, and clinically relevant applications are especially encouraged. By bringing together interdisciplinary research from computer science, biomedical engineering, and pathology, this topic seeks to advance the development of reliable AI tools that can support pathologists and improve cancer diagnosis and patient outcomes.This Research Topic welcomes original research articles addressing recent advances in deep learning for computational pathology and oncology diagnostics. Submissions should focus on innovative approaches for the analysis of digital histopathology data and their potential impact on cancer detection, classification, and prognosis.Relevant themes include, but are not limited to:- Deep learning methods for medical image analysis- Tumor detection, classification, and grading in digital pathology- Weakly supervised, self-supervised, and data-efficient learning approaches- Multimodal learning combining pathology with clinical, genomic, or radiological data- Explainable and trustworthy AI for pathology applications- Benchmark datasets, evaluation frameworks, and reproducibility in computational pathology- Clinical translation and validation of AI systems for oncology diagnostics Topic editors: Juan Pedro Dominguez-Morales, Niccolò Marini Source: https://www.frontiersin.org/research-topics/83161/advances-in-deep-learning-for-computational-pathology-and-oncology-diagnostics