Work

An update of artificial intelligence in pathology: Current applications, challenges, and future directions

Year: 2026

Type: article

Abstract: Artificial intelligence (AI) is rapidly transforming the medical field, particularly digital pathology, by offering unprecedented capabilities for feature extraction and interpretation of whole slide images (WSIs). Deep learning (DL) networks, notably convolutional neural networks (CNNs), have demonstrated significant value in this domain. This review synthesises the current landscape of AI applications in pathology, including diagnostic assistance, classification, quantitative analysis, disease grading, and predictive modelling. Despite substantial progress, several major challenges persist, preventing AI from realising its full potential. Future directions should prioritise the development of more complex neural networks that deliver robustness, generalizability, and explainability. Promising advances include explainable AI (XAI) models, multi-modal networks, graph neural networks (GNNs), and weakly-supervised models. In particular, XAI may support model inspection and improve communication with users. Ultimately, this structured narrative review proposes that pathology and AI share a synergistic and reciprocal relationship: AI is not only revolutionising digital pathology, but the demanding requirements of pathologists are also driving innovation at the forefront of AI development.

Source: Intelligent Systems with Applications

DOI: 10.1016/j.iswa.2026.200736