Narrative review on AI and digital pathology for breast and gynecologic cancers

Photo by Logan Voss / Unsplash

Frontiers in Medicine Published April 24, 2026 Medically reviewed April 24, 2026 DOI ↗ By Dr. Julia Lee, PhD · Oncology, Genomics & Drug Development

Key Takeaway

Consider the potential of AI in pathology for breast and gynecologic cancers, but note significant standardization and workflow barriers.

This is a narrative review that synthesizes current evidence on the integration of digital and molecular pathology with artificial intelligence (AI) and machine learning (ML) for patients with breast and gynecologic cancers. The scope includes potential applications for assessing treatment response, survival, tumor heterogeneity, and interactions between tumor cells, stroma, and immune cells.

The authors argue that multimodal data integration could enhance diagnostic and prognostic capabilities. However, the review does not report pooled effect sizes or specific quantitative findings, as it is a qualitative synthesis rather than a meta-analysis.

Key limitations noted by the authors include challenges with standardization, reproducibility, regulation, and workflow integration. The review does not report specific study populations, interventions, comparators, or adverse events.

Practice relevance is framed around the clinical adoption of multimodal data, but the authors emphasize that current barriers must be addressed before widespread implementation. The evidence presented is preliminary and should be interpreted with caution.

Study Details

Study typeSystematic review

EvidenceLevel 1

PublishedApr 2026

View Original Abstract ↓

Breast and gynecologic cancers consist of two groups of complex solid tumors, each with unique genomic features, immune microenvironments, and treatment responses. Recent advances in next-generation sequencing, spatial profiling, and digital pathology have transformed diagnostic methods, enabling seamless integration of morphological and molecular data. Artificial intelligence (AI) and machine learning (ML) are now essential tools for linking histomorphology, immunophenotype, and molecular alterations in ways that were previously unachievable. This review discusses recent progress in integrating digital and molecular pathology for these cancers, with an emphasis on practical clinical applications. We highlight emerging research in breast, endometrial, ovarian, and cervical cancers, where combined image-based and molecular approaches can predict treatment response and survival. Additionally, spatial transcriptomics and proteomics are deepening our understanding of tumor heterogeneity and the interactions between tumor cells, stroma, and immune cells that drive disease progression. We also address current challenges, such as standardization, reproducibility, regulation, and workflow integration, and propose priorities to facilitate the clinical adoption of multimodal data.