Tumor habitat imaging and the feasibility of virtual biopsy

Tumors are not uniform entities but mosaics of cellular and molecular niches. Tumor habitat imaging, the partitioning of lesions into spatially distinct subregions on medical images offers a powerful route to quantify intra-tumoral heterogeneity and move towards a virtual biopsy paradigm. Instead of sampling a single core of tissue, imaging enables non-invasive, whole-tumor, and repeatable assessment of phenotypic diversity.This Research Topic aims to collect methodological innovations, translational studies, and clinical perspectives on habitat imaging pipelines and their integration into oncology workflows. The focus is specifically on tumor habitats (i.e. explicitly spatial clustering into sub-regions, habitat maps) rather than more generic radiomics or AI for medical imaging & pathology. We invite original papers and datasets that explore:• Biological attribution and validation: radiogenomic and radiopathomic correlation, fusion-guided biopsies, digital pathology alignment, and spatial multi-omics integration, including imaging-pathology co-localisation. Studies that co-register PET/MR/CT habitats with digital pathology, spatial transcriptomics, and multiplex imaging, including fusion-guided biopsies and surgical co-localization.• Computational pipelines: unsupervised and semi-supervised clustering, self-supervised learning, AI agents, deep learning and foundation-model embeddings, and hybrid radiomics–AI approaches for habitat segmentation. Studies extending existing AI/radiomics topics by focusing on spatially resolved phenotypes and actionable outputs for clinical decision-making.• Clinical translation: prognostication, early response monitoring, risk stratification, biopsy guidance, habitat-guided radiotherapy, spatial dose painting, adaptive therapy and dosage, as well as incorporation into PACS and planning systems. Studies focusing on habitat-guided biopsy and dose escalation, and post-treatment assessment.• Cross-cutting challenges: standardization (compliant reporting, harmonization across scanners), reproducibility and multi-center validation, uncertainty quantification, explainability, and regulatory considerations for software as a medical device. Studies focusing on methods that make habitat imaging reliable and regulatory-ready.By bringing together engineers, radiologists, nuclear medicine specialists, oncologists, pathologists, and data scientists, this research topic will showcase how spatially resolved, imaging-driven phenotypes can complement pathology and accelerate the adoption of virtual biopsy in precision oncology. Contributions highlighting multimodal fusion (e.g., simultaneous PET/SPECT-CT/MRI), spatial multi-omics co-registration and workflow studies are particularly encouraged. AI methodologies and techniques should play a central role in all papers considered for acceptance. Topic editors: Rajat Vashistha, Viktor Vegh Source: https://www.frontiersin.org/research-topics/74798/tumor-habitat-imaging-and-the-feasibility-of-virtual-biopsy