Validating Novel Imaging Algorithms for Breast Cancer: Small Cohort Studies Bridging Simulation and Clinical Practice

Breast cancer imaging has entered an era where novel algorithms, from deep learning classifiers to radiomics pipelines and reconstruction methods, are proposed faster than they can be clinically validated across established modalities such as mammography, MRI, ultrasound, and digital breast tomosynthesis. Many promising tools are developed and tested only in simulation, using digital phantoms or synthetic datasets, because large annotated clinical cohorts remain scarce, expensive, or restricted by privacy constraints. This creates a persistent gap: algorithms that perform well in silico often lack evidence of real-world reliability, while clinicians remain cautious about adopting tools validated on data that never touched a patient. Small cohort studies, when designed and reported rigorously, offer a practical bridge between these two worlds, providing an early, resource-realistic checkpoint before larger trials are justified.This Research Topic aims to consolidate research that takes novel breast imaging algorithms, whether developed for mammography, MRI, ultrasound, tomosynthesis, or other modalities, across that bridge, from simulation and phantom-based development into preliminary clinical testing. The goal is to give methodologically sound small-cohort validation work a dedicated, visible home, rather than leaving it scattered across broader oncology or engineering venues where its translational value is easy to overlook. We are especially interested in studies that are explicit about their validation pathway: what was tested in simulation, what changed when real patient data was introduced, and what that implies for scaling up.To capture the breadth and depth of current research, we welcome articles including but not limited to the following topics:- Phantom-based or in-silico validation pipelines for breast imaging algorithms across mammography, MRI, ultrasound, or tomosynthesis- Deep learning or radiomics models tested on small, well-characterized clinical cohorts- Methods for managing the simulation-to-clinic domain gap and synthetic-to-real data transfer- Modality-specific algorithm validation across mammography, MRI, ultrasound, or tomosynthesis- Reproducibility, generalizability, and reporting standards for small-cohort AI validation studiesStudies validated exclusively in simulation with no real-patient or phantom-to-clinic transition step are outside the scope of this collection.If your work moves a breast imaging algorithm even one step closer to clinical usability, this is the right home for it. We particularly welcome studies that report how algorithm performance shifted between development and validation stages, in mammography, MRI, ultrasound, tomosynthesis, or other emerging breast imaging modalities. We invite Original Research, Brief Research Reports, Methods papers, Study Protocols, and Technology and Code articles that document algorithm performance, validation design, and lessons learned along the simulation-to-clinic pathway. Topic editors: Jiabao He, Isabella Castellano, Wen Li Source: https://www.frontiersin.org/research-topics/85457/validating-novel-imaging-algorithms-for-breast-cancer-small-cohort-studies-bridging-simulation-and-clinical-practice