PLOS One Issues Expression of Concern Over Data Availability

The PLOS One Editors have issued an Expression of Concern regarding a previously published article titled “Nuclei segmentation and classification from histopathology images using federated learning for end-edge platform.” This official notice, published on October 30, 2026, highlights specific compliance issues related to the availability of the underlying data and source code necessary to verify the study’s results. The Editors determined that the individual-level data supporting the article’s findings were not provided as required by the journal’s Data Availability policy.

The core issue identified by the Editors is a discrepancy between the Data Availability statement in the original publication and the actual materials supplied. Although the authors provided information on third-party datasets used in the research, specifically referencing a dataset available on Kaggle, they failed to provide the specific data underlying the figures and graphs presented in the article. The Editors consider the provision of this specific underlying data to be necessary for full compliance with PLOS Data Availability standards, which mandate that individual-level data supporting results be made available.

In addition to the data availability concerns, the notice addresses the availability of the source code for the original article. The authors were initially unable to provide the source code used to generate the results. However, they have since supplied recreated code and made it publicly available through a GitHub repository and an archived release on Zenodo.

The Expression of Concern serves as a formal record of these issues, noting that the Editors regret that these matters were not addressed prior to the article’s publication.

Subject of Research: Data availability compliance issues in a PLOS One article on nuclei segmentation

Article Title: Expression of Concern: Nuclei segmentation and classification from histopathology images using federated learning for end-edge platform

Article References: The PLOS One Editors (2026). Expression of Concern: Nuclei segmentation and classification from histopathology images using federated learning for end-edge platform. PLOS One, 21(10), e0361682. https://doi.org/10.1371/journal.pone.0361682

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0361682

Keywords: PLOS One, Expression of Concern, Data Availability, Federated Learning, Histopathology, Nuclei Segmentation, Source Code, Scientific Publishing, Reproducibility, Expression, Concern, Nuclei