The PLOS One Editors published an Editorial Note to inform readers about issues identified after publication of the article titled “Brain MRI detection and classification: Harnessing convolutional neural networks and multi-level thresholding.” The note summarizes two categories of post-publication concerns: (1) the presence of cited works that had been retracted prior to the article’s publication, and (2) an incorrect Data Availability statement that required correction.
PLOS identified that three works cited in the article, listed as References 12, 19, and 20, had been retracted before the article’s publication date. The Editors state that these references are not considered crucial to supporting the research findings or the article’s conclusions. However, the Editorial Note specifies that because those sources were retracted, certain sentences in the article’s Literature Review are no longer supported by the cited literature.
Two specific sentences are named as unsupported: the first sentence of the third paragraph of the Literature Review (which described “An augmentation-based 2D convolutional neural network (CNN) system … proposed by Chanu et al.”) and the second sentence of the fifth paragraph (which described a technique by Arpit Kumar Sharma et al. based on a modified ResNet50 architecture and an enhanced watershed algorithm). PLOS indicates these statements rely on references that had been retracted prior to the article’s publication, and therefore those particular literature claims should not be considered supported by the cited sources.
The Editorial Note corrects the article’s Data Availability statement. The updated statement indicates that the data underlying the study are publicly available from the Harvard Whole Brain Atlas (http://www.med.harvard.edu/AANLIB/). The Editors emphasize that the authors did not receive any special access privileges and that all data are accessible to other researchers under the same conditions.
The note further states that the specific Whole Brain Atlas cases used in the study are listed in the article’s Supporting Information file designated as S1 Table. PLOS provides a link to that supplementary file for readers who wish to see the exact cases included in the analysis.
PLOS explicitly reports that, in its assessment, the retracted references are not critical for supporting the research methods, primary results, or conclusions reported in the article. Nonetheless, the Editorial Note documents that particular sentences in the Literature Review are unsupported because they cite retracted work. The note does not assert that the main findings or conclusions are invalid; instead, it clarifies the limited scope of the problem (unsupported literature statements) and documents the corrected data availability information.
The Editorial Note includes publication metadata: it was published by The PLOS One Editors and is assigned a DOI. The note reproduces the reference to the original article (Kamireddy RR et al., Brain MRI detection and classification: Harnessing convolutional neural networks and multi-level thresholding, PLoS One. 2024;19(8):e0306492) and provides links to view the original article, the supporting information (including S1 Table), and article metrics and comments. The Editors also provide the formal citation for the Editorial Note and note that it is distributed under the Creative Commons Attribution License.
Readers seeking the corrected Data Availability details or the list of specific Whole Brain Atlas cases are directed to the Supporting Information file labeled S1 Table, available from the article page. The Editorial Note links to the Harvard Whole Brain Atlas as the public repository for the underlying data and clarifies that no special access was granted to the authors. For full context, readers can consult the Editorial Note alongside the original article and its supplementary materials.
This Editorial Note documents the Editors’ post-publication actions: identifying that three cited works had been retracted prior to publication, naming the limited Literature Review statements that are no longer supported, and issuing a corrected Data Availability statement pointing to the Harvard Whole Brain Atlas and the article’s S1 Table. The note serves to inform readers about these specific issues and to provide links to the corrected and supporting information. If readers require further details beyond what is reported in the Editorial Note, those details are not provided in the note itself and would require consultation of the original article or its supplementary files.