Emerging Concern of Scientific Fraud: Deep Learning and Image Manipulation
Open accessAbstract
Scientific fraud by image duplications and manipulations within western blot images is a rising problem. Currently, problematic western blot images are mainly detected by checking repeated bands or through visual observation. However, the completeness of the above methods in detecting problematic images has not been demonstrated. Here we show that Generative Adversarial Nets (GANs) can generate realistic western blot images that indistinguishable from real western blots. The overall accuracy of researchers for identifying synthetic western blot images is 0.52, which almost equal to blind guess (0.5). We found that GANs can generate western blot images with bands of the expected lengths, widths, and angles in desired positions that can fool researchers. For the case study, we find that the accuracy of detecting the synthetic western blot images is related to years of researchers performed studies relevant to western blots, but there was no apparent difference in accuracy among researchers with different academic degrees. Our results demonstrate that GANs can generate fake western blot images to fool existing problematic image detection methods. Therefore, more information is needed to ensure that the western blots appearing in scientific articles are real. We argue to require every western blot image to be uploaded along with a unique identifier generated by the laboratory machine and to peer review these images along with the corresponding submitted articles, which may reduce the incidence of scientific fraud.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
If a western blot image has been synthetically generated, either entirely or locally, there is essentially no hope to spot such traces by visual inspection.
“If the western blot image under analysis has been synthetically generated, either in its entirety or locally only in specific pixel regions, there would be essentially no hope to spot such traces by visual inspection [10].”
From Forensic Analysis of Synthetically Generated Western Blot Images (2022) · cites this paper as [10]
The indexed Abstract reports that researchers identified synthetic western blots with 0.52 accuracy, close to guessing, supporting the point that entirely generated blots are difficult to detect visually. It does not examine blots generated only in specific pixel regions, so the claim's extension to local generation is unsupported. (LLM, conf=0.85)
Experimental — accuracy not yet measuredStandard GAN-based image generation techniques can synthesize western blots that are almost indistinguishable from real ones, even to experts' eyes.
“Indeed, during some preliminary experiments, the authors of [10] verified that standard image generation techniques based on GANs [14], [15] can synthesize almost indistinguishable western blots with respect to the real ones, even at the experts' eyes.”
From Forensic Analysis of Synthetically Generated Western Blot Images (2022) · cites this paper as [10]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
The citation lists haven’t been fetched yet
References 13 · Cited by 21
Research works printed but not linked: 11
- Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proceedings of the National Academy of Sciences USA. 1979, 76 (9): 4350-54.
- Goodfellow I, Pouget-Abadie J, Mirza M, et al. Generative adversarial nets. Advances in neural information processing systems. 2014: 2672-2680.
- Yu J, Lin Z, Yang J, et al. Generative image inpainting with contextual attention. Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 5505-5514.
- Reed S, Akata Z, Yan X, et al. Generative adversarial text to image synthesis. arXiv preprint arXiv:1605.05396, 2016. · arXiv
- Zhu J Y, Park T, Isola P, et al. Unpaired image-to-image translation using cycle-consistent adversarial networks. Proceedings of the IEEE international conference on computer vision. 2017: 2223-2232.
- Iizuka S, Simo-Serra E, Ishikawa H. Globally and locally consistent image completion. ACM Transactions on Graphics, 2017, 36(4): 1- 14.
- You C, Li G, Zhang Y, et al. CT super-resolution GAN constrained by the identical, residual, and cycle learning ensemble. IEEE transactions on medical imaging, 2019, 39(1): 188-203.
- Bradski G, Kaehler A. Learning OpenCV: Computer vision with the OpenCV library. " O'Reilly Media, Inc.", 2008.
- Mahmood T, Yang P C. Western blot: technique, theory, and trouble shooting. North American journal of medical sciences, 2012, 4(9): 429.
- Isola P, Zhu J Y, Zhou T, et al. Image-to-image translation with conditional adversarial networks. Proceedings of the IEEE conference on computer vision and pattern recognition. 2017: 1125- 1134.
- Park T, Liu M Y, Wang T C, et al. Semantic image synthesis with spatially-adaptive normalization. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2019: 2337-2346. Fig. 2. Results of External Evaluation.
- Authors
- Qi, C., Zhang, J., & Luo, P.
- Affiliations
- Zhujiang Hospital (CN)
- Corresponding author
- Luo, Peng
- ORCID
- Venue
- bioRxiv (Cold Spring Harbor Laboratory)
- Publisher
- openRxiv
- Published
- 2020-11-24
- OpenAlex
- W3110153736
- Cited by
- 21
- References
- 13
- Citations by year
- 2025: 3 · 2024: 7 · 2023: 3 · 2022: 4 · 2021: 4
- Type
- posted content
- Language
- en
- Open access
- green
- doi.org
- Locations
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion)
- Topics
- Social Sciences › Safety Research › Academic integrity and plagiarism
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Physical Sciences › Mathematics › Statistics and Probability › Benford’s Law and Fraud Detection
- Keywords
- scientific fraud, image editing, Western blot images, generative adversarial networks, synthetic image generation, image duplication, research integrity
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
