{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,1,29]],"date-time":"2026-01-29T21:14:54Z","timestamp":1769721294721,"version":"3.49.0"},"reference-count":35,"publisher":"MDPI AG","issue":"2","license":[{"start":{"date-parts":[[2026,1,29]],"date-time":"2026-01-29T00:00:00Z","timestamp":1769644800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Universidad Militar Nueva Granada-Vicerrector\u00eda de Investigaciones","award":["INV-ING-4157"],"award-info":[{"award-number":["INV-ING-4157"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Information"],"abstract":"<jats:p>Distribution shifts commonly arise in real-world machine learning scenarios in which the fundamental assumption that training and test data are drawn from independent and identically distributed samples is violated. In the case of medical data, such distribution shifts often occur during data acquisition and pose a significant challenge to the robustness and reliability of artificial intelligence systems in clinical practice. Additionally, quantifying these shifts without training a model remains a key open problem. This paper proposes a comprehensive methodological framework for evaluating the impact of such shifts on medical image datasets under artificial transformations that simulate acquisition variations, leveraging the Cumulative Spectral Gradient (CSG) score as a measure of multiclass classification complexity induced by distributional changes. Building on prior work, the proposed approach is meaningfully extended to twelve 2D medical imaging benchmarks from the MedMNIST collection, covering both binary and multiclass tasks, as well as grayscale and RGB modalities. We evaluate the metric analyzing its robustness to clinically inspired distribution shifts that are systematically simulated through motion blur, additive noise, brightness and contrast variation, and sharpness variation, each applied at three severity levels. This results in a large-scale benchmark that enables a detailed analysis of how dataset characteristics, transformation types, and distortion severity influence distribution shifts. Thus, the findings show that while the metric remains generally stable under noise and focus distortions, it is highly sensitive to variations in brightness and contrast. On the other hand, the proposed methodology is compared against Cleanlab\u2019s widely used Non-IID score on the RetinaMNIST dataset using a pre-trained ResNet-50 model, including both class-wise analysis and correlation assessment between metrics. Finally, interpretability is incorporated through class activation map analysis on BloodMNIST and its corrupted variants to support and contextualize the quantitative findings.<\/jats:p>","DOI":"10.3390\/info17020125","type":"journal-article","created":{"date-parts":[[2026,1,29]],"date-time":"2026-01-29T10:08:57Z","timestamp":1769681337000},"page":"125","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Complexity-Driven Adversarial Validation for Corrupted Medical Imaging Data"],"prefix":"10.3390","volume":"17","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8073-3594","authenticated-orcid":false,"given":"Diego","family":"Renza","sequence":"first","affiliation":[{"name":"Facultad de Ingenier\u00eda, Universidad Militar Nueva Granada, Carrera 11 101-80, Bogota 110111, Colombia"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5430-8778","authenticated-orcid":false,"given":"Jorge","family":"Brieva","sequence":"additional","affiliation":[{"name":"Facultad de Ingenier\u00eda, Universidad Panamericana, Augusto Rodin 498, Mexico City 03920, Mexico"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9637-786X","authenticated-orcid":false,"given":"Ernesto","family":"Moya-Albor","sequence":"additional","affiliation":[{"name":"Facultad de Ingenier\u00eda, Universidad Panamericana, Augusto Rodin 498, Mexico City 03920, Mexico"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,1,29]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","unstructured":"Clement David-Olawade, A., Olawade, D.B., Vanderbloemen, L., Rotifa, O.B., Fidelis, S.C., Egbon, E., Akpan, A.O., Adeleke, S., Ghose, A., and Boussios, S. (2025). AI-Driven Advances in Low-Dose Imaging and Enhancement\u2014A Review. Diagnostics, 15.","DOI":"10.3390\/diagnostics15060689"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"43035","DOI":"10.1007\/s11042-023-17326-1","article-title":"Recent trend in medical imaging modalities and their applications in disease diagnosis: A review","volume":"83","author":"Abhisheka","year":"2024","journal-title":"Multimed. Tools Appl."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"28","DOI":"10.1186\/s41747-025-00562-5","article-title":"AI for image quality and patient safety in CT and MRI","volume":"9","author":"Melazzini","year":"2025","journal-title":"Eur. Radiol. Exp."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"651","DOI":"10.1097\/CM9.0000000000003489","article-title":"Artificial intelligence in medical imaging: From task-specific models to large-scale foundation models","volume":"138","author":"Bian","year":"2025","journal-title":"Chin. Med. J."},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Jim\u00e9nez-S\u00e1nchez, A., Avlona, N., Boer, S., Campello, V., Feragen, A., Ferrante, E., Ganz, M., Gichoya, J., Gonzalez, C., and Groefsema, S. (2025). In the picture: Medical imaging datasets, artifacts, and their living review. Proceedings of the 2025 ACM Conference on Fairness, Accountability, and Transparency (FAccT \u201925), Athens, Greece, 23\u201326 June 2025, Association for Computing Machinery.","DOI":"10.1145\/3715275.3732035"},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"eabq6147","DOI":"10.1126\/sciadv.abq6147","article-title":"Disparities in dermatology AI performance on a diverse, curated clinical image set","volume":"8","author":"Daneshjou","year":"2022","journal-title":"Sci. Adv."},{"key":"ref_7","unstructured":"Yao, H., Choi, C., Cao, B., Lee, Y., Koh, P., and Finn, C. (December, January 28). Wild-time: A benchmark of in-the-wild distribution shift over time. Proceedings of the 36th International Conference on Neural Information Processing Systems, New Orleans, LA, USA."},{"key":"ref_8","unstructured":"Zhang, H., Singh, H., Ghassemi, M., and Joshi, S. (2023, January 23\u201329). \u201cWhy did the model fail?\u201d: Attributing model performance changes to distribution shifts. Proceedings of the 40th International Conference on Machine Learning, Honolulu, HI, USA."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Kim, T., Park, S., Lim, S., Jung, Y., Muandet, K., and Song, K. (2025, January 11\u201315). Sufficient Invariant Learning for Distribution Shift. Proceedings of the 2025 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.","DOI":"10.1109\/CVPR52734.2025.00467"},{"key":"ref_10","unstructured":"Taori, R., Dave, A., Shankar, V., Carlini, N., Recht, B., and Schmidt, L. (2020, January 6\u201312). Measuring robustness to natural distribution shifts in image classification. Proceedings of the 34th International Conference on Neural Information Processing Systems, Vancouver, BC, Canada."},{"key":"ref_11","unstructured":"Izzo, Z., Ying, L., and Zou, J. (2021, January 18\u201324). How to learn when data reacts to your model: Performative gradient descent. Proceedings of the 38th International Conference on Machine Learning, Online."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Renza, D., Moya-Albor, E., and Chavarro, A. (2024). Adversarial Validation in Image Classification Datasets by Means of Cumulative Spectral Gradient. Algorithms, 17.","DOI":"10.3390\/a17110531"},{"key":"ref_13","unstructured":"Koh, P., Sagawa, S., Marklund, H., Xie, S., Zhang, M., Balsubramani, A., Hu, W., Yasunaga, M., Phillips, R., and Gao, I. (2021, January 18\u201324). WILDS: A benchmark of in-the-wild distribution shifts. Proceedings of the 38th International Conference on Machine Learning, Online."},{"key":"ref_14","unstructured":"Chen, A., Lee, Y., Setlur, A., Levine, S., and Finn, C. (2023). Confidence-based model selection: When to take shortcuts for subpopulation shifts. arXiv."},{"key":"ref_15","first-page":"1373","article-title":"Confident Learning: Estimating Uncertainty in Dataset Labels","volume":"70","author":"Northcutt","year":"2021","journal-title":"J. Artif. Intell. Res. JAIR"},{"key":"ref_16","unstructured":"Yao, H., Wang, Y., Li, S., Zhang, L., Liang, W., Zou, J., and Finn, C. (2022, January 17\u201323). Improving out-of-distribution robustness via selective augmentation. Proceedings of the 39th International Conference on Machine Learning, Baltimore, MD, USA."},{"key":"ref_17","unstructured":"Liang, W., Yang, X., and Zou, J. (2022). Metashift: A dataset of datasets for evaluating contextual distribution shifts. arXiv."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"\u00d6zt\u00fcrk, \u015e., Duran, O., and \u00c7ukur, T. (2025). DenoMamba: A fused state-space model for low-dose CT denoising. IEEE J. Biomed. Health Inform.","DOI":"10.1109\/JBHI.2025.3629034"},{"key":"ref_19","unstructured":"Kabas, B., Arslan, F., Nezhad, V., Ozturk, S., Saritas, E., and \u00c7ukur, T. (2025). Physics-Driven Autoregressive State Space Models for Medical Image Reconstruction. arXiv."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"112451","DOI":"10.1016\/j.patcog.2025.112451","article-title":"Wavelet-based physically guided normalization network for real-time traffic dehazing","volume":"172","author":"Zhang","year":"2026","journal-title":"Pattern Recognit."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"10189","DOI":"10.1109\/TITS.2025.3550267","article-title":"VNDHR: Variational Single Nighttime Image Dehazing for Enhancing Visibility in Intelligent Transportation Systems via Hybrid Regularization","volume":"26","author":"Liu","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"ref_22","doi-asserted-by":"crossref","unstructured":"Aguilera-Gonz\u00e1lez, S., Renza, D., and Moya-Albor, E. (2024, January 13\u201315). Evaluation of Dataset Distribution in Biomedical Image Classification Against Image Acquisition Distortions. Proceedings of the 20th International Symposium on Medical Information Processing and Analysis (SIPAIM), Antigua, Guatemala.","DOI":"10.1109\/SIPAIM62974.2024.10783583"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Branchaud-Charron, F., Achkar, A., and Jodoin, P. (2019, January 15\u201320). Spectral Metric for Dataset Complexity Assessment. Proceedings of the 2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA.","DOI":"10.1109\/CVPR.2019.00333"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Gonz\u00e1lez-Santoyo, C., Renza, D., and Moya-Albor, E. (2025). Identifying and Mitigating Label Noise in Deep Learning for Image Classification. Technologies, 13.","DOI":"10.3390\/technologies13040132"},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1038\/s41597-022-01721-8","article-title":"MedMNIST v2-A large-scale lightweight benchmark for 2D and 3D biomedical image classification","volume":"10","author":"Yang","year":"2023","journal-title":"Sci. Data"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Yang, J., Shi, R., and Ni, B. (2021, January 13\u201316). MedMNIST Classification Decathlon: A Lightweight AutoML Benchmark for Medical Image Analysis. Proceedings of the IEEE 18th International Symposium on Biomedical Imaging (ISBI), Nice, France.","DOI":"10.1109\/ISBI48211.2021.9434062"},{"key":"ref_27","unstructured":"Cummings, J., Snorrason, E., and Mueller, J. (2023, January 23\u201329). Detecting Dataset Drift and Non-IID Sampling via k-Nearest Neighbors. Proceedings of the ICML 2023 Data-centric Machine Learning Research Workshop, Honolulu, HI, USA."},{"key":"ref_28","unstructured":"Wightman, R., Touvron, H., and J\u00e9gou, H. (2021). Resnet strikes back: An improved training procedure in timm. arXiv."},{"key":"ref_29","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27\u201330). Deep Residual Learning for Image Recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"35","DOI":"10.2152\/jmi.66.35","article-title":"Standardization of imaging features for radiomics analysis","volume":"66","author":"Haga","year":"2019","journal-title":"J. Med. Investig."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"9","DOI":"10.1016\/j.nicl.2014.08.008","article-title":"Statistical normalization techniques for magnetic resonance imaging","volume":"6","author":"Shinohara","year":"2014","journal-title":"Neuroimage Clin."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"193","DOI":"10.1007\/s12021-024-09655-9","article-title":"Deepn4: Learning N4ITK bias field correction for T1-weighted images","volume":"22","author":"Kanakaraj","year":"2024","journal-title":"Neuroinformatics"},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Xu, C., Sun, Y., Zhang, Y., Liu, T., Wang, X., Hu, D., Huang, S., Li, J., Zhang, F., and Li, G. (2025). Stain Normalization of Histopathological Images Based on Deep Learning: A Review. Diagnostics, 15.","DOI":"10.3390\/diagnostics15081032"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Bhati, D., Neha, F., and Amiruzzaman, M. (2024). A Survey on Explainable Artificial Intelligence (XAI) Techniques for Visualizing Deep Learning Models in Medical Imaging. J. Imaging, 10.","DOI":"10.20944\/preprints202408.0765.v1"},{"key":"ref_35","doi-asserted-by":"crossref","unstructured":"Chavarro, A., Renza, D., and Moya-Albor, E. (2024). ConvNext as a Basis for Interpretability in Coffee Leaf Rust Classification. Mathematics, 12.","DOI":"10.3390\/math12172668"}],"container-title":["Information"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2078-2489\/17\/2\/125\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,29]],"date-time":"2026-01-29T10:20:50Z","timestamp":1769682050000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2078-2489\/17\/2\/125"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,1,29]]},"references-count":35,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,2]]}},"alternative-id":["info17020125"],"URL":"https:\/\/doi.org\/10.3390\/info17020125","relation":{},"ISSN":["2078-2489"],"issn-type":[{"value":"2078-2489","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,1,29]]}}}