Geographic Distribution of US Cohorts Used to Train Deep Learning Algorithms
Open accessAbstract
This study describes the US geographic distribution of patient cohorts used to train deep learning algorithms in published radiology, ophthalmology, dermatology, pathology, gastroenterology, and cardiology machine learning articles published in 2015-2019.
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?
In a review of 5 years of peer-reviewed articles describing AI algorithms for diagnostic tasks in patient care, 71% used data from California, Massachusetts, or New York.
“researchers analyzing 5 years of peer-reviewed articles describing AI algorithms for diagnostic tasks in patient care found that 71% used data from California, Massachusetts, or New York.”
From Managing a 'responsibility vacuum' in AI monitoring and governance in healthcare: a qualitative study (2025) · cites this paper as [15]
The Methods describe a review of articles published from 2015 through 2019 on image-based diagnostic deep-learning algorithms. The Results report that 40 of 56 studies with geographically identifiable US training cohorts (71%) used a cohort from California, Massachusetts, or New York. The claim omits that denominator: 71% does not apply to all 74 included studies. (LLM, conf=0.88)
Experimental — accuracy not yet measuredApproximately 60% of the reviewed articles only used data from California, Massachusetts, or New York, 34 states were completely unrepresented, and 13 states contributed limited data.
“Approximately 60% of the articles only used data from these three states, 34 states were completely unrepresented, and 13 states contributed limited data [15].”
From Managing a 'responsibility vacuum' in AI monitoring and governance in healthcare: a qualitative study (2025) · cites this paper as [15]
The Results report that 34 states contributed no patient cohorts and the other 13 contributed between 1 and 5 cohorts, supporting those parts of the claim. They report that 40 of 56 studies with geographically identifiable cohorts (71%) used a cohort from at least one of California, Massachusetts, or New York; they do not establish that approximately 60% of articles used data only from those three states. (LLM, conf=0.90)
Experimental — accuracy not yet measuredData for healthcare modeling is heavily skewed by geography.
“Data for healthcare modeling is also heavily skewed by geography.”
From Managing a 'responsibility vacuum' in AI monitoring and governance in healthcare: a qualitative study (2025) · cites this paper as [15]
The All of Us Research Program has enrolled about 400,000 diverse participants of the 1 million people planned across the USA, and is focused on enrolling individuals from broadly defined underrepresented groups in biomedical research.
“The All of Us Research Program 90 constitutes another novel and ambitious initiative by the NIH that has enrolled about 400,000 diverse participants of the 1 million people planned across the USA, and is focused on enrolling individuals from broadly defined underrepresented groups in biomedical research, which is especially needed in medical AI 91,92 .”
From Multimodal biomedical AI (2022) · cites this paper as [92]
A review of published AI articles between 2015 and 2019 in six disciplines found that data used for imaging AI models largely come from three states-California, Massachusetts, and New York-with none from 34 states.
“A review of published AI articles between 2015 and 2019 in six disciplines found that data used for imaging AI models largely come from three states—California, Massachusetts, and New York—with none from 34 states. 22”
From AI pitfalls and what not to do: mitigating bias in AI (2023) · cites this paper as [22]
Existing AI data sets are acquired from limited geographical regions and most lack information necessary to evaluate subgroup performance of models.
“Existing AI data sets are acquired from limited geographical regions; most lack information necessary to evaluate subgroup performance of models.”
From AI pitfalls and what not to do: mitigating bias in AI (2023) · cites this paper as [22]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Dissecting racial bias in an algorithm used to manage the health of populations
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019) · Science · doi:10.1126/science.aax2342 · Cited by 7,137 · see citing works
Zech, J. R., Badgeley, M. A., Liu, M., Costa, A. B., Titano, J. J., & Oermann, E. K. (2018) · PLOS Medicine · doi:10.1371/journal.pmed.1002683 · Cited by 1,719 · see citing works
Machine Learning and Health Care Disparities in Dermatology
Adamson, A. S., & Smith, A. (2018) · JAMA Dermatology · doi:10.1001/jamadermatol.2018.2348 · Cited by 639 · see citing works
How to Read Articles That Use Machine Learning
Liu, Y., Chen, P.-H. C., Krause, J., & Peng, L. H. (2019) · JAMA · doi:10.1001/jama.2019.16489 · Cited by 530 · see citing works
A Mathematical Approach to Medical Diagnosis
Warner, H. R. (1961) · JAMA · doi:10.1001/jama.1961.03040290005002 · Cited by 374 · see citing works
Inconsistent Performance of Deep Learning Models on Mammogram Classification
Wang, X., Liang, G., Zhang, Y., Blanton, H., Bessinger, Z., & Jacobs, N. (2020) · Journal of the American College of Radiology · doi:10.1016/j.jacr.2020.01.006 · Cited by 176 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
On the Opportunities and Risks of Foundation Models
Bommasani, R., Hudson, D. A., Adeli, E., Altman, R., Arora, S., von Arx, S., Bernstein, M. S., Bohg, J., Bosselut, A., Brunskill, E., et al. (2021) · arXiv · doi:10.48550/arxiv.2108.07258 · Cited by 2,676 · see citing works
Self-citationAcosta, J. N., Falcone, G. J., Rajpurkar, P., & Topol, E. J. (2022) · Nature Medicine · doi:10.1038/s41591-022-01981-2 · Cited by 1,365 · see citing works
Sources of bias in artificial intelligence that perpetuate healthcare disparities—A global review
Celi, L. A., Cellini, J., Charpignon, M.-L., Dee, E. C., Dernoncourt, F., Eber, R., Mitchell, W. G., Moukheiber, L., Schirmer, J., Situ, J., et al. (2022) · PLOS Digital Health · doi:10.1371/journal.pdig.0000022 · Cited by 576 · see citing works
Wu, E., Wu, K., Daneshjou, R., Ouyang, D., Ho, D. E., & Zou, J. (2021) · Nature Medicine · doi:10.1038/s41591-021-01312-x · Cited by 494 · see citing works
The current and future state of AI interpretation of medical images
Rajpurkar, P., & Lungren, M. P. (2023) · New England Journal of Medicine · doi:10.1056/nejmra2301725 · Cited by 443 · see citing works
Digital medicine and the curse of dimensionality
Berisha, V., Krantsevich, C., Hahn, P. R., Hahn, S., Dasarathy, G., Turaga, P., & Liss, J. (2021) · npj Digital Medicine · doi:10.1038/s41746-021-00521-5 · Cited by 344 · see citing works
Health data poverty: an assailable barrier to equitable digital health care
Ibrahim, H., Liu, X., Zariffa, N., Morris, A. D., & Denniston, A. K. (2021) · The Lancet Digital Health · doi:10.1016/s2589-7500(20)30317-4 · Cited by 326 · see citing works
AI pitfalls and what not to do: mitigating bias in AI
Gichoya, J. W., Thomas, K., Celi, L. A., Safdar, N., Banerjee, I., Banja, J. D., Seyyed-Kalantari, L., Trivedi, H., & Purkayastha, S. (2023) · The British Journal of Radiology · doi:10.1259/bjr.20230023 · Cited by 277 · see citing works
“22 While there have been initiatives to build data sets from “data deserts”, presently, AI models are trained on data collected from tertiary care academic institutions.”
Randomized Clinical Trials of Machine Learning Interventions in Health Care: A Systematic Review
Plana, D., Shung, D. L., Grimshaw, A. A., Saraf, A., Sung, J. J. Y., & Kann, B. H. (2022) · JAMA Network Open · doi:10.1001/jamanetworkopen.2022.33946 · Cited by 236 · see citing works
Federated benchmarking of medical artificial intelligence with MedPerf
Karargyris, A., Umeton, R., Sheller, M. J., Aristizabal, A., George, J., Wuest, A., Pati, S., Kassem, H., Zenk, M., Baid, U., et al. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00652-2 · Cited by 188 · see citing works
Understanding Liability Risk from Using Health Care Artificial Intelligence Tools
Mello, M. M., & Guha, N. (2024) · New England Journal of Medicine · doi:10.1056/nejmhle2308901 · Cited by 155 · see citing works
Guo, L. N., Lee, M. S., Kassamali, B., Mita, C., & Nambudiri, V. E. (2022) · Journal of the American Academy of Dermatology · doi:10.1016/j.jaad.2021.06.884 · Cited by 122 · see citing works
Characterizing the clinical adoption of medical AI devices through U.S. insurance claims
Wu, K., Wu, E., Theodorou, B., Liang, W., Mack, C., Glass, L., Sun, J., & Zou, J. (2024) · NEJM AI · doi:10.1056/aioa2300030 · Cited by 103 · see citing works
Continual learning in medical devices: FDA's action plan and beyond
Vokinger, K. N., Feuerriegel, S., & Kesselheim, A. S. (2021) · The Lancet Digital Health · doi:10.1016/s2589-7500(21)00076-5 · Cited by 95 · see citing works
Regulating AI in medicine in the United States and Europe
Vokinger, K. N., & Gasser, U. (2021) · Nature Machine Intelligence · doi:10.1038/s42256-021-00386-z · Cited by 68 · see citing works
Lekadir, K., Osuala, R., Gallin, C., Lazrak, N., Kushibar, K., Tsakou, G., Aussó, S., Alberich, L. C., Marias, K., Tsiknakis, M., et al. (2021) · arXiv · doi:10.48550/arxiv.2109.09658 · Cited by 58 · see citing works
Kim, J. Y., Hasan, A., Kellogg, K. C., Ratliff, W., Murray, S. G., Suresh, H., Valladares, A., Shaw, K., Tobey, D., Vidal, D. E., et al. (2024) · PLOS Digital Health · doi:10.1371/journal.pdig.0000390 · Cited by 61 · see citing works
Machine Learning for Health: Algorithm Auditing & Quality Control
Oala, L., Murchison, A. G., Balachandran, P., Choudhary, S., Fehr, J., Leite, A. W., Goldschmidt, P. G., Johner, C., Schörverth, E. D. M., Nakasi, R., et al. (2021) · Journal of Medical Systems · doi:10.1007/s10916-021-01783-y · Cited by 48 · see citing works
Solving medicine's data bottleneck: Nightingale Open Science
Mullainathan, S., & Obermeyer, Z. (2022) · Nature Medicine · doi:10.1038/s41591-022-01804-4 · Cited by 43 · see citing works
Berisha, V., & Liss, J. M. (2024) · npj Digital Medicine · doi:10.1038/s41746-024-01199-1 · Cited by 41 · see citing works
Predicting patient demographics from chest Radiographs with deep learning
Adleberg, J., Wardeh, A., Doo, F. X., Marinelli, B., Cook, T. S., Mendelson, D. S., & Kagen, A. (2022) · Journal of the American College of Radiology · doi:10.1016/j.jacr.2022.06.008 · Cited by 32 · see citing works
Schmidt, E., Putora, P. M., & Fijten, R. (2025) · Philosophy & Technology · doi:10.1007/s13347-024-00834-9 · Cited by 14 · see citing works
“Transfer of AI tools across different populations is a real-world problem, as evidenced by the studies in Kaushal et al. ( 2020 ), by AI systems used to detect intracranial hemorrhages (O’Connor, 2021 ), or by Watson for oncology, which is used world-wide despite being trained by doctors at one US…”
Owens, K., Griffen, Z., & Damaraju, L. (2025) · BMC Health Services Research · doi:10.1186/s12913-025-13388-z · Cited by 19 · see citing works
“Approximately 60% of the articles only used data from these three states, 34 states were completely unrepresented, and 13 states contributed limited data [15].”
- Authors
- Kaushal, A., Altman, R., & Langlotz, C.
- Affiliations
- Stanford University (US)
- Department of Radiology, Stanford University School of Medicine, Stanford, California
- Corresponding author
- Kaushal, Amit
- Venue
- JAMA
- Publisher
- American Medical Association (AMA)
- Published
- 2020-09-22
- Volume
- 324
- Issue
- 12
- Pages
- 1212
- ISSN
- 0002-9955 (ISSN-L)
- 0098-7484
- PMC
- PMC7509620
- OpenAlex
- W3087217556
- Cited by
- 239
- References
- 6
- Citations by year
- 2026: 28 · 2025: 37 · 2024: 38 · 2023: 51 · 2022: 46 · 2021: 37 · 2020: 2
- FWCI
- 10.01
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- JAMA (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › scientometrics and bibliometrics research
- Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
- Keywords
- deep learning training data, health disparities, radiology, ophthalmology, dermatology, pathology, gastroenterology
- MeSH terms
- Deep Learning, Algorithms, Diagnosis, Computer-Assisted, Humans, United States
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
