Optimizing Communication of Radiation Exposure in Medical Imaging, the Radiologist Challenge
Open accessAbstract
Since I started my residency program in Radiology, I have been committed to promoting radiation protection, paying particular attention to the justification and optimization of the examinations [...]
AI snapshot
Written from the paper’s full text.
TL;DR
This commentary argues that radiologists face a persistent challenge in communicating radiation risks and benefits in medical imaging, especially because the risks are stochastic, delayed, and uncertain. It recommends tailored oral communication, familiar comparisons, transparency, and future AI/GPT support, while noting that no empirical study is presented.
Key findings
- CT is responsible for up to 67% of radiation exposure in medical imaging, and a cumulative effective dose above 100 mSv is delivered to more than 1% of patients undergoing multiple CT examinations.
- Radiation risk communication is difficult because patients may be told there is a cancer risk, that it is very low or possibly negligible below 100 mSv, and that it is delayed by 10–20 years.
- The paper recommends oral, tailored communication; avoiding jargon and acronyms; using familiar comparisons such as transcontinental flight, background radiation, cigarette smoking, or driving accident risk; using a risk-ranking scale; describing benefits separately; showing the machine; and explaining dose optimization and dose management systems/DRLs.
- It identifies a legal duty under Council Directive 2013/59/EURATOM, chapter 7, article 57(d), to provide patients with adequate information on benefits and risks, and suggests GPT/AI tools may support radiation protection education and patient communication in the future.
Methods
- Narrative commentary/opinion piece, not an empirical study.
- No dataset, sample, or formal analysis is reported; the text draws on a personal clinical anecdote, cited literature, and regulatory guidance.
- Main content is a discussion of communication challenges and proposed strategies, with a brief speculative section on AI/GPT.
Limitations
- Stated in the paper: individual perception limits communication, and AI models have several limitations.
- Evident from the text: there is no empirical evaluation, systematic review, or tested intervention; recommendations are based on commentary and cited sources rather than new data.
- Evident from the text: the AI/GPT discussion is speculative and does not report validation or outcomes.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
0 of 11 cited sources checked · 0 supported · 0 flagged
- Premisecitation
More than 1% of patients undergoing multiple CT examinations receive a cumulative effective dose above 100 mSv.
“As highlighted by a large multicentre study, a cumulative effective dose above 100 mSv is delivered to more than 1% of patients undergoing multiple CT examinations [2].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [2] Rehani, 2020, Patients Undergoing Recurrent CT Scans: Assessing the Magnitude, Eur. Radiol. source
Not checked in the free run.
- Premisecitation
GPT-based models have been extensively tested and developed in healthcare at several levels, including research, education, clinical decisions, and patient communication.
“In recent years, generative pre-trained transformer (GPT)-based models, AI tools generating large language models (LLMs), have been extensively testing and developing in healthcare at several levels, such as research, education, clinical decisions and patient communication [11,12,13].”
Quote in the paperNot a substantial claim0 of 3 sources checked- cites [11] Sejnowski, 2023, Large Language Models and the Reverse Turing Test, Neural Comput. source
Not checked in the free run.
- cites [12] Lecler, A., Duron, L., and Soyer, P. (2023). Revolutionizing Radiology with GPT-Based Models: Current Applications, Future Possibilities and Limitations of ChatGPT. Diagn. Interv. Imaging, S2211-5684(23)00027-X. source
Not checked in the free run.
- cites [13] Cascella, 2023, Evaluating the Feasibility of ChatGPT in Healthcare: An Analysis of Multiple Clinical and Research Scenarios, J. Med. Syst. source
Not checked in the free run.
- Premisecitation
CT is responsible for up to 67% of radiation exposure in medical imaging.
“The increased use of CT, responsible for up to 67% of radiation exposure in medical imaging, represents a major concern [1].”
Quote in the paperNot a substantial claim0 of 1 sources checked- cites [1] Mettler, 2020, Patient Exposure from Radiologic and Nuclear Medicine Procedures in the United States: Procedure Volume and Effective Dose for the Period 2006–2016, Radiology source
Not checked in the free run.
- May combine two claims. 4
- Classified as probably not a substantial claim. 1
May combine two claims.
PremisecitationPatients want to be informed about radiation-related risks but have a poor understanding of radiation exposure and the risk of medical imaging.
“Patients want to be informed about radiation-related risks, but they have a poor understanding of radiation exposure and the risk of medical imaging [4].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [4] Bastiani, 2021, Patient Perceptions and Knowledge of Ionizing Radiation from Medical Imaging, JAMA Netw. Open source
Not checked in the free run.
May combine two claims.
PremisecitationCouncil Directive 2013/59/EURATOM establishes that radiologists must ensure justification and optimization of examinations and inform patients about benefits and risks associated with imaging.
“In particular, radiologists are called to ensure the justification and optimization of the examinations and to inform patients about the benefits and risks associated with imaging, as established by the Council Directive 2013/59/EURATOM.”
- Also stated in: § 2. The Challenge · hedged
“wherever practicable and prior to the exposure taking place, the practitioner or the referrer, as specified by Member States, ensures that the patient or their representative is provided with adequate information relating to the benefits and risks associated with the radiation dose from the medical exposure.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [3] (2023, February 24). Council Directive 2013/59/Euratom of 5 December 2013 Laying down Basic Safety Standards for Protection against the Dangers Arising from Exposure to Ionising Radiation, and Repealing Directives 89/618/Euratom, 90/641/Euratom, 96/29/Euratom, 97/43/Euratom and 2003/122/Euratom. Available online: https://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2014:013:0001:0073:EN:PDF.
Not checked in the free run.
- Also stated in: § 2. The Challenge · hedged
May combine two claims.
PremisecitationThe uncertainty of the information about radiation risk and the absence of supporting data make patient information and communication difficult.
“The uncertainty of this frightening information and the absence of data available in support makes the patient information and communication difficult [5].”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [5] Lam, 2015, Communicating Potential Radiation-Induced Cancer Risks From Medical Imaging Directly to Patients, AJR Am. J. Roentgenol. source
Not checked in the free run.
Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
PremisecitationChatGPT has been trained to understand and generate language for conversation with users.
“Already ChatGPT, variant of the third GPT generation, has been trained to understand and generate language for conversation with users [12,13,14].”
Quote in the paperNot a substantial claim0 of 3 sources checked- cites [12] Lecler, A., Duron, L., and Soyer, P. (2023). Revolutionizing Radiology with GPT-Based Models: Current Applications, Future Possibilities and Limitations of ChatGPT. Diagn. Interv. Imaging, S2211-5684(23)00027-X. source
Not checked in the free run.
- cites [13] Cascella, 2023, Evaluating the Feasibility of ChatGPT in Healthcare: An Analysis of Multiple Clinical and Research Scenarios, J. Med. Syst. source
Not checked in the free run.
- cites [14] Patel, 2023, ChatGPT: Friend or Foe?, Lancet Digit. Health source
Not checked in the free run.
The paper relies on this without support.
There is a risk of cancer from medical imaging, and cancer is a potentially lethal disease.
“there is a risk of cancer, which is a potentially lethal disease”
The paper relies on this without support.
The risk of cancer from medical imaging is supposed to be very low, possibly negligible below 100 mSv.
“the risk is supposed to be very low, maybe negligible below 100 mSv”
The paper relies on this without support.
The risk of cancer from medical imaging is remote, delayed by 10–20 years.
“the risk is remote, delayed by 10–20 years”
The paper relies on this without support.
The signature on written informed consent does not reflect real patient understanding.
“its signature does not reflect real patient understanding”
The paper relies on this without support.
Patients are often stressed about their health condition, and their attitude may reflect this; they may overestimate or minimize negative information, or not listen to the entire speech, focusing on what they want to hear.
“In clinical situations, patients are often stressed about their health condition, and their attitude may reflect this. They may overestimate or minimize negative information, or they may not listen to the entire speech, focusing on what they want to hear.”
The paper relies on this without support.
Patients want to be reassured that clinicians are taking care of them and are scrupulously committed to their work.
“Patients want to be reassured of the fact that we are taking care of them and that we are scrupulously committed to our work.”
The paper relies on this without support.
Radiologists and technologists are trained in radiation protection and receive continuing education throughout their careers.
“Radiologists and technologists are trained in radiation protection and receive continuing education throughout their careers.”
The paper relies on this without support.
The absence of keys for correct communication may lead to lack of communication, inappropriately understating risks, or overloading the patient with unnecessary details.
“Consequently, the absence of keys for correct communication may lead to: (1) a lack of communication of the risks to avoid useless discussion and concern; (2) inappropriately understating the risks to reassure the patient; and (3) overloading the patient with an excessively scientific lexicon and unnecessary details, miscommunicating the essential information.”
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?
ChatGPT can provide initial observations that highlight regions of interest and assist radiologists in their analyses.
“It has the ability to provide initial observations, which can highlight regions of interest and assist radiologists in their analyses [13,25,26,27,28,29,30,31,32,33].”
ChatGPT has the ability to analyse and generate descriptions for medical images, which can enhance the workflow of radiologists.
“The potential of ChatGPT lies in its ability to analyse and generate descriptions for medical images, which can enhance the workflow of radiologists.”
Cites
Cascella, M., Montomoli, J., Bellini, V., & Bignami, E. (2023) · Journal of Medical Systems · doi:10.1007/s10916-023-01925-4 · Cited by 1,120 · see citing works
Lecler, A., Duron, L., & Soyer, P. (2023) · Diagnostic and Interventional Imaging · doi:10.1016/j.diii.2023.02.003 · Cited by 385 · see citing works
Health, T. L. D. (2023) · The Lancet Digital Health · doi:10.1016/s2589-7500(23)00023-7 · Cited by 257 · see citing works
Mettler, F. A., Mahesh, M., Bhargavan, M., Chambers, C. E., Elee, J., Frush, D. P., Miller, D. L., Royal, H. D., Milano, M. T., Spelic, D. C., et al. (2020) · Radiology · doi:10.1148/radiol.2020192256 · Cited by 236 · see citing works
Patients undergoing recurrent CT scans: assessing the magnitude
Rehani, M. M., Yang, K., Melick, E. R., Heil, J., Šalát, D., Sensakovic, W. F., & Liu, B. (2019) · European Radiology · doi:10.1007/s00330-019-06523-y · Cited by 177 · see citing works
Large Language Models and the Reverse Turing Test
Sejnowski, T. J. (2023) · Neural Computation · doi:10.1162/neco_a_01563 · Cited by 169 · see citing works
Communicating Potential Radiation-Induced Cancer Risks From Medical Imaging Directly to Patients
Lam, D. L., Larson, D. B., Eisenberg, J. D., Forman, H. P., & Lee, C. I. (2015) · American Journal of Roentgenology · doi:10.2214/ajr.15.15057 · Cited by 90 · see citing works
Patient Perceptions and Knowledge of Ionizing Radiation From Medical Imaging
Bastiani, L., Paolicchi, F., Faggioni, L., Martinelli, M., Gerasia, R., Martini, C., Cornacchione, P., Ceccarelli, M., Chiappino, D., Latta, D. D., et al. (2021) · JAMA Network Open · doi:10.1001/jamanetworkopen.2021.28561 · Cited by 71 · see citing works
Self-citationLoose, R., Vano, E., Mildenberger, P., Tsapaki, V., Caramella, D., Sjöberg, J., Paulo, G., Torresin, A., Schindera, S. T., Frija, G. G., et al. (2020) · European Radiology · doi:10.1007/s00330-020-07290-x · Cited by 52 · see citing works
Shyu, J. Y., & Sodickson, A. D. (2015) · British Journal of Radiology · doi:10.1259/bjr.20150868 · Cited by 30 · see citing works
Earl, V. J., Potter, A. O. G., & Perdomo, A. A. (2022) · Journal of Medical Radiation Sciences · doi:10.1002/jmrs.632 · Cited by 28 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Younis, H. A., Eisa, T. A. E., Nasser, M., Sahib, T. M., Noor, A. A., Alyasiri, O. M., Salisu, S., Hayder, I. M., & Younis, H. A. (2024) · Diagnostics · doi:10.3390/diagnostics14010109 · Cited by 242 · see citing works
Shen, L., Lu, J., Zhou, C. J., Bi, Z., Ye, X., Zhao, Z., Xu, M. G., Zeng, M., & Wang, M. (2025) · European Radiology · doi:10.1007/s00330-025-11681-3 · Cited by 6 · see citing works
Jankowski, S., Rotzinger, D. C., Ria, F., & Pozzessere, C. (2024) · Radiology · doi:10.1148/radiol.233465 · Cited by 5 · see citing works
Self-citationBefore We Trust: Concerns for Using ChatGPT in Medical Communication
Wang, S. (2025) · Radiology · doi:10.1148/radiol.241888 · Cited by 2 · see citing works
Chat GPT vs an experienced ophthalmologist: evaluating chatbot writing performance in ophthalmology
Katz, G., Zloto, O., Hostovsky, A., Huna-Baron, R., Mizrachi, I. B.-B., Burgansky, Z., Skaat, A., Vishnevskia-Dai, V., Fabian, I. D., Sagiv, O., et al. (2025) · Eye · doi:10.1038/s41433-025-03779-1 · Cited by 1 · see citing works
ChatGPT versus Radiology Institutional Websites: What Is the Patients' Point of View?
Crimì, F., Sussan, G., & D’Alessandro, C. (2025) · Radiology · doi:10.1148/radiol.241941
- Authors
- Pozzessere, C.
- Affiliations
- Centre Hospitalier Universitaire Vaudois (CH)
- Corresponding author
- Pozzessere, Chiara
- ORCID
- Pozzessere, Chiara
- Venue
- Tomography
- Publisher
- MDPI AG
- Published
- 2023-03-23
- Volume
- 9
- Issue
- 2
- Pages
- 717-720
- ISSN
- 2379-1381 (ISSN-L)
- 2379-139X
- PMC
- PMC10142536
- OpenAlex
- W4360852150
- Semantic Scholar
- a7744908e910ba9691ff05e9089a7c5968696365
- Cited by
- 6
- References
- 11
- Citations by year
- 2025: 4 · 2024: 2
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Tomography (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- SERVAL (Université de Lausanne) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- IRIS (submittedVersion, cc-by)
- Topics
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiation Dose and Imaging
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiology practices and education
- Health Sciences › Dentistry › Oral Surgery › Dental Radiography and Imaging
- Keywords
- radiation protection, diagnostic imaging, radiation dose optimization, radiologist role, patient radiation dose
- MeSH terms
- Radiation Exposure, Radiologists, Communication, Humans, Radiation Protection, Tomography, X-Ray Computed
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.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-10.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-10.
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
