Basics — what to protect, why, and how far
This page explains the ideas on which radiation protection is built, as general knowledge. It starts from the three principles of the ICRP and then goes through medical, occupational and public exposure. Our own research is presented on the Research page, which you can also reach from the "Related research in our lab" box at the end of each section.
Radiation protection is not the science of avoiding radiation. Radiation is an indispensable tool in diagnosis, treatment and research, and radiation protection is a body of ideas and techniques for reducing only unnecessary exposure while keeping those benefits. The International Commission on Radiological Protection (ICRP) organises this system around three principles.
The ICRP is an independent scientific organisation, founded in 1928, that issues international recommendations on radiological protection. They are not laws, but national regulations are built on them. Japanese translations of the main publications are published by the Nuclear Regulation Authority.
ICRP website / Japanese translations of ICRP publications (Nuclear Regulation Authority, in Japanese)
Does the use of radiation produce benefits that outweigh the harm it causes? This is the judgement of whether to do it at all.
Once it is decided to proceed, keep the number of people exposed, individual doses and the likelihood of exposure as low as reasonably achievable, taking economic and societal factors into account.
In planned exposure situations, sets upper limits that individuals must not exceed. Note that they do not apply to the medical exposure of patients.
The same word "exposure" is governed by completely different rules depending on who receives it and why. This page follows these three categories in order.
| ① Medical exposure patients |
Because the patients themselves benefit, dose limits do not apply. Patients are protected instead by justification (is the examination really needed?) and optimisation (lowering the dose while keeping the image quality needed for diagnosis). Diagnostic reference levels (DRLs) are used as a guide. |
|---|---|
| ② Occupational exposure medical staff |
Exposure incurred through work. Dose limits apply (effective dose 100 mSv in 5 years and 50 mSv in any year; lens of the eye 20 mSv per year averaged over 5 years). Monitoring with personal dosimeters is a prerequisite. |
| ③ Public exposure and the environment residents |
In planned exposure situations, 1 mSv per year (additional exposure of the public). In existing exposure situations, such as after an accident, reference levels rather than dose limits guide optimisation. For the long-term recovery phase they are selected from the lower half of the 1–20 mSv per year band, and exposures are reduced step by step toward the lower end of the band (ICRP Publication 146). |
The practical three principles for reducing external exposure — time, distance and shielding — and the use of protective equipment are covered in Education — three principles of protection and protective equipment. Here we explain what comes before that: what the problems are, and in what framework decisions are made.
→ Related research in our labFive research areas / Our approach — measure, simulate, visualize, make
→ Glossary (in Japanese)Effective dose
According to Volume I of the 2020/2021 Report of the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), about 4.2 billion radiological examinations are performed worldwide each year, and the medical exposure per person is estimated at about 0.57 mSv per year. Japan has a very large number of CT scanners per head of population and is considered to have a high level of medical exposure by international standards (country comparisons and estimates for Japan are in the glossary article "Dose optimisation" (in Japanese)).
| World population | ×1.27 |
|---|---|
| Number of radiological procedures | ×1.70 (×1.34 per 1,000 population) |
| Dose per person | ×1.33 |
| CT examinations | about 220 million (UNSCEAR 2008) → about 400 million (2020/2021), +82% |
UNSCEAR 2020/2021 Report, Volume I (United Nations)
The table summarises values published in Annex A, "Evaluation of medical exposure to ionizing radiation", of that report. We do not reproduce the report's figures or tables.
Setting the share of examinations beside the share of dose for each type of examination makes priorities clear. CT accounts for only about one in ten examinations but for six-tenths of the dose. IVR and nuclear medicine each account for less than 1% of examinations, yet for 8% and 7.2% of the dose — too much to ignore.
Disparities by income level are also large: high-income countries receive about 10 times the frequency and 13 times the dose of low-income countries — about 13 times for CT and 18 times for radiotherapy, the gap widening with more advanced care. Optimising medical exposure is therefore also a question of access to necessary care.
A series of CT overexposure incidents in the late 2000s prompted countries to build systems for measuring and collecting doses. The systems differ from country to country, but they share a cycle of record → evaluate → optimise.
The American College of Radiology (ACR) launched the Dose Index Registry (DIR) in 2011. Doses are collected automatically in structured form via DICOM-RDSR and compared with data from other facilities nationwide, supporting optimisation and the detection of outliers. As of January 2025 it held data on more than 236 million examinations from about 2,400 facilities (ACR NRDR).
European Directive 2013/59/Euratom requires Member States to apply justification and optimisation and to keep records of patient exposure. EuroSafe Imaging, led by professional societies, supports implementation in practice and in 2021 published European recommendations on introducing dose management systems (DMS).
Since the revised regulations under the Medical Care Act took effect in 2020, recording and managing doses is mandatory. Hospitals must appoint a person responsible for medical radiation safety, prepare guidelines and provide training, and optimise their own protocols with reference to the Japanese diagnostic reference levels (DRLs) developed by professional societies. Doses are evaluated at least once a year.
Digitisation brought a new difficulty to radiation protection. Unlike film, an image does not fail when the dose is excessive — it actually looks better, so "dose creep", a gradual increase in dose, cannot be detected by eye. To address this, the ICRP set up Task Group 108, which has produced two publications.
| ICRP Publication 154 (2023) |
Presents a strategic framework for achieving optimisation, built on three pillars: ① collaboration between professions (radiologists, radiographers and medical physicists); ② methodology (evaluation with DRLs and objective image-quality assessment with model observers); and ③ processes (quality-control systems and the PDCA cycle). |
|---|---|
| ICRP Publication 160 (2024) |
Modality-specific practice building on the first report: AEC, copper filtration and collimation in radiography; pulse rate and skin-dose monitoring in fluoroscopy and IVR; ATCM and deep-learning reconstruction (DLR) in CT; and considerations for children and pregnant patients. It also touches on AI and the automation of dose monitoring. |
Publication 154 also presents a maturity model so that each facility can see where it stands. The quality of protection is not fixed at the moment a system is introduced; it builds up gradually through organisation, methodology and processes.
| D Preliminary | Performance testing limited to regulatory compliance. Medical physics support limited or absent. Limited documentation of procedures. |
|---|---|
| C Basic | A medical physicist for diagnostic radiology is appointed. Procedures documented for some modalities and high-frequency examinations. DRLs introduced. |
| B Intermediate | An optimisation team of physicians, radiographers and physicists is established. Local DRLs are set, and patient dose surveys are used to develop protocols. |
| A Advanced | Multidisciplinary teams work together routinely. Dose-monitoring software is used to optimise on the basis of individual doses, and image quality is also assessed objectively. |
→ Related research in our labPrecise dose evaluation for medical exposure / How can CT doses be optimised? (Questions behind our research)
→ Glossary (in Japanese)Diagnostic reference levels (DRL) / Dose concepts and evaluation in angiography / Diagnostic medical physicists
Optimisation is the most misunderstood word in radiation protection. Properly defined, it means "keeping the number of people exposed, individual doses and the likelihood of exposure as low as reasonably achievable, taking economic and societal factors into account." The aim is not to minimise dose.
Thinking about clinical practice makes this clear. Lowering the dose degrades the image, and if the diagnosis cannot be made, a repeat examination may increase exposure instead. Adding more protective equipment burdens the operator and lengthens the procedure. Counting the disadvantages of reducing dose as well, and looking for the choice that is better overall — that is optimisation. This is why we need yardsticks that show how we compare with others, such as diagnostic reference levels (DRLs).
The same structure holds outside medicine. Decontamination lowers the dose where people live, but at the same time workers are exposed, the local concentration rises where the soil is collected, and cost, years and storage space are required. Not decontaminating, on the other hand, leaves residents with higher doses (③ Public and environmental protection and decontamination).
| What improves | Doses fall where people live / contamination is moved away from living areas / covering with soil lowers doses further |
|---|---|
| What happens instead | Workers are exposed / local concentrations rise where material is collected / cost, years and storage space are needed |
| Non-radiological risks | Inhalation of dust / occupational accidents with heavy machinery and work at height / chemicals such as asbestos / traffic accidents involving transport vehicles / changes to landscape and ecosystems / health effects of relocation |
For protection after a large nuclear accident, ICRP Publication 146 (2020) gives the formal recommendations. Reflecting the experience of Chernobyl and the Fukushima Daiichi accident, it covers everything from the emergency response to long-term recovery. Importantly, it treats decontamination not as "removing as much radioactive material as possible" but as "environmental remediation that reduces exposure and its social and environmental impacts as a whole."
Do the benefits of protective actions such as decontamination outweigh the disadvantages for workers and society?
Choose by weighing technical, economic, social and environmental factors, not dose alone. Reference levels are used as a tool to drive this optimisation.
Reference levels — a tool for optimisation
For protection of the public, the reference level for the long-term phase is selected from the lower half of the 1–20 mSv per year band. The goal of optimisation is to reduce exposures gradually toward the lower end of the band, or below it where possible. A reference level is not a line of safety but a marker for prioritising the reduction of exposures above it.
Source: ICRP Publication 146 (2020), Main Points (i) and paragraph (193).
Reference levels need to be read with care. You start from the high side and work downward; they are not raised from 1 mSv. The lower end of the band (1 mSv per year) is not a boundary between safe and dangerous either, but a marker for continued improvement.
Not deciding on the basis of dose alone applies equally in medicine and in the environment.
→ Related research in our labGround contamination dose simulator — estimating quantitatively how well decontamination works
→ Glossary (in Japanese)Dose optimisation
Our laboratory publishes web materials that let you try out these ideas in a browser. No installation is needed. The full list is on the Education page.
Estimates the patient's entrance surface dose from tube voltage, filtration and distance by Monte Carlo calculation (Birch–Marshall).
Computes the 3D dose distribution inside the body during CT by Monte Carlo simulation and shows the dose to each organ.
Learn, by operating the screens, the cycle of recording, evaluating and optimising patient doses required under Japan's Medical Care Act.
Seven steps and 24 scenarios of water-phantom measurement, checking distances (SSD/SCD/SAD) and correction factors in 3D as you go.
This section summarises materials published by the organisations concerned; we do not reproduce their figures or tables. For the originals, see the ICRP website, the Japanese translations of ICRP publications (Nuclear Regulation Authority) and UNSCEAR.
As radiological procedures increase, not only radiologists but also physicians in cardiology, gastroenterology, neurosurgery, orthopaedics, urology and gastrointestinal surgery, and the nurses who assist them, are exposed under X-ray fluoroscopy. Unlike patients, medical staff are subject to dose limits.
The ICRP's 2011 Seoul Statement (formally published as Publication 118, 2012) substantially lowered the equivalent dose limit for the lens of the eye (applied in Japan from April 2021), making occupational dose management for staff in IVR (image-guided interventional procedures) an urgent issue.
Yet radiation is invisible and gives no sensation, so protective behaviour is hard to sustain — a fundamental difficulty. Solving this "invisibility" with technology is the consistent theme of our laboratory.
| Equivalent dose limit for the lens of the eye | 20 mSv per year (averaged over 5 years, not exceeding 50 mSv in any year). Applied in Japan from April 2021. Source: ICRP Seoul Statement, 2011 (ICRP Publication 118, 2012). Occupational protection in IVR: ICRP Publication 139 (2018) |
|---|---|
| Effective dose limits | 100 mSv in 5 years and 50 mSv in any one year (radiation workers). Source: Ordinance on Prevention of Ionizing Radiation Hazards (Japan) |
| Health effects of fluoroscopically guided procedures | A survey of spine surgeons (227 surgeons, 454 hands) reported that, compared with the hand receiving less direct X-ray exposure, the more exposed hand had an adjusted odds ratio of 3.18 (95% CI 2.24–4.52) for longitudinal melanonychia and 2.26 (95% CI 1.67–3.06) for hand eczema. Source: Hijikata et al., Eur. Spine J. 30:3702–3708, 2021 / Miura, Rinsho Seikei Geka 55, 2020 |
| International concern | The IAEA identifies the monitoring of medical staff doses as a challenge common to all countries, and highlights real-time electronic dosimeters, automatic staff tracking and virtual simulators. Source: IAEA (press release on radiation protection in medicine) |
The values above are quoted from the published sources; we do not reproduce the original figures or tables.
As a research group of the Japanese Society of Radiological Technology chaired by Fujibuchi, we surveyed the lens-of-the-eye equivalent doses of radiation workers at National Hospital Organization hospitals by profession and by task (4,493 person-years in total). Among medical staff under non-uniform exposure management, 2.9% exceeded 20 mSv per year (values not accounting for the effect of protective eyewear).
| Annual equivalent dose | Physicians | Nurses | Radiological technologists | Others | Total |
|---|---|---|---|---|---|
| 0 mSv | 1,151 | 505 | 94 | 115 | 1,865 |
| up to 2.5 mSv | 711 | 257 | 446 | 104 | 1,518 |
| up to 5 mSv | 191 | 67 | 152 | 5 | 415 |
| up to 10 mSv | 173 | 57 | 86 | 1 | 317 |
| up to 20 mSv | 159 | 46 | 41 | — | 246 |
| up to 30 mSv | 67 | 7 | 1 | — | 75 |
| up to 50 mSv | 37 | 2 | 1 | — | 40 |
| up to 75 mSv | 10 | 1 | — | — | 11 |
| up to 100 mSv | 5 | 1 | — | — | 6 |
| Total | 2,504 | 943 | 821 | 225 | 4,493 |
Source: Fujibuchi T, et al. "Proposal for Reduction Measures of Eye Lens Exposure Based on Actual Exposure Management in Radiation-exposed Medical Staff." Japanese Journal of Radiological Technology 77(2):160–171, 2021 (in Japanese)
The distribution shows that those most likely to exceed 20 mSv per year are physicians and nurses involved in fluoroscopy — by department, angiography, fluoroscopy, endoscopy and operating rooms; by specialty, cardiology, gastroenterology, gastrointestinal surgery, radiology and orthopaedics. Radiological technologists may also receive high doses when briefly assisting in radiography or CT.
| Highly exposed workers | Some physicians exceed 20 mSv per year. |
|---|---|
| Insufficient use of protective equipment | Protective eyewear is little used. |
| Deficiencies in personal dosimetry | In unannounced checks of medical staff, the proportion not wearing their chest/abdomen personal dosimeter was 60.9% for physicians, 23.0% for nurses and 6.6% for radiological technologists. |
| Lack of awareness of legal requirements | In facilities outside the scope of the Radioisotope Regulation Act, responsibility for management tends to be unclear. |
Source: Science Council of Japan, opinion "Towards improving radiation management relating to the occupational exposure of medical workers" (19 September 2023, in Japanese)
Recommended measures include appointing a radiation safety manager with clearly defined authority, establishing in-house rules (covering dose measurement, training and health checks), mandatory training before starting work and refresher training, thorough management of non-uniform exposure, and centralised management of personal doses across multiple facilities. The Japan Radiological Society has also prepared guidelines on radiation safety for medical staff (2025).
The biggest reason protective behaviour is not followed is that nobody sees where and how scattered radiation spreads. Our laboratory publishes scatter distributions calculated by Monte Carlo simulation in a form that can be viewed in a browser or on a tablet.
Shows on a floor plan where, and how much, scattered radiation spreads in a radiography room — useful for deciding where to stand.
Explore the 3D scatter distribution of a CT room in your browser — the web version of X-SERVE.
Change the exposure conditions and see interactively how the dose distribution in the room changes.
Follow how photons interact and scatter, and see what goes on inside a Monte Carlo calculation. For basic learning.
→ Related research in our labX-SERVE — AR scatter visualization for X-ray rooms (Education) /Scatter visualization and real-time imaging
→ Glossary (in Japanese)Visualising exposure and optimisation
→ As teaching materialEducation
The ideas of radiation protection work in the same way outside medicine. Environmental radiation after a nuclear disaster is the best teaching material for this.
The inverse-square law we learn at school applies only to point sources. Over widely contaminated ground (an area source), the area of each ring of distant contamination grows in proportion to its radius, so raising the height hardly lowers the dose.
A person standing on the ground receives dose not only from the ground beneath their feet but from a wide surrounding area. Knowing how much each area contributes (the contribution) makes it possible to estimate quantitatively how effective decontamination will be.
Decontamination therefore has to be designed by contribution, not by area. Decontaminating only your own plot leaves the contribution from outside, while places with a large contribution — under gutters or in drainage ditches — can give a large effect for a small amount of removal. If caesium has penetrated into the soil, self-shielding increases and the contribution from far away falls, so the effective range shifts toward your feet and the same decontamination area gives different results.
What does 0.23 µSv/h mean? This frequently quoted figure is the ambient dose rate corresponding to an additional 1 mSv per year — an administrative threshold used to designate the Intensive Contamination Survey Areas (0.19 from the accident plus an assumed 0.04 from natural background). It is not a decontamination target, not a criterion for return, and not a boundary between safe and dangerous. Both "below it is safe" and "above it is dangerous" are wrong.
Calculates the dose distribution from contaminated ground; compare contribution rings, decontamination area, house shielding and penetration into the soil.
Dialogue practice with an AI playing a member of the public or a patient. It scores you on empathy, scientific accuracy, clarity and support for decision-making, and suggests improvements.
"Let's look inside the body with X-rays" — an AR lesson for children that also shows the X-rays bouncing off (scattered radiation).
All of these run in a browser. The values shown are model calculations, not measurements at any specific location.
Establishing systems and installing equipment will not lower doses unless behaviour in the workplace changes. What is needed in the end is a radiation protection culture — not following rules, but sharing a state in which each person can think and judge for themselves.
You can measure for yourself. You understand what the numbers mean, and their limits. You can say "I don't know". So you can make your own judgements.
Measure together (pass on how to measure, not just numbers). Communicate conditions and uncertainties too. Create places where questions can be raised. Disclose errors instead of hiding them.
Experts are not the only ones who decide. The role of professionals is to provide the information and conditions people need to judge: present scientific evaluations clearly, and do not make value judgements on others' behalf.
| ① Measure for yourself | When you hold a survey meter, change the height and position and look at the difference. Check whether the reading changes when you step a few paces away. |
|---|---|
| ② Always check units and conditions | Bq/m², µSv/h or mSv per year? When, where, and at what height? |
| ③ Be able to say "I don't know" | Not asserting something is not evasion. Separating what you can show from what you cannot is honesty. |
| ④ Don't take calculations at face value | A beautiful figure on the screen is not in itself proof that it is right. Check the conditions and assumptions. |
When asked "Is it safe?", what to present is the measured values and conditions, the natural background and the additional dose, the expected annual dose and its range, and the options for lowering it. What to avoid is groundless assertion and reaching a conclusion without asking about the person's own values. This procedure has exactly the same shape as risk communication in the examination room when a patient says "I'm worried about the radiation from CT."
We also publish a web app for practising this dialogue ("Radiation Risk Communication Training" in Education — web materials).
→ Related activities of our labEducation / Putting XR materials into teaching — and testing whether they work
In-depth articles on individual terms and topics in radiation protection. They are currently available in Japanese only.
New articles are published first on our note (in Japanese).
How our laboratory studies these questions, and how we turn the results into teaching materials, is presented on the following pages.
Measure, simulate, visualize: the eye-lens exposure survey, scatter visualization (X-SERVE), dose evaluation for medical exposure, the ground contamination dose simulator and more.
Time, distance and shielding in practice, XR teaching materials and how we test them, and our browser-based web materials.
Please email us directly. We usually reply within a few working days.
Radiation protection training and FD seminars for medical staff.
Email about trainingFor prospective undergraduate, master's, doctoral and international students.
Email about joining the lab