Before a radiotherapy treatment the radiographers often take a quick X-ray picture or scan on the treatment machine and compare it with the planning scan, to check that you are in the planned position. If the position is out, the couch is moved to correct it before the beam is switched on. How often this is done depends on the part of the body and the type of treatment.
Image guidance compares an image taken on the treatment machine with the reference image from planning and corrects the difference. No source read grades its recommendations by evidence level; the fullest current document is the UK Radiotherapy Board's On target 2 (2021), read in full, beside the IAEA's implementation report (2019) and the AAPM practice guideline on IGRT systems (MPPG 2.b, 2021). The US ASTRO white paper update (2023) and ACR-ARS practice parameter (2025) could be read only as abstracts.
On target 2 asks each centre for site-specific protocols tailored to the site, with frequency, imaging dose and complexity reflecting treatment intent, site and fractionation; it says automatic matches should always be checked visually and gross errors always investigated. The IAEA has the radiation oncologist, with the multidisciplinary team, choose and prescribe the image guidance protocol, and says the protocol should state the frequency of imaging and the actions when a shift is above a threshold. MPPG 2.b, a guideline on the imaging systems, lists the duties of each profession. On imaging dose, On target 2 cites UK law (IR(ME)R) that exposures are justified and optimised, the IAEA says doses are justified, prescribed and reported, and MPPG 2.b says the dose must be measured at least annually.
They differ in three places. Correction rule: On target 2 recommends no action threshold for daily couch corrections where the correction is automated (every measured shift is applied), while the IAEA describes both that approach and action-level protocols that shift only above a set figure (its example is 2 mm), and MPPG 2.b asks physicists and oncologists to set repositioning thresholds. Online or offline: On target 2 notes a shift towards more online verification and says offline protocols are now much less commonly used, but that many of its site chapters still include them and that they so far have the largest evidence base; the IAEA says the choice depends on the site treated and the department's resources. Who decides: On target 2 puts online matching with trained radiographers at the machine with an independent second check; MPPG 2.b lists timely review of images among the radiation oncologist's duties; the IAEA gives the therapist final responsibility for the change, in some instances with input from the radiation oncologist, lists review and approval of online and offline images among the radiation oncologist's duties, and calls later offline review and approval good practice.
Imaging dose, each body's figures kept apart: On target 2 says three- and four-dimensional cone-beam CT dose is reported as approximately 1 to 5 cGy per scan and that daily imaging may result in total organ doses of 1 to 3% of the prescribed dose; the IAEA says kilovoltage cone-beam CT doses can be of the order of 1 cGy. On target 2 (5% of the prescribed dose) and the AAPM TG-180 abstract (5% of the therapeutic target dose) give the point above which imaging dose should be considered in planning. In the UK, national dose reference levels for cone-beam CT verification in adults were adopted in 2025, set by site as a weighted cone beam dose index, from 3.5 mGy (brain) to 21 mGy (prostate and gynaecological); this index is a different measure from the per-scan and organ doses above. A patient's own department's protocol comes first: this page explains the check and is not a set of instructions.
Why: The same guidance says how often to image, the imaging dose and the complexity should reflect the aim of treatment, the part of the body and the number of treatments.
“Each radiotherapy centre should have in place site-specific IGRT protocols that are tailored to the needs of that site and take into account the factors affecting the accuracy of set-up.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Main recommendation (Radiotherapy Board consensus; no evidence grade)
Why: Most imaging on the machine adds a radiation dose on top of the treatment, and the IAEA says these doses are to be justified, prescribed and reported.
“The radiation oncologist, in consultation with the radiotherapy multidisciplinary team, chooses and prescribes the appropriate image guidance protocol.”
International Atomic Energy Agency 2019 · Process description in an IAEA report (no grade)
Why: The plan is built on the position and preparation at the planning scan. The same guidance says patient comfort is essential to keep the position reproducible.
“Patient position and anatomy should be reproducible at daily treatment, using the same preparation, immobilisation and motion management.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Recommendation (Radiotherapy Board consensus; no evidence grade)
Why: A scan typically gives much better contrast, which matters where soft tissue has to be seen; flat pictures are usually matched on bone or implanted markers.
“3D volumetric techniques should be considered where possible because of improved contrast and for analysis and visual verification.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Recommendation (Radiotherapy Board consensus; no evidence grade)
Why: A poor image cannot be trusted for a match, and a large mistake should be seen before any fine adjustment is made.
“It is important to assess the acquired image immediately for image quality and to determine whether it is fit for its intended purpose.”
International Atomic Energy Agency 2019 · Guidance text in an IAEA report (no grade)
Why: If the software matches on a bone that has rotated, such as the thigh bone, the target can be put in the wrong place even though the match looks good.
“It is important not to include anatomy that could adversely affect the registration, such as the femur in patients with prostate cancer where rotation of the femurs could dominate the registration and not result in a true representation of the pelvic position, or the mandible in patients with oral cancer, which may not be stable.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 10.1.4; not a boxed recommendation)
Why: Matching software can fail. The guidance gives alignment on the wrong bone of the spine as an example of a matching error that checks are there to catch.
“In all cases, automatic algorithms can fail, so the results should always be checked visually.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 5.2; not a boxed recommendation)
Why: Correcting on the day removes both the regular error and that day's chance error, and an automatic couch move costs almost no time.
“Since there is no or little workload and minimal uncertainty associated with automated patient set-up corrections, it is recommended to use no action thresholds for daily table corrections.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Recommendation (Radiotherapy Board consensus; no evidence grade)
Why: Decisions at the machine are made quickly, so each shift is checked before the beam is on. The IAEA says shifts need to be recorded and reviewed, and later review can show trends.
“It is important to verify every patient shift prior to treatment delivery.”
International Atomic Energy Agency 2019 · Guidance text in an IAEA report (no grade)
Why: The guidance lists possible causes of a gross error: the wrong patient, part of the body or orientation, the wrong field or treatment centre, or a large change in the body. Planning margins are not meant to cover errors of this size.
“It is important to always investigate the source of such gross errors prior to the decision to treat.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 6.4.2); the boxed recommendation reads 'Gross errors should always be investigated.'
Why: Some changes seen on the image are medical problems that need attention, and others affect where the tumour sits or how accurately the dose is delivered.
“It is also imperative to detect changes that may be related to medical conditions (for example, lung collapse) and take appropriate action, such as warning the clinician or delaying treatment.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 5.2); the boxed recommendation says such changes must be detected and reported
Why: Averaging a few days separates the regular error, which matters most for the dose, from day-to-day chance variation. On target 2 says a correction from a single image can make errors larger, and asks for at least three images.
“The extended NAL protocol (eNAL) includes once-weekly imaging in addition to imaging in the first 3–4 fractions.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (Appendix 14.1; not a boxed recommendation)
Why: The people at the machine can act at once, and delay matters because organs move with time; training keeps different staff matching the same way.
“For efficiency, the online registration and correction process should, where possible, be performed by radiographers involved in the actual patient set-up.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Recommendation (Radiotherapy Board consensus; no evidence grade)
Why: The guidance says a second check can reduce the chance of errors. It adds that the amount of checking should fit the clinical goal and should not compromise workflow.
“An independent second check in particular for high-dose hypofractionated regimes (for example, SABR) is strongly recommended, but ideally should be done for standard fractionation regimes as this can reduce the chance of errors.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text, 'strongly recommended' (section 5.2)
Why: The guideline asks for this so that staff stay competent on the systems in use and are trained when a system is upgraded or changed.
“In addition to initial training, it is important that each facility develop a periodic training review program to ensure competency on current systems and augment with training for system upgrades/changes.”
American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 · Guidance text in an AAPM practice guideline (the sentence before it reads 'Training for the operation of the IGRT system must be provided.')
Why: Each X-ray image adds a small dose to healthy tissue; planning for it in advance means it is weighed against the benefit rather than added case by case.
“The expected number and dose of concomitant exposures, including scope for repeat imaging where required, should be included within site-specific protocols.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 5.5); UK law (IR(ME)R) requires exposures to be justified and optimised
Why: The task group's abstract says the dose from current kilovoltage imaging is generally below this threshold, and that doses should still be kept as low as reasonably achievable.
“The recommended threshold beyond which imaging dose should be considered in the treatment planning process is 5% of the therapeutic target dose.”
American Association of Physicists in Medicine (Medical Physics) 2018 · Task group recommendation (abstract; no evidence grade)
Why: The cameras typically use no ionising radiation, but in many cases the skin surface matches the position of the anatomy inside poorly.
“However, because of the poor correlation between external surface and internal anatomy in many cases, surface-based approaches are not a replacement for other forms of IGRT.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Guidance text (section 4.5; not a boxed recommendation)
Why: The safety margin drawn around the target at planning is only right if it matches the errors the department actually measures.
“Routine prospective IGRT data collection for the individual patient, individual treatment protocol and anatomical sites is essential to calculate systematic and random errors and inform local margins.”
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · Main recommendation (Radiotherapy Board consensus; no evidence grade)
How care is usually carried out, read from the guidelines and standards cited. Your own team's protocol and their judgement come first; this page is for understanding it, not for doing it yourself.
A written, site-specific protocol; imaging prescribed for the patient; the same set-up every day; an image fit for the task; a match on the right anatomy, checked by eye; correction before the beam; gross errors investigated before treating; trained and assessed staff with a second check for treatments in few fractions; and imaging dose justified, optimised and recorded. Each point is stated by the UK Radiotherapy Board (On target 2, 2021) or the IAEA (2019), the two documents read in full on clinical practice; the AAPM guideline (MPPG 2.b, 2021) covers the imaging systems, staff duties and training. None grades by evidence level, and no source read contradicts these points. The three bodies differ on three points, and each position is shown. Action threshold: On target 2 recommends none for automated daily couch corrections; the IAEA describes zero action level and action-level protocols (example 2 mm) without choosing; MPPG 2.b asks each department to set repositioning thresholds. Online or offline: On target 2 describes a shift towards more online verification and says offline protocols are now much less commonly used, while many of its site chapters keep them and it says they so far have the largest evidence base; the IAEA says the choice depends on the site and the department's resources. Who matches and who reviews: On target 2 gives online matching to competent radiographers with an independent second check; MPPG 2.b lists timely review of images as a radiation oncologist's responsibility; the IAEA gives the therapist final responsibility for the change, in some instances with input from the radiation oncologist, with later offline review and approval. The current US documents on clinical practice (ASTRO 2023, ACR-ARS 2025) were read as abstracts only, so their detailed positions on frequency and review are not stated here.
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · International Atomic Energy Agency 2019 · American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 · American Society for Radiation Oncology (Practical Radiation Oncology) 2023 · American College of Radiology and American Radium Society (American Journal of Clinical Oncology) 2025 · UK Health Security Agency 2025
Where it differs: The couch is moved only when the measured error is larger than a set figure, instead of applying every measured shift. The IAEA report describes it beside the zero action level approach: "This is by way of contrast to action level protocols, where shifts are only applied if they are larger than a predetermined action level (e.g. 2 mm)". MPPG 2.b lists setting repositioning thresholds among the duties of the physicist and the radiation oncologist.
Evidence: Neither document grades the choice. On target 2 recommends no action threshold where the couch correction is automated, on the grounds of little workload and minimal uncertainty; it keeps tolerances for errors that are not corrected, such as rotations, and says such values should be determined for each treatment site and protocol.
International Atomic Energy Agency 2019 · American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 · Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021
Where it differs: Images are taken on selected fractions and analysed after treatment, and the average error is applied to later fractions. On target 2 describes the no action level approach as calculating the systematic error after 3 to 4 fractions and correcting its whole magnitude, suggests correcting only systematic errors above 2 mm, and adds weekly imaging in the extended form; it says data from a minimum of three images should be combined. The IAEA figures show imaging for typically three successive fractions. In the shrinking action level approach the running average is compared with a limit that falls as more fractions are imaged.
Evidence: The IAEA report: "Random deviations can only be corrected for with an on-line IGRT protocol." and "Off-line correction protocols are therefore most suitable when the ratio of random to systematic deviation is small." On target 2 says offline protocols may be used depending on accuracy requirements and can reduce resources on the treatment unit, that they are now much less commonly used because most centres have automated couch correction, that many of its site-specific protocols still include them and that they have the largest evidence base so far, and that the no action level protocol does not act on gross errors, which should be corrected before a further fraction is given.
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · International Atomic Energy Agency 2019
Where it differs: MPPG 2.b lists among the radiation oncologist's responsibilities specifying imaging modalities and frequencies, identifying registration targets and repositioning thresholds, and conducting a timely review of clinical IGRT images; the therapist acquires positioning images for review by the therapist or the radiation oncologist. The IAEA report: "In the on-line process, subsequent off-line review and approval of the images acquired according to the on-line protocol is a good practice to establish the types and sizes of effect, plus any trends, and to deduce possible causes."
Evidence: MPPG 2.b states that its staff responsibilities are consistent with the ACR-ASTRO practice guideline for IGRT, citing the 2009 edition; the current ACR-ARS revision (2025) could be read only as an abstract. The IAEA report lists among the radiation oncologist's duties direct or indirect supervision of positioning corrections, including systematic or periodic review and approval of online and offline images. On target 2 instead places the online decision with trained radiographers and warns that without appropriate competency more cases would be referred to the clinician, delaying the process. No comparative study of the two models was read.
American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 · International Atomic Energy Agency 2019 · Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · American College of Radiology and American Radium Society (American Journal of Clinical Oncology) 2025
Where it differs: Two flat X-ray images, or megavoltage portal images, are matched on bone or implanted markers rather than a cone-beam CT matched on soft tissue.
Evidence: On target 2 says a cone-beam CT always yields more clinical information but that this may not translate into greater accuracy, so planar imaging can still be justified in a limited number of cases, giving short-course palliative treatment as its example. The IAEA report gives megavoltage portal imaging doses of the order of 10 cGy per image, kilovoltage planar imaging in the range of 0.1 mGy to 2 cGy depending on the imaging strategy, and kilovoltage cone-beam CT of the order of 1 cGy.
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · International Atomic Energy Agency 2019
Where it differs: Cameras track the skin surface to help the first set-up and to watch for movement during treatment, with no radiation dose.
Evidence: On target 2 says surface systems can help initial set-up and give real-time monitoring but are not a replacement for other forms of IGRT because the surface often correlates poorly with internal anatomy. The ESTRO-ACROP guideline on surface guided radiation therapy (2022) gives recommendations on roles, commissioning and quality assurance; only its abstract was read.
Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 · European Society for Radiotherapy and Oncology (Radiotherapy and Oncology) 2022
Where it differs: MPPG 2.b prefers a daily check of imaging and treatment isocentre coincidence and couch repositioning (2 mm; 1 mm for radiosurgery) but accepts weekly checks for systems used with standard fractionation, and requires the check on days when stereotactic radiosurgery or stereotactic body radiotherapy is scheduled. These are tolerances for the machine, not for a patient's set-up.
Evidence: MPPG 2.b: "While daily checks of these parameters are preferred, weekly checks are considered acceptable for IGRT systems used with standard fractionation schemes." The machine checks themselves are in the linear accelerator quality assurance protocol.
American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021
| Recommendation | Grade, as printed | Source |
|---|---|---|
| Frequency, dose and complexity of imaging should reflect intent, site and fractionation “The frequency, imaging dose and complexity of the IGRT process should reflect the treatment intent, anatomical site and fractionation (as detailed in the site-specific guidance in Section 10).” | Main recommendation (Radiotherapy Board consensus; no evidence grade) | Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 |
| No action threshold for automated daily couch corrections (UK) “Since there is no or little workload and minimal uncertainty associated with automated patient set-up corrections, it is recommended to use no action thresholds for daily table corrections.” | Recommendation (Radiotherapy Board consensus; no evidence grade) | Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 |
| Zero action level protocols apply every shift (IAEA describes both approaches) “In zero action level protocols, the shift is applied to the patient set-up irrespective of the size of the shift.” | Description in an IAEA report (no grade) | International Atomic Energy Agency 2019 |
| The choice of online or offline depends on the site and the department's resources (IAEA) “The decision of which to use depends on the site being treated and departmental resources, including the need to balance treatment accuracy with workload and expertise.” | Guidance text in an IAEA report (no grade) | International Atomic Energy Agency 2019 |
| Daily online imaging can often be justified (UK) “Daily online IGRT can often be justified depending on clinical requirements and taking the additional time spent on the treatment unit into consideration.” | Guidance text (section 10.1; not a boxed recommendation) | Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 |
| Gross errors are always investigated “Gross errors should always be investigated.” | Recommendation (Radiotherapy Board consensus; no evidence grade) | Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 |
| Staff training on the imaging system “Training for the operation of the IGRT system must be provided.” | must (AAPM practice guideline wording) | American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 |
| Imaging dose measured at least annually for each mode of clinical operation “The imaging dose from an IGRT system must be measured at least annually for at least one acquisition technique of each mode of clinical operation.” | must (AAPM practice guideline wording) | American Association of Physicists in Medicine (Journal of Applied Clinical Medical Physics) 2021 |
| Imaging dose is considered in planning above 5% of the prescribed dose (UK) “In cases such as this, the imaging dose should be considered in the treatment planning process if it exceeds 5% of the prescribed dose.” | Guidance text (section 5.5; not a boxed recommendation) | Radiotherapy Board (The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and Engineering in Medicine) 2021 |
| Imaging dose is considered in planning above 5% of the therapeutic target dose (US) “The recommended threshold beyond which imaging dose should be considered in the treatment planning process is 5% of the therapeutic target dose.” | Task group recommendation (abstract; no evidence grade) | American Association of Physicists in Medicine (Medical Physics) 2018 |
| Imaging doses are justified, prescribed and reported “In summary, it is important that IGRT imaging doses are justified, prescribed and reported.” | Summary statement in an IAEA report (no grade) | International Atomic Energy Agency 2019 |
| A quality assurance programme built on established guidance is required (US) “A comprehensive quality-assurance program must be developed, using established guidance, to ensure IGRT is performed in a safe and effective manner.” | Consensus white paper (abstract; statements reached consensus at 75% agreement or more) | American Society for Radiation Oncology (Practical Radiation Oncology) 2023 |
| Image guidance complements other quality assurance (US) “IGRT should complement and be used in combination with other quality assurance processes.” | Practice parameter (abstract; no evidence grade) | American College of Radiology and American Radium Society (American Journal of Clinical Oncology) 2025 |
Shares Linear accelerator quality assurance: daily, weekly, monthly and annual machine checks, Patient identification and pre-treatment time-out in radiotherapy, IMRT / IGRT (modern external beam).
Shares Patient identification and pre-treatment time-out in radiotherapy, In-room imaging for radiotherapy (cone-beam CT, ExacTrac, CT-on-rails, HyperSight), Radiotherapy, IMRT / IGRT (modern external beam).
Shares Fiducial markers, Deep inspiration breath-hold for left-sided breast radiotherapy, Breathing motion in lung and upper abdominal radiotherapy: 4D CT, gating, breath-hold and compression, Radiotherapy planning scan (CT simulation): positioning, skin marks and what the patient does.
Shares Fiducial markers, GTV, CTV and PTV (target volumes), IMRT / IGRT (modern external beam).
Shares Breast radiotherapy positioning, immobilisation and daily set-up verification, Radiotherapy planning scan (CT simulation): positioning, skin marks and what the patient does, Surface-guided radiotherapy (SGRT).
Shares IMRT / IGRT / VMAT (modern external beam radiotherapy), GTV, CTV and PTV (target volumes), IMRT / IGRT (modern external beam).
Shares GTV, CTV and PTV (target volumes), Stereotactic radiosurgery and SBRT: planning and delivery safeguards, IMRT / IGRT (modern external beam).
Shares Bladder filling and rectal preparation for prostate radiotherapy, Adaptive radiotherapy (online replanning), IMRT / IGRT (modern external beam).