Train radiology techs to tell artifacts from findings with a case bank that pairs each artifact with a similar-looking anatomical finding, so they practise the decision and not just the label. This guide is for diagnostic services directors and radiology educators. It covers five steps: describing the performance problem, diagnosing the gaps, designing practice, working through an example, and assessing transfer. It is about training design, not interpretation guidance. Use reviewed images, and have a qualified radiologist validate every case before anyone is trained on it.
What the research says, and does not say
Barrett and Keat's 2004 article Artifacts in CT: Recognition and Avoidance in RadioGraphics groups CT artifacts by cause and describes how to recognise and avoid them. It is a useful reference for building a case bank's categories.
It describes the artifacts. It does not show that a particular training method works, and it does not replace the review of your own images by a qualified person. Use it to structure your cases, then test the training with your own techs.
Quick facts
| Question | Short answer |
|---|---|
| What is the task? | Distinguish plausible artifacts from findings that need review. |
| What is the safe default? | When unsure, escalate for review under your approved procedure. |
| What makes practice useful? | Unseen cases, a reasoned answer and feedback tied to the decision. |
| What makes scores trustworthy? | Assessors scoring independently, then comparing evidence. |
| What should you avoid scoring? | Speed or confidence alone. |
| What can it not decide? | Interpretation. That belongs to qualified readers. |
1. Describe the performance problem
Start with the problem: a case bank pairs artifacts with similar-looking anatomical findings, and the tech has to decide which is which. State what the role must do: distinguish plausible artifacts from findings requiring review. Note that the cost of a wrong call is not symmetrical. Calling a real finding an artifact is a different error from escalating an artifact, and your criteria should say so.
2. Diagnose the learning and context gaps
Compare the required task with the current activities and assessments. If techs only see a slide deck of labelled artifacts, they practise recognition of named examples, not the decision on an unfamiliar image. Our guide to matching learning activities to Bloom's taxonomy helps you tell recall from analysis. Before prescribing training, check the other causes: image quality, protocols, staffing and who the tech is meant to escalate to.
3. Design purposeful preparation and practice
Plan three layers:
- Preparation. A short self-paced check of prerequisites: terms, common artifact causes and your escalation procedure. A recall check confirms prerequisites only.
- Blind case drills. Show an unseen case, ask for a classification and the reasoning behind it, then give feedback linked to the decision.
- Live input where it adds value. A radiologist or senior tech talks through an ambiguous case. Asynchronous review is enough for clear ones.
A worked case with feedback and a retry
Illustrative. A radiologist or radiology educator should validate the images, case and answer criteria before use.
The case. A reviewed image shows a bright streak that could be an artifact or could resemble a real finding. The case bank pairs it with a similar image of a true finding.
The task. Decide whether this needs review. Explain your decision, cite the evidence and say what further information you need.
A first attempt. "It is a streak, so it is an artifact." A reviewer asks what could cause the streak, whether it follows the geometry of the scan, and what would change the answer. The learner has no answer to any of them.
Feedback and retry. The reviewer notes that the label was applied from the look of the image, with no reasoning and no escalation path. The learner is asked to try again with the paired image beside it.
A stronger attempt. The learner compares the two images, names what each feature does and does not match, and says the call is uncertain. They would escalate under the approved procedure and ask for the patient history, the scan settings and a repeat acquisition if the procedure allows it.
Why each response is defensible or incomplete. Judge against approved criteria, not the reviewer's preference. The first attempt classified by appearance and did not escalate, which is a critical error. The second reasoned from evidence and stayed within the tech's remit.
Building the case bank
A case bank is only as good as its review. Keep it small at first and make every entry earn its place:
- Choose cases from your own department. Artifacts your techs meet weekly teach more than rare textbook examples.
- Pair each artifact with a look-alike finding. The pair is what forces a decision rather than a label.
- Have a radiologist review each image and write the answer. Record the cause, the features that give it away and the escalation step.
- Remove anything identifiable. Follow your department's rules for patient data before images go into training.
- Retire cases that techs have memorised. A case seen twice tests memory, not judgement.
Keep a note of which cases techs get wrong most often. Those are the ones to discuss live, and they tell you where the procedure or the protocol may need attention as well as the training.
4. Assess capability and plan transfer
Use classification accuracy and explanation quality on a new case. Score each criterion from 0 to 2:
| Criterion | 0: absent or unsupported | 1: partial, with omissions | 2: complete and justified |
|---|---|---|---|
| Task execution | Classifies by appearance. | Classifies with a partial reason. | Classifies, cites evidence and records the action. |
| Reasoning | Gives no reason. | Gives a reason without naming uncertainty. | Explains alternatives, constraints and uncertainty. |
| Boundaries | Does not escalate. | Escalates without saying why. | Uses the approved procedure and seeks help when information is insufficient. |
Calibrate the scoring: have two assessors score the same answer independently, compare their evidence and revise any criterion where they disagree. This rubric is a proposed starting point, not a validated readiness standard. Define critical errors separately, for example failing to escalate a plausible finding. An overall score must never hide a critical failure. For criteria on reasoning, see our guide to assessing judgement and problem-solving in training.
Add a workplace follow-up: a supervised review of real escalations after a few weeks. Compare baseline and unseen follow-up cases, define the population, denominator and timing before you start, and look for concurrent changes, such as a new scanner or protocol, before you credit the training.
Where interactive delivery helps
In a live session, show an unseen image and ask each tech to submit a classification privately, then display the spread before discussing the reasoning. Self-paced delivery suits the prerequisite check and the review of reasoning afterwards. AhaSlides, an all-in-one engagement platform for teams and classrooms, supports interactive learning and both live and self-paced delivery. Check how you would run the drills in your own workflow before you rely on it. Image display and storage rules in your department still apply.
What training will not fix
- Poor image quality at source. Training cannot fix a faulty acquisition.
- Unclear escalation routes. Techs cannot escalate to a process that does not exist.
- A case bank nobody reviewed. A wrong label teaches the wrong lesson.
- Evidence limits. The source above describes artifacts, not training outcomes for your team.
Frequently asked questions
What is an artifact in radiology?
A feature in the image that does not correspond to the anatomy being imaged, usually caused by the equipment, the technique or the patient.
Can a tech decide that something is an artifact?
That depends on your approved procedure. The training should teach the tech what to escalate and to whom.
How many cases should a case bank hold?
Enough that techs see unseen cases each time. Build it from reviewed images and add to it as your department meets new examples.
How do you check that assessors agree?
Have them score the same answer independently, compare their evidence and revise any criterion where they differ.
Who should review the case bank?
A radiologist or radiology educator, before any tech is trained or scored on it.
Where to start
Pick one modality and choose five artifacts that your department meets. Pair each with a similar-looking finding, and have a radiologist review the images. Run one blind drill, compare two assessors' scores and revise the criteria where they differ.
To see how a drill could run live or self-paced, explore AhaSlides.








