How to read a radiology report
Read Your Scan Editorial Team·Last updated ·8 min read
Written from published radiology references and reviewed for plain language. Not written or signed by a physician, and not a diagnosis. How we write these pages →
The short answer
A radiology report has five parts: clinical history, technique, comparison, findings and impression. The impression is the radiologist’s answer to the question your doctor asked, so read it first. The findings section above it is a complete inventory of everything visible, normal structures included, which is why it reads far more alarmingly than it is.
- Read the impression first. It is the conclusion; the findings section above it is raw inventory.
- The findings list normal structures too. A paragraph naming an organ is usually saying it looks fine.
- Hedging language — “cannot be excluded”, “clinical correlation is recommended” — is professional caution addressed to your doctor, not a warning aimed at you.
- A finding is not a diagnosis. Only the clinician who ordered the scan can read it next to your symptoms and history.
Why the report frightens people who were never its intended reader
A radiology report is a letter from one clinician to another. The radiologist has never met you, was given one or two lines about why you were scanned, and is answering a specific question for the doctor who ordered the study. Nothing in the document is written with a patient in mind, and until recently no patient read one before their appointment. Electronic portals changed that: the report now often arrives days before the conversation about it.
That is the whole source of the distress. The vocabulary is technical because precision between specialists matters, and the structure front-loads everything the radiologist saw rather than what any of it means. Read in that order — every observation first, the conclusion last — an ordinary study reads like a catalogue of problems. Read in the order below, the same document usually says something much shorter and much calmer.
The five parts of a radiology report
Nearly every report, on every modality, follows the same skeleton. Headings and layout differ between hospitals; the sequence does not.
1.Clinical history (or Indication)
One or two lines saying why the scan was ordered — the symptom, the injury, the follow-up interval. It is the question the radiologist was asked. It is also frequently abbreviated or incomplete, because it was typed by whoever booked the study, so do not read a diagnosis into it.
2.Technique
How the images were taken: the modality, which sequences or phases, which region, whether contrast was given. This section makes no claim about your health at all. Its practical use is telling you what the scan could and could not have seen.
3.Comparison
Which earlier imaging, if any, was on the screen alongside this one. “No prior comparison available” means only that this is the first study of its kind on file. When a prior does exist, the report can describe change over time, which is usually far more informative than any single image.
4.Findings
The inventory. The radiologist works through the anatomy region by region and describes how each part looks — including, and mostly, the parts that look normal. This is the longest section and the one that alarms readers, because a page of dense description reads like a page of problems when most of it is the opposite.
5.Impression (or Conclusion)
The answer. Two or three sentences in which the radiologist says what they believe matters, in order of importance, and what should happen next. If you read one section, read this one. Where the impression and the findings seem to disagree, the impression is the considered view.
Some reports carry an Addendum — text appended after the report was first signed, often after a second read, a correction, or a prior study that arrived late. An addendum supersedes what it corrects. Read it last and treat it as current.
Read it in this order, not top to bottom
1.Start at the impression
Read it twice. Note whether it calls anything new, and whether it asks for anything — a follow-up scan, a different test, a referral. A request for follow-up is a common, routine instruction and is not by itself bad news.
2.Go back to the clinical history
Now that you know the answer, check what the question was. A report reads very differently once you know the radiologist was looking for a kidney stone rather than screening an entire abdomen.
3.Skim the findings for the region that matters
Find the paragraph about the part of you that hurts, or the part the impression named. Most of the rest is the radiologist confirming that other structures look unremarkable.
4.Look up words, not sentences
Single terms — effusion, opacity, hyperintensity — have stable meanings you can learn once. Whole sentences do not translate: what they mean depends on the study and the person, and that is exactly what your doctor supplies.
5.Write down two questions
One about anything the impression asks for, and one about anything in the findings the impression did not repeat. If the impression left it out there is usually a reason, and hearing that reason stated is worth more than another hour of searching.
What the hedging language is actually doing
A large share of any report is careful, deliberately non-committal wording, and it is the part patients most often mistake for a verdict. Imaging shows shape, density and signal; it rarely proves cause. A radiologist who writes “consistent with” rather than “is” is describing the limit of what a picture can establish, not withholding something from you.
| The phrase | What it is doing |
|---|---|
| Unremarkable | Looks normal. The most reassuring word in the document, and the least reassuring-sounding. |
| No evidence of X | This study does not show X. Not the same as proving X absent, because every scan has a resolution limit and a field of view. |
| Cannot be excluded | The images do not rule the possibility out. Written for completeness, and used most often about things the radiologist considers unlikely. |
| Clinical correlation is recommended | The picture alone does not settle it; the finding needs to be read next to your symptoms and examination. Addressed to your doctor, not to you. |
| Nonspecific | Real, visible, and explainable by several ordinary things. Common in reports describing inflammation, ageing changes and old injuries. |
| Consistent with | Fits the suspected cause well. Stronger than “cannot be excluded”, weaker than a diagnosis. |
The practical rule: hedging tells you how confident the radiologist is, and nothing about how serious the thing is. A confidently worded finding can be trivial, and a heavily hedged one can be the sentence that matters most. Which is which is a clinical judgement, not a linguistic one.
Incidental findings, and why nearly everyone has one
Modern scanners see a great deal that has nothing to do with why you were scanned. A CT ordered for abdominal pain images the liver, kidneys, spine and lung bases on the way through, and something in that territory is very often mildly unusual: a simple cyst, a small nodule, a benign-looking lesion, degenerative changes appropriate for your age. These are incidental findings, and reporting them is expected even when the radiologist is confident they are harmless.
Most incidental findings need nothing at all. Some need one repeat scan at a stated interval to confirm they are not changing — which is what published follow-up frameworks such as the Fleischner Society criteria for lung nodules exist to standardise, so the decision is not made ad hoc for each patient. A recommendation for a repeat study in six or twelve months is usually the system working as designed, not a signal of concern.
What a report cannot tell you
- How you feel. Imaging and symptoms disagree constantly — dramatic-looking spine degeneration in someone with no pain, a clean study in someone in severe pain. Both are ordinary.
- What happens next. The impression may recommend, but the plan belongs to the doctor who ordered the scan and knows the rest of your history.
- Anything outside the field of view. A lumbar spine MRI is not a scan of the hips, and a report is silent about anatomy it did not image.
- Certainty. Every report is one expert reading of one set of images at one moment. Second reads exist, and disagreement between two competent radiologists on borderline findings is a documented, normal part of the discipline.
See this on your own scan
Upload your images (de-identified in your browser) and your radiologist’s report, and every finding is mapped onto the slice it came from, explained in plain language and scored — free, no card needed. Informational only — not a diagnosis.
Frequently asked questions
- Should I read the impression or the findings first?
- The impression. It is the radiologist’s conclusion, written after reviewing everything, and it states what they think matters and what should happen next. The findings section above it is a complete inventory of what was visible, normal structures included, so reading it first gives a far more alarming picture of an ordinary study than it deserves.
- Does “cannot be excluded” mean the radiologist thinks I have it?
- Usually the opposite. It is completeness language: the images do not positively rule the possibility out, so it is named. Radiologists write it most often about possibilities they consider unlikely but cannot disprove from a picture. It describes the limits of the scan, not the odds for you.
- My report is three pages long. Is that bad?
- No. Report length tracks how much anatomy was imaged and how thorough the radiologist is, not how much is wrong. A whole-abdomen CT covers many organs and each gets a sentence whether it looks normal or not. A short report on a small body part can carry far more serious news than a long one on a large one.
- Why does my report mention things I was not scanned for?
- Because they were inside the field of view. Anything the scanner imaged has to be described, so a chest CT may comment on the thyroid, the spine or the upper abdomen. These incidental findings are reported for completeness and most need no action beyond your doctor noting them.
- Can I get a second opinion on a radiology report?
- Yes, and it is routine. Independent second reads are a normal part of radiology, particularly before surgery or a major treatment decision. Ask the doctor who ordered the scan how to arrange one; you will usually need the images themselves, not only the report text.
Words on this page
Each of these has its own page in the report glossary.
Report glossary, A–Z →Sources
- Patient-facing procedure descriptions (MRI, CT, X-ray, ultrasound)RadiologyInfo.org — Radiological Society of North America & American College of Radiology
- ACR Appropriateness Criteria and practice parameters for reportingAmerican College of Radiology
- MRI scan — overview, how it is performed and resultsNHS (United Kingdom)
Keep reading
Read Your Scan is informational only — not a medical diagnosis, and not a substitute for a licensed radiologist or your doctor. This page explains what a report says and what a radiologist looks at; it cannot tell you what your own images show. If you have urgent symptoms, seek care.