Explaining Imaging Findings to Patients in Plain Language

The problem

Patients are increasingly given access to their own radiology reports — through online patient portals, GP letters, or requests under data subject access rights. The problem is that these reports are written for radiologists and clinicians, not for the patients they concern.

A patient who reads "moderate osteoarthritis with joint space narrowing, osteophyte formation, and subchondral sclerosis at the medial tibiofemoral compartment" does not know what to think. They may catastrophise. They may dismiss it. They may arrive at their next appointment having spent three days convinced they need a knee replacement when you were planning conservative management.

A plain-language explanation before the patient reads the full report — or alongside it — reduces anxiety, improves the consultation, and saves clinic time spent managing misunderstanding.

How AI helps

An AI tool can translate a summary of radiology findings — not the full report, and never the identifiable document — into a plain-language patient explanation. You provide the key clinical findings in anonymised form. The tool drafts an explanation. You review it for accuracy and calibrate the tone for your patient.

This is particularly useful for MRI reports of common orthopaedic pathology: rotator cuff tears, meniscal tears, spinal degenerative changes, anterior cruciate ligament (ACL) tears, and osteoarthritis of major joints. These findings have standard explanations that the tool can produce reliably, subject to your clinical review.

A real example

Dr Serena is an orthopaedic registrar seeing a 52-year-old woman in a shoulder clinic. The patient had an MRI three weeks ago. The report describes a partial thickness articular surface tear of the supraspinatus tendon with mild subacromial bursitis. The patient has already seen this report on the GP's patient portal and has been frightened by the word "tear."

Dr Serena opens an AI tool and types:

Write a plain-language explanation of the following MRI shoulder finding 
for a woman in her early fifties who has already seen the report and is 
worried by the word 'tear'.

Findings: partial thickness articular surface tear of the supraspinatus 
tendon with mild subacromial bursitis.

Explain:
- What the supraspinatus is and what it does
- What a partial thickness tear means (not a complete rupture)
- What subacromial bursitis is
- That this finding is common in people her age and does not always 
  mean surgery is needed
- That the next step is a clinical assessment and discussion of options

Do not recommend a treatment. Tone: calm and reassuring without 
minimising. Maximum 250 words.

The tool produces a plain-language explanation. Dr Serena reads it. She adjusts one sentence that slightly overstates how common these findings are — accurate but she feels it could inadvertently minimise the patient's symptoms. She uses the explanation as a handout to give the patient before the consultation begins.

Try it yourself

Things to watch for

Accuracy requires your review. The tool can produce plausible-sounding explanations that are slightly inaccurate — particularly around what a finding means clinically for this patient. Read every sentence as a clinician, not just as a writer. An explanation that is technically accurate in the abstract may not be accurate for this patient's specific situation.

Do not use this as a substitute for a clinical explanation. The plain-language document supports the consultation — it does not replace the conversation where you explain what the finding means for this patient's treatment plan. Give it to the patient before or after the consultation, not instead of it.

The tool does not know the clinical context. A partial ACL tear in a 25-year-old athlete and the same finding in a 60-year-old sedentary patient have very different implications. The tool will explain the anatomy accurately but cannot calibrate the clinical significance for this individual. You do that.

Reassurance must be calibrated. The tool tends toward reassurance in its language. For findings that do warrant concern — a significant rotator cuff tear in a patient who has been symptomatic for six months — check that the plain-language explanation is honest about the clinical significance without being alarmist.

Remember: AI is a helpful assistant, not a clinician. You make the call.