Science & Technology (Commonwealth Union) – Artificial Intelligence (AI) continues to expand into a wide variety of fields. These include utilizing less data to read medical reports, improving the hospitality sector by improving Concierge services and even enhancing the agricultural sector.
AI may soon play a key role in helping patients better interpret complex medical imaging results, according to a major new study from the University of Sheffield.
The study found that radiology reports for X-rays, CT scans and MRIs became significantly easier to understand when rewritten using advanced AI tools such as ChatGPT. Patients rated the AI-revised versions as nearly twice as clear as the original reports.
Researchers also found that the reading level of the reports dropped from university standard to a level more suitable for 11- to 13-year-old students.
Hospitals often process vast amounts of data in a variety of forms such as medical imaging to blood tests, ultra sounds. Medical staff are required to guide patients to the appropriate treatment options as they continuously update and process large amounts of data. This is where AI can show great potential.
The results indicate that AI-generated explanations could become a routine addition to medical reports, improving clarity, openness and patient confidence across healthcare systems, including the National Health Service (NHS) of the UK.
The research team analysed 38 studies published between 2022 and 2025, examining more than 12,000 radiology reports that had been simplified using AI. These revised reports were assessed by patients, members of the public and clinicians to measure both how well they were understood and whether they remained clinically accurate.
Radiology reports have historically been prepared with medical professionals in mind, not patients. But with the rise of patient-focused healthcare initiatives such as the NHS App, and new rules requiring greater openness around medical records, more patients are now able to view these reports directly.
The study’s lead author, Dr Samer Alabed, a Senior Clinical Research Fellow at the University of Sheffield and Honorary Consultant Cardio Radiologist at Sheffield Teaching Hospitals NHS Foundation Trust, indicated that the core problem is that these reports are not designed for patients.
He indicated that They are packed with specialist terminology and abbreviations that can be hard to interpret, which may cause unnecessary worry, misplaced reassurance, or simple confusion.
“Patients with lower health literacy or English as a second language are particularly disadvantaged. Clinicians frequently have to use valuable appointment time explaining report terminology instead of focusing on care and treatment. Even small time savings per patient could add up to significant benefits across the NHS.”
Clinicians who assessed the AI-rewritten reports found that the overwhelming majority were both accurate and thorough. However, about one per cent included mistakes, such as an incorrect diagnosis. This suggests that although the technology shows strong potential, it still requires close supervision.
Among the 38 studies analysed, none had been carried out in the UK or within NHS environments — a notable shortcoming that Dr Samer says the team is now working to address.
Dr Samer pointed out that the study has identified several major priorities, chief among them the need for real-world trials within NHS clinical workflows to properly evaluate safety, efficiency and patient outcomes.
He further indicated that it would involve human oversight systems, where clinicians check and approve AI-generated explanations before they are provided to patients and their aim is not to replace doctors, but to help deliver clearer, more compassionate and more equitable healthcare communication.
The findings reinforce the University’s commitment to turning innovative ideas into meaningful impact, reflecting its spirit of independent thought and collective ambition.
