🤖 Generated with AI assistance · Does not replace individualized medical advice.
Source: PubMed PMID 42671701 · The international journal of cardiovascular imaging (2026)
Score: 5.0/10 · Generated: 2026-09-01T07:28:10.484734 · Engine: qwen36-dsv4 local
AI transparency notice: This article was generated with artificial intelligence assistance. The information is for educational purposes and does not constitute medical advice.
Keyfacts
- Study: 75 outpatients at two Italian tertiary centers.
- Intervention: ChatGPT-4o generated simplified explanations and clinical recommendations from physician-dictated reports.
- Key results: Experts rated accuracy (4.63 ± 0.88) and completeness (4.42 ± 1.06) with excellent agreement (ICC 0.95-0.99) [DOI: 10.1007/s10554-026-03813-7].
- Patient impact: AI-enhanced versions scored significantly higher in all domains (p < 0.001), with overall satisfaction of 8.87 ± 1.11 vs. 6.56 ± 2.22 (+35%) [DOI: 10.1007/s10554-026-03813-7].
- Safety: High-risk recommendations were infrequent but not negligible (1.3% in patient content, 6.7% for physicians) [DOI: 10.1007/s10554-026-03813-7].
Context: Why it matters in oncology
Oncology patients increasingly access their electronic health records, facing dense imaging reports full of technical jargon. Poor comprehension can generate anxiety, misunderstandings about prognosis, and difficulties in shared decision-making. In oncology, where CT, MRI, or PET-CT reports guide vital decisions (surgery, chemotherapy, immunotherapy), clarity is as crucial as accuracy. This study, although focused on cardiology, offers a validated model for adapting imaging findings communication to accessible language while preserving clinical integrity.
Study details
Researchers processed 75 cardiac MRI reports with ChatGPT-4o, generating two products: a simplified explanation for the patient and personalized clinical recommendations. Two expert cardiologists evaluated the accuracy and completeness of the explanations; three additional cardiologists assessed the