To evaluate GPT-5 for grading cochlear and vestibular endolymphatic hydrops (EH) on delayed gadolinium-enhanced 3D-FLAIR MRI in Ménière's disease (MD) and to assess prompting strategies and human-AI collaboration.
✦ The floor
Discussion
Signed responses from readers of the wire.
Findings are preliminary for clinical adoption; audiologists and ENT teams should monitor this space, but no change to current MRI reporting or Ménière's disease diagnostic workflows is warranted yet.
If validated, AI-assisted grading of endolymphatic hydrops on MRI could standardise and accelerate diagnosis of Ménière's disease, reducing reliance on scarce specialist radiologist expertise.
- 01GPT-5 was evaluated for grading cochlear and vestibular endolymphatic hydrops on delayed gadolinium-enhanced 3D-FLAIR MRI.
- 02Study published in European Radiology (PubMed PMID 42706360, DOI 10.1007/s00330-026-12848-2).
- 03Human-AI collaboration model was the focus, not fully autonomous AI grading.
- 04Endolymphatic hydrops (excess inner-ear fluid) grading is a key step in confirming Ménière's disease diagnosis.
- 05Results could inform future standardisation of MRI-based Ménière's disease assessment.
GPT-5 can assist in grading endolymphatic hydrops on delayed gadolinium-enhanced 3D-FLAIR MRI in Ménière's disease patients.
studypartially supportedHuman-AI collaboration improves MR grading accuracy or consistency for endolymphatic hydrops.
studyunclear- PMID
- 42706360
- DOI
- 10.1007/s00330-026-12848-2.
- Journal
- European Radiology
- Publication type
- research_article
- Evidence level
- 2b
- Population
- Patients with Ménière's disease undergoing delayed gadolinium-enhanced 3D-FLAIR MRI
- Intervention
- GPT-5 AI-assisted grading of endolymphatic hydrops on MRI
- Comparator
- Human radiologist grading
Primary outcomes
Accuracy of cochlear endolymphatic hydrops grading; Accuracy of vestibular endolymphatic hydrops grading