EUVIMEDEUVIMED entdecken
Crossref

Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports

Connor J Mayes, Chloe Reyes, Mia E Truman, Christopher A Dodoo, Cameron R Adler, Imon Banerjee, Ashish Khandelwal, Lauren F Alexander, Shannon P Sheedy, Cole P Thompson, Jacob A Varner, Maria Zulfiqar und 1 weitere

Abdominal Radiology · 27.06.2025 · 51 (1) · 513-520

DOI: 10.1007/s00261-025-05079-4 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Abdominal Radiology, online erschienen am 27.06.2025. Verfasst von Connor J Mayes und 12 weiteren Autor:innen. Beteiligte Einrichtung: Mayo Clinic College of Medicine and Science, Phoenix, USA. Hauptthemen laut MEDLINE: Tomography, X-Ray Computed, Radiography, Abdominal, Radiology Information Systems, Artificial Intelligence. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 21 Literaturangaben. Gefördert von: Radiological Society of North America.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE: Radiology reports are essential for communicating imaging findings to guide diagnosis and treatment. Although most radiology reports are accurate, errors can occur in the final reports due to high workloads, use of dictation software, and human error. Advanced artificial intelligence models, such as GPT-4, show potential as tools to improve report accuracy. This retrospective study evaluated how GPT-4 performed in detecting and correcting errors in finalized radiology reports in real-world settings for abdominopelvic computed tomography (CT) reports. METHODS: We evaluated finalized CT abdominopelvic reports from a tertiary health system by using GPT-4 with zero-shot learning techniques. Six radiologists each reviewed 100 of their finalized reports (randomly selected), evaluating GPT-4's suggested revisions for agreement, acceptance, and clinical impact. The radiologists' responses were compared by years in practice and sex. RESULTS: GPT-4 identified issues and suggested revisions for 91% of the 600 reports; most revisions addressed grammar (74%). The radiologists agreed with 27% of the revisions and accepted 23%. Most revisions were rated as having no (44%) or low (46%) clinical impact. Potential harm was rare (8%), with only 2 cases of potentially severe harm. Radiologists with less experience (≤ 7 years of practice) were more likely to agree with the revisions suggested by GPT-4 than those with more experience (34% vs. 20%, P = .003) and accepted a greater percentage of the revisions (32% vs. 15%, P = .003). CONCLUSIONS: Although GPT-4 showed promise in identifying errors and improving the clarity of finalized radiology reports, most errors were categorized as minor, with no or low clinical impact. Collectively, the radiologists accepted 23% of the suggested revisions in their finalized reports. This study highlights the potential of GPT-4 as a prospective tool for radiology reporting, with further refinement needed for consistent use in clinical practice.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Tomography, X-Ray ComputedRadiography, AbdominalRadiology Information SystemsArtificial IntelligenceDiagnostic ErrorsQuality ImprovementHumansRetrospective StudiesFemaleMaleMiddle Aged

Stichwörter der Autor:innen

Artificial intelligenceGPT-4RadiologyRadiology reports

Autor:innen und Institutionen

  • Mayo Clinic College of Medicine and Science, Phoenix, USA.
  • Mayo Clinic, Phoenix, USA.
  • Mayo Clinic Scottsdale, Scottsdale, USA.
  • Mayo Clinic Rochester, Rochester, USA.
  • Mayo Clinic Jacksonville, Jacksonville, USA.
  • Mayo Clinic, Phoenix, USA. Tan.Nelly@mayo.edu.

Förderung

  • Radiological Society of North America

Zitieren

APA

Mayes, C. J., Reyes, C., Truman, M. E., Dodoo, C. A., Adler, C. R., Banerjee, I., Khandelwal, A., Alexander, L. F., Sheedy, S. P., Thompson, C. P., Varner, J. A., Zulfiqar, M., & Tan, N. (2025). Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports. Abdominal Radiology, 51(1), 513-520. https://doi.org/10.1007/s00261-025-05079-4

Vancouver

Mayes CJ, Reyes C, Truman ME, Dodoo CA, Adler CR, Banerjee I, et al. Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports. Abdominal Radiology. 2025;51(1):513-520. doi:10.1007/s00261-025-05079-4

BibTeX
@article{Mayes2025,
  author = {Mayes, Connor J and Reyes, Chloe and Truman, Mia E and Dodoo, Christopher A and Adler, Cameron R and Banerjee, Imon and Khandelwal, Ashish and Alexander, Lauren F and Sheedy, Shannon P and Thompson, Cole P and Varner, Jacob A and Zulfiqar, Maria and Tan, Nelly},
  title = {Improving radiology reporting accuracy: use of GPT-4 to reduce errors in reports},
  journal = {Abdominal Radiology},
  year = {2025},
  volume = {51},
  number = {1},
  pages = {513--520},
  doi = {10.1007/s00261-025-05079-4}
}
Kostenlos starten →