Could generative artificial intelligence replace fieldwork in pain research?

acute pain artificial intelligence climbing generative pretrained transformer

Journal

Scandinavian journal of pain
ISSN: 1877-8879
Titre abrégé: Scand J Pain
Pays: Germany
ID NLM: 101520867

Informations de publication

Date de publication:
01 Jan 2024
Historique:
received: 22 11 2023
accepted: 11 12 2023
medline: 7 3 2024
pubmed: 7 3 2024
entrez: 7 3 2024
Statut: epublish

Résumé

Generative artificial intelligence (AI) models offer potential assistance in pain research data acquisition, yet concerns persist regarding data accuracy and reliability. In a comparative study, we evaluated open generative AI models' capacity to acquire data on acute pain in rock climbers comparable to field research. Fifty-two rock climbers (33 m/19 f; age 29.0 [24.0-35.75] years) were asked to report pain location and intensity during a single climbing session. Five generative pretrained transformer models were tasked with responses to the same questions. Climbers identified the back of the forearm (19.2%) and toes (17.3%) as primary pain sites, with reported median pain intensity at 4 [3-5] and median maximum pain intensity at 7 [5-8]. Conversely, AI models yielded divergent findings, indicating fingers, hands, shoulders, legs, and feet as primary pain localizations with average and maximum pain intensity ranging from 3 to 4.4 and 5 to 10, respectively. Only two AI models provided references that were untraceable in PubMed and Google searches. Our findings reveal that, currently, open generative AI models cannot match the quality of field-collected data on acute pain in rock climbers. Moreover, the models generated nonexistent references, raising concerns about their reliability.

Sections du résumé

BACKGROUND BACKGROUND
Generative artificial intelligence (AI) models offer potential assistance in pain research data acquisition, yet concerns persist regarding data accuracy and reliability. In a comparative study, we evaluated open generative AI models' capacity to acquire data on acute pain in rock climbers comparable to field research.
METHODS METHODS
Fifty-two rock climbers (33 m/19 f; age 29.0 [24.0-35.75] years) were asked to report pain location and intensity during a single climbing session. Five generative pretrained transformer models were tasked with responses to the same questions.
RESULTS RESULTS
Climbers identified the back of the forearm (19.2%) and toes (17.3%) as primary pain sites, with reported median pain intensity at 4 [3-5] and median maximum pain intensity at 7 [5-8]. Conversely, AI models yielded divergent findings, indicating fingers, hands, shoulders, legs, and feet as primary pain localizations with average and maximum pain intensity ranging from 3 to 4.4 and 5 to 10, respectively. Only two AI models provided references that were untraceable in PubMed and Google searches.
CONCLUSION CONCLUSIONS
Our findings reveal that, currently, open generative AI models cannot match the quality of field-collected data on acute pain in rock climbers. Moreover, the models generated nonexistent references, raising concerns about their reliability.

Identifiants

pubmed: 38452184
pii: sjpain-2023-0136
doi: 10.1515/sjpain-2023-0136
doi:

Types de publication

Journal Article

Langues

eng

Sous-ensembles de citation

IM

Informations de copyright

© 2024 the author(s), published by De Gruyter.

Références

Stokel-Walker C, Van Noorden R. ChatGPT and generative AI’s impact on science. Nature. 2023;614(7947):214–6. 10.1038/d41586-023-00340-6 .
doi: 10.1038/d41586-023-00340-6 pubmed: 36747115
Schedlbauer J, Raptis G, Ludwig B. Web crawling, web scraping, and text mining for medical informatics labor market analysis. Int J Med Inform. 2021;150:104453. 10.1016/j.ijmedinf.2021.104453 .
doi: 10.1016/j.ijmedinf.2021.104453 pubmed: 33862508
Borji A. A categorical archive of ChatGPT failures. ArXiv; 2023. 10.48550/arXiv.2302.03494 .
doi: 10.48550/arXiv.2302.03494
Spitale G, Biller-Andorno N, Germani F. GPT-3 AI model (dis)informs better than humans. Sci Adv. 2023;9(26):eadh1850. 10.1126/sciadv.adh1850 .
doi: 10.1126/sciadv.adh1850 pubmed: 37379395 pmcid: PMC10306283
Moskatel LS, Zhang N. An observational, qualitative study on the utility of ChatGPT in assessing literature on migraine prevention. Front Neurol. 2023;14:1225223. 10.3389/fneur.2023.1225223 .
doi: 10.3389/fneur.2023.1225223 pubmed: 37662036 pmcid: PMC10469750

Auteurs

Suzana Bojic (S)

Department of Anesthesiology and Intensive Care, Faculty of Medicine, University of Belgrade, Dr. Subotica 8, 11 000 Belgrade, Serbia.
Department of Anesthesiology and Intensive Care, University Clinical Hospital Centre "Dr. Dragisa Misovic - Dedinje", Heroja Milana Tepica 1, 11 000 Belgrade, Serbia.

Nemanja Radovanovic (N)

Department of Anesthesiology and Intensive Care, University Clinical Centre of Serbia, 11 000 Belgrade, Serbia.

Milica Radovic (M)

Intensive Care Unit, University Clinical Hospital Center Zemun, 11 000 Belgrade, Serbia.

Dusica Stamenkovic (D)

Department of Anesthesiology and Intensive Care, Medical Faculty, University of Defense, Veljka Lukica Kurjaka 1, 11 000 Belgrade, Serbia.
Department of Anesthesiology and Intensive Care, Military Medical Academy, Crnotravska 17, 11 000 Belgrade, Serbia.

Classifications MeSH