Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Recolector de Cienci...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Recolector de Ciencia Abierta, RECOLECTA
Bachelor thesis . 2025
License: CC BY NC ND
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
versions View all 2 versions
addClaim

Fiabilidad diagnóstica de modelos de lenguaje basados en Inteligencia Artificial

Authors: Zhou Xu, Qi-Heng;

Fiabilidad diagnóstica de modelos de lenguaje basados en Inteligencia Artificial

Abstract

Este trabajo analiza la fiabilidad diagnóstica de tres modelos de inteligencia artificial basados en procesamiento del lenguaje natural: ChatGPT-4o, DeepSeek-V3 y Grok-3. Se evaluaron sus capacidades mediante 20 casos clínicos seleccionados de The New England Journal of Medicine, aplicando seis prompts en castellano e inglés, con el fin de valorar su precisión en el diagnóstico diferencial, diagnóstico definitivo y grado de confianza en sus respuestas. Se emplearon pruebas estadísticas como el test Q de Cochran, test de McNemar y chi-cuadrado. Los resultados muestran que todos los modelos presentan un buen rendimiento en la generación de diagnósticos diferenciales (hasta un 90 % de aciertos), pero disminuyen notablemente en la precisión del diagnóstico único (en torno al 50–60 %). ChatGPT mostró un rendimiento estable en ambos idiomas, mientras que DeepSeek y Grok fueron más precisos en inglés. No se encontraron diferencias estadísticamente significativas entre los modelos ni entre idiomas.

Grado en Medicina

32 p.

Country
Spain
Related Organizations
Keywords

Modelos de lenguaje, Diagnóstico médico, ChatGPT, Grok, Medicina, Medicine, DeepSeek, Inteligencia artificial

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green
Related to Research communities
Spanish Research Portal