Reproducibility of natural-language–driven statistical workflows compared with conventional software for diagnostic metrics in prostate cancer research
Investigative and Clinical Urology · 2026 · 67 (4) · 340
DOI: 10.4111/icu.20250642 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift Investigative and Clinical Urology, 2026. Verfasst von Horst Emanuel Lagos-Beitz und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Urology, Instituto Nacional de Ciencias Médicas y Nutrición "Salvador Zubirán", Mexico City, Mexico. Hauptthemen laut MEDLINE: Prostatic Neoplasms, Workflow, Software. Die Arbeit nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
PURPOSE: To evaluate whether a natural language-driven analytical workflow specified through a large language model (LLM) can accurately encode, execute, and reproducibly replicate a predefined statistical process traditionally performed in SPSS, and to assess its efficiency and auditability in urological research. MATERIALS AND METHODS: A dataset of 531 patients from a prostate cancer diagnostic study was analyzed in parallel using SPSS and ChatGPT's Data Analyst environment. Equivalent analytical procedures and variables were applied to calculate sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy. Concordance between methods was assessed using absolute deviation and reproducibility rates. Processing time and user workload were also compared. RESULTS: After two prompt-refinement steps to align analytical definitions, the LLM-based analysis achieved full concordance with SPSS across all diagnostic metrics, with a mean absolute deviation of
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Urology, Instituto Nacional de Ciencias Médicas y Nutrición "Salvador Zubirán", Mexico City, Mexico.
- Department of Urology, Instituto Nacional de Ciencias Médicas y Nutrición "Salvador Zubirán", Mexico City, Mexico. gerardo.tenag@incmnsz.mx.
ORCID-Profile: Horst Emanuel Lagos-Beitz, Guillermo Trujillo-Martínez, Carlos Antonio González-Quirós, Rodrigo Escalante-Armenta, Guillermo Martínez-Delgado, Ricardo Alonso Castillejos-Molina, Gerardo Tena Gonzalez-Mendez
Zitieren
APA
Lagos-Beitz, H. E., Trujillo-Martínez, G., González-Quirós, C. A., Escalante-Armenta, R., Martínez-Delgado, G., Castillejos-Molina, R. A., & Tena Gonzalez-Mendez, G. (2026). Reproducibility of natural-language–driven statistical workflows compared with conventional software for diagnostic metrics in prostate cancer research. Investigative and Clinical Urology, 67(4), 340. https://doi.org/10.4111/icu.20250642
Vancouver
Lagos-Beitz HE, Trujillo-Martínez G, González-Quirós CA, Escalante-Armenta R, Martínez-Delgado G, Castillejos-Molina RA, et al. Reproducibility of natural-language–driven statistical workflows compared with conventional software for diagnostic metrics in prostate cancer research. Investigative and Clinical Urology. 2026;67(4):340. doi:10.4111/icu.20250642
BibTeX
@article{LagosBeitz2026,
author = {Lagos-Beitz, Horst Emanuel and Trujillo-Martínez, Guillermo and González-Quirós, Carlos Antonio and Escalante-Armenta, Rodrigo and Martínez-Delgado, Guillermo and Castillejos-Molina, Ricardo Alonso and Tena Gonzalez-Mendez, Gerardo},
title = {Reproducibility of natural-language–driven statistical workflows compared with conventional software for diagnostic metrics in prostate cancer research},
journal = {Investigative and Clinical Urology},
year = {2026},
volume = {67},
number = {4},
pages = {340},
doi = {10.4111/icu.20250642}
}
EUVIMED