Integrating ChatGPT into Real-Time Multidisciplinary Oncology Discussions
Brazilian Journal of Oncology · 26.06.2026 · 22 (CP)
DOI: 10.1055/s-0046-1825344 ↗Kurz gefasst
Fachartikel in der Zeitschrift Brazilian Journal of Oncology, erschienen am 26.06.2026. Verfasst von Beatriz Favero Perez und 5 weiteren Autor:innen.
Abstract (Originalsprache)
All articles of this category (opens in new window) Address for correspondence Beatriz Favero Perez (bia.perez@hotmail.com)
Introduction
With advances in Oncology, from diagnosis to therapeutic options, multidisciplinary team (MDT) meetings have become essential in clinical decision-making. In parallel, the use of artificial intelligence (AI) has increased due to the growing volume of data and constant updates. Unlike prior studies that retrospectively assessed agreement between human decisions and algorithms, this prospective study proposes integrating AI as an active participant in real-time clinical discussions.
Objectives
To evaluate the impact of real-time AI use during MDT meetings and measure potential improvements in diagnostic hypotheses and clinical decisions.
Methods
ChatGPT was pre-programmed with a standardized prompt guiding how to process clinical data in two stages. During MDT meetings, case presentations were recorded in-app and submitted to the AI, which generated initial diagnostic hypotheses and therapeutic suggestions. These were not shared with the board at that point. In parallel, the specialists’ discussion was also recorded and submitted to the AI. Based on this input, ChatGPT reassessed its recommendations. Only after this second review was its final opinion revealed to the board. For each case, it was assessed whether the board's initial plan was maintained or modified following AI input. This model was implemented in April 2025 and remains active in a Clinical Oncology residency program.
Results
An interim analysis was performed after 25 cases, with average participation of 5 medical oncologists, 2 surgical oncologists, 1 radiation oncologist, 2 hematologists, and 10 Oncology/Hematology residents. Among cases requiring diagnostic workup, there was a 53% increase in diagnostic hypotheses after AI intervention. Broader treatment strategies were proposed in 52% of cases. Disagreement between professionals and AI occurred in 16% of discussions. In 63% of these, the AI changed its position and aligned with the board. Average time per case was 11.2 minutes.
Conclusion
This interim analysis suggests that real-time AI use in tumor boards can be a strategic ally in Oncology, expanding diagnostic reasoning and therapeutic planning amid growing clinical complexity. Publication History Article published online: 26 June 2026 © 2026. The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution 4.0 International License, permitting copying and reproduction so long as the original work is given appropriate credit (https://creativecommons.org/licenses/by/4.0/) Thieme Revinter Publicações Ltda. Rua Rego Freitas, 175, loja 1, República, São Paulo, SP, CEP 01220-010, Brazil Bibliographical Record Beatriz Favero Perez, Jean Marcelo Strippoli Doffini, Fabio de Azevedo Gonçalves, Claudia Vaz de Melo Sette, Auro del Giglio, Daniel de Iracema Gomes Cubero. Integrating ChatGPT into Real-Time Multidisciplinary Oncology Discussions. Brazilian Journal of Oncology 2026; 22. DOI: 10.1055/s-0046-1825344
Zitieren
APA
Perez, B. F., Doffini, J. M. S., Gonçalves, F. D. A., Sette, C. V. D. M., Giglio, A. D., & Cubero, D. D. I. G. (2026). Integrating ChatGPT into Real-Time Multidisciplinary Oncology Discussions. Brazilian Journal of Oncology, 22(CP). https://doi.org/10.1055/s-0046-1825344
Vancouver
Perez BF, Doffini JMS, Gonçalves FDA, Sette CVDM, Giglio AD, Cubero DDIG. Integrating ChatGPT into Real-Time Multidisciplinary Oncology Discussions. Brazilian Journal of Oncology. 2026;22(CP). doi:10.1055/s-0046-1825344
BibTeX
@article{Perez2026,
author = {Perez, Beatriz Favero and Doffini, Jean Marcelo Strippoli and Gonçalves, Fabio de Azevedo and Sette, Claudia Vaz de Melo and Giglio, Auro del and Cubero, Daniel de Iracema Gomes},
title = {Integrating ChatGPT into Real-Time Multidisciplinary Oncology Discussions},
journal = {Brazilian Journal of Oncology},
year = {2026},
volume = {22},
number = {CP},
doi = {10.1055/s-0046-1825344}
}
EUVIMED