TY - GEN
T1 - Soft-Prompt Tuning to Predict Lung Cancer Using Primary Care Free-Text Dutch Medical Notes
AU - Elfrink, Auke
AU - Vagliano, Iacopo
AU - Abu-Hanna, Ameen
AU - Calixto, Iacer
N1 - Funding Information:
Acknowledgement. This study received approval from the Ethics Committee of the Medical University of Graz (approval no. 30-146 ex 17/18) and was supported by a research grant from ERA PerMed within the project called “PreCareML”. The source codes of this study are publicly available in PreCareML/XAI Repository.
Funding Information:
Acknowledgements. Parts of this work were generously supported by a grant of the German Federal Ministry of Research and Education (01ZZ1802H).
Funding Information:
Acknowledgements. This work was funded by Research Fund Flanders (FWO fellowship 1S38023N) and supported by the Flemish government (through the AI Research Program) and Stichting MS Research (through a Monique Blom-de Wagt grant). We furthermore thank Professor Bénédicte Dubois, neurologist at UZ Leuven, for collecting the data that was used retrospectively in this work.
Funding Information:
Acknowledgement. This research was partially funded by the German Federal Ministry of Health as part of the KINBIOTICS project.
Funding Information:
Acknowledgments. This work was done and funded in the scope of the European Union’s Horizon 2020 research and innovation program, under project VALU3S (grant agreement no. 876852). This work has also received funding from UIDP/00760/2020. A publicly available dataset was utilized in this work. The data can be found at: https://hdl.handle.net/2268/191620.
Funding Information:
Acknowledgements. This work received joint funding from the European Regional Development Fund (ERDF), the Spanish Ministry of the Economy, Industry and Competitiveness (MINECO), the Carlos III Research Institute (grant no. PI19/00375), and received support from the Generalitat de Catalunya 2021 SGR 01125.
Funding Information:
the support. RT, SD have received funding from the Swiss National Science Foundation (SNSF), grant agreement No 200021 197021.
Funding Information:
Partly supported by FAPESP grants 2020/16543-7 and 2020/06443-5, and by Coor-dena¸cão de Aperfei¸coamento de Pessoal de Ńıvel Superior - Brasil (CAPES) - Finance Code 001. Carried out at the Center for Artificial Intelligence (C4AI-USP), supported by FAPESP grant 2019/07665-4 and by the IBM Corporation.
Funding Information:
Acknowledgements. This work was supported by the internship programme at the European Astronaut Centre (EAC), European Space Agency (ESA). The authors would like to thank dr. Guillauime Weerts and dr. Sergi Vaquer Araujo as the Space Medicine Team Leads and dr. Ulrich Straube for their support throughout this project. Mona Nasser’s research is partially supported by the National Institute for Health Research Applied Research Collaboration South West Peninsula. The views expressed in this publication are those of the author(s) and not necessarily those of the National Institute for Health Research or the Department of Health and Social Care or the European Space Agency.
Funding Information:
Acknowledgements. This work was supported by the Slovenian Research Agency Program Grant P2-0209. We would also like to thank the Slovenian National Institute of Public Health for their constructive cooperation.
Funding Information:
Supported by the Slovenian Research Agency grants P2-0209 and L2-3170.
Funding Information:
This study is co-funded by the Normandy County Council and the European Union (PredicAlert European Project - FEDER fund). Part of this work was performed using computing resources of CRIANN (Normandy, France). This work was performed using HPC resources from GENCI-IDRIS (Grant 2022-102446).
Funding Information:
Acknowledgments. This study received funding from the Region North Denmark Health Innovation Foundation. This study is also supported by the Poul Due Jensen Foundation.
Funding Information:
This work has been supported by project B-TIC-324-UGR20 FEDER/Junta de Andalucía and Universidad de Granada.
Funding Information:
Acknowledgments. This research was supported by the National Health Institute (NIH) under grant number R01DK122073.
Funding Information:
Acknowledgements. This work was financially supported by the Polish National Center for Research and Development grant number INFOSTRATEG-I/0022/2021-00, and carried out with the support of the Laboratory of Bioinformatics and Computational Genomics and the High Performance Computing Center of the Faculty of Mathematics and Information Science, Warsaw University of Technology.
Funding Information:
This work was funded by NHSX and the Innovate UK Regu-
Funding Information:
Acknowledgement. The work presented in this paper was supported in part by NIH grants R01EB032752 and R01DK131586. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of NIH.
Funding Information:
Acknowledgements. this work was partially funded by the CONFAINCE project (Ref: PID2021-122194OB-I00) by MCIN/AEI/10.13039/501100011033 and, as appropriate, by “ERDF A way of making Europe”, by the “European Union”, and by the GRALENIA project (Ref: 2021/C005/00150055) supported by the Spanish Ministry of Economic Affairs and Digital Transformation, the Spanish Secretariat of State for Digitization and Artificial Intelligence, Red.es and by the NextGenerationEU funding. This research was also partially funded by a national grant (Ref: FPU18/02220), of the Spanish Ministry of Science, Innovation and Universities (MCIU) and by a mobility grant (Ref: R-933/2021), of the University of Murcia.
Funding Information:
Supported by the Defense Advanced Research Projects Agency (DARPA) through Cooperative Agreement D20AC00002 awarded by the U.S. Department of the Interior, Interior Business Center. The content of the article does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.
Funding Information:
Acknowledgements. This research was supported by the University of Padova project C94I19001730001, by the Italian Ministry of Health grant RF-2016-02362405, and by the Italian Ministry of Education, University and Research (PRIN) grant 2017SNW5MB.
Funding Information:
Acknowledgment. This work was partially supported by National Science and Technology Development Agency grant no. P-20-52599, Faculty of Medicine Siriraj Hospital, Mahidol University grant no. R016536004, a grant from the Mahidol University Office of International Relations to Haddawy in support of the Mahidol-Bremen Medical Informatics Research Unit, a Study Group grant from the Hanse-Wissenschaftskolleg Institute for Advanced Study to Haddawy, a fellowship from the Hanse-Wissenschaftskolleg Institute for Advanced Study, and by a Young Researcher grant from Mahidol University to Su Yin.
Funding Information:
Acknowledgements. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No 860974. This publication reflects only the authors’ view, and the funding agencies are not responsible for any use that maybe made of the information it contains.
Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - We examine the use of large Transformer-based pretrained language models (PLMs) for the problem of early prediction of lung cancer using free-text patient medical notes of Dutch primary care physicians. Specifically, we investigate: 1) how soft prompt-tuning compares to standard model fine-tuning; 2) whether simpler static word embedding models (WEMs) can be more robust compared to PLMs in highly imbalanced settings; and 3) how models fare when trained on notes from a small number of patients. All our code is available open source in https://bitbucket.org/aumc-kik/prompt_tuning_cancer_prediction/.
AB - We examine the use of large Transformer-based pretrained language models (PLMs) for the problem of early prediction of lung cancer using free-text patient medical notes of Dutch primary care physicians. Specifically, we investigate: 1) how soft prompt-tuning compares to standard model fine-tuning; 2) whether simpler static word embedding models (WEMs) can be more robust compared to PLMs in highly imbalanced settings; and 3) how models fare when trained on notes from a small number of patients. All our code is available open source in https://bitbucket.org/aumc-kik/prompt_tuning_cancer_prediction/.
KW - Cancer
KW - Natural Language Processing
KW - Prediction models
UR - https://www.scopus.com/pages/publications/85163924870
U2 - 10.1007/978-3-031-34344-5_23
DO - 10.1007/978-3-031-34344-5_23
M3 - Conference contribution
C2 - 37313528
SN - 9783031343438
VL - 13897 LNAI
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 193
EP - 198
BT - Artificial Intelligence in Medicine - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023, Proceedings
PB - Springer Science and Business Media Deutschland GmbH
T2 - 21st International Conference on Artificial Intelligence in Medicine, AIME 2023
Y2 - 12 June 2023 through 15 June 2023
ER -
SDG 3 Good Health and Well-being