TrialMatchAI: An End-to-End AI-powered Clinical Trial Recommendation System to Streamline Patient-to-Trial Matching
Authors/Creators
-
Abdallah, Majd
(Project leader)1, 2
-
Nakken, Sigve
(Project member)3
-
Bierkens, Mariska
(Project member)4
-
galvis, johanna
(Project member)5
-
Groppi, Alexis
(Project member)6
-
Karkar, Slim
(Project member)6
-
Meiqari, Lana
(Project member)7
-
Rujano, Maria Alexandra
(Project member)8
-
Canham, Steve
(Project member)
-
Dienstmann, Rodrigo
(Project member)
-
Fijneman, Remond
(Project member)4
-
Hovig, Eivind
(Project manager)9, 3
-
Meijer, Gerrit
(Project member)10, 11
-
Nikolski, Macha
(Project leader)5, 2
- 1. IBGC
- 2. Cbib
-
3.
Oslo University Hospital
-
4.
The Netherlands Cancer Institute
-
5.
Institut de Biochimie et Génétique Cellulaires
-
6.
Université de Bordeaux
-
7.
Vrije Universiteit Amsterdam
-
8.
European Clinical Research Infrastructure Network
- 9. University of Oslo
-
10.
University Medical Center Utrecht
- 11. Netherlands Cancer Institute
Description
This repository contains the supplementary materials for the paper titled "TrialMatchAI: An End-to-End AI-powered Clinical Trial Recommendation System to Streamline Patient-to-Trial Matching." It includes:
-
Synthetic patient dataset used for evaluation (Ideal Candidates)
-
Fine-tuned Large Language Models (LLMs) for tasks: named entity recognition (NER), trial re-ranking, and Chain-of-Thought (CoT) reasoning
-
Training and fine-tuning data used to develop the models
-
Matching results on synthetic patient profiles (Ideal Candidates + TREC 2021 & 2022)
-
Normalization dictionaries used for standardizing the extracted biomedical entities
These resources are provided to facilitate reproducibility and further research on AI-driven clinical trial matching systems.
Files
Restricted
The record is publicly accessible, but files are restricted. <a href="https://zenodo.org/account/settings/login?next=https://zenodo.org/records/15045515">Log in</a> to check if you have access.
Additional details
Identifiers
Funding
Software
- Repository URL
- https://github.com/cbib/TrialMatchAI
- Programming language
- Python , Java
- Development Status
- Active