Background
Radiation therapy clinical trials generate data including pre-treatment, intra-treatment, and post-treatment images, dose files and treatment logs. These data can be utilised to develop and assess future radiation therapy innovations using analysis methods including statistical techniques and machine learning algorithms. Currently, no streamlined workflow exists to transfer, organise and manage these large multi-parametric datasets in radiotherapy.
Aims
This work describes an open-source project for transferring, storing, organising and analysing clinical trial data into a well organised, secure and usable central searchable service. The method extracts useful features from the data and provides a role-based access to analyse clinical data for radiation therapy applications.
Methods
The database solution utilises a cloud-based RESTful API service and provides client libraries for various software platforms with a secure role-based authentication system granting access via a site-to-site VPN technology. The technology stack of this solution tried and tested database back-ends coupled with cloud-based lightweight RESTful API techniques to provide a useful toolkit to analyse the data. The clinical data are stored in a secure, enterprise-grade with strict access controls, which contains past as well as ongoing trial data.
Results
The solution deployed at the University of Sydney enabled easy integration for four ongoing multi-institutional clinical trials as well as new deep learning and data visualisation efforts with existing data. The code is available for the medical physics community to implement and use: https://github.com/Image-X-Institute/The-Real-Time-Imaging-Database/tree/main.
Conclusions
A flexible platform has been developed for clinicians and researchers to manage and analyse clinical trial data for radiotherapy applications via an easy-to-use and secure interface.