Collective Intelligence In Practice: Real-Time Data Sharing to Support Radiotherapy Medical Physics
Abstract
Purpose
Clinical practice in radiotherapy medical physics is strengthened through shared experience and peer learning. Although peer-reviewed outputs and educational resources are essential, they are often insufficiently timely, specific, or practical for day-to-day implementation challenges, particularly for emerging techniques such as new dose calculation algorithms. Access to expert support and reassurance is especially limited for rural or geographically isolated centers. This study implements real-time data-sharing enabling direct peer comparison with the goal of providing additional practical support to radiotherapy centers.
Methods
A web-based application was developed to support a remote patient-specific quality assurance (PSQA) auditing study, designed to ensure direct comparability of results between participants operating matched linear accelerator models. Submitted results are processed immediately, enabling real-time benchmarking against peer centers. In addition, the application provides comparisons of local beam model parameters (e.g. MLC offset) against both an aggregated database and previously published reference data. The platform was designed to be scalable, supporting expansion to additional domains such as machine quality assurance data.
Results
To date, 176 PSQA audit results have been uploaded from 13 countries, allowing the center to compare their PSQA results with relevant data from others. A survey and other communications have revealed that at least five centers reevaluated their clinical protocol as a result. The system also contains 159 clinical or ex-clinical beam model parameter sets from 77 centers. Thirty-six of these beam models are using the recently implemented Eclipse v18 enhanced leaf model. There has been great interest in the developed data sharing/comparison platform, with >95% of Australasian centers contributing their data. To date, at least five centers have used the system to inform the parameter choice for their beam model.
Conclusion
Real-time data sharing has proven to be valuable in supporting radiotherapy medical physics, especially when implementing new technologies where guidance is currently limited.