Session Item

In 2009, a group of Danish researchers saw the need to collect and store radiotherapy(RT) plans for clinical studies. In the following years, the system – DcmCollab – was developed and implemented with the support of clinical researchers from all of the Danish RT centres. Since then, DcmCollab has been under continous development with features requested by the users, as well as upgrades to support the stability and security of the system.
The vision for DcmCollab is to support easy and secure collection of comprehensive RT data from all Danish patients(pts) treated with RT, and facilitate the re-usability of and access to data.

The DcmCollab system is based on a central SQL database, a DICOM server, and a web server for user interaction. A pre-existing dedicated national health computer network (SDN) facilitates secure and direct connection to the system from the treatment planning systems(TPS) at each Danish RT centre.
Data submitted to DcmCollab are not initially linked to any scientific project and are only accessible to users from the submitting centre. For each project, permission to view data from other centres is granted on a per-user basis (fig. 1) and a number of settings can be applied, including structure name mapping, and dose and delineation statistics. Furthermore, the system provides independently calculated dose volume histograms(DVH) and an RT plan viewer.

Presently, data from >11,000 pts included in 58 research projects are collected. Fig. 2 shows a cumulative distribution of treatment dates for pts included in the DcmCollab. Over the years, the inflow of data has been steady.
 
Development of DcmCollab has focused on both adding features and on improving security. The main features now are:
•    High focus on security
•    Easy data submission directly from the local TPS
•    Web upload of data, supporting international participation
•    Storage of all raw data submitted
•    Individual structure name mapping schemes for each research project
•    Display of RT plans and DVH
•    Creation of anonymized data for quality assurance(QA), e.g. delineation audits

The DcmCollab system was created to support and facilitate RT research by storing and managing complete RT planning datasets for national projects and thereby provide a powerful basis to facilitate research.
The aforementioned features have been fundamental for the success of DcmCollab, evident by the large amount of data uploaded. Similar features should be considered when developing a similar system. Additionally, the national unique personal identifier, the CPR-number, has facilitated linking data from DcmCollab to other databases, and the SDN network is the basis for the direct data submission from the local TPS.
The system has matured over time, and has now become a unique and invaluable database of detailed, high quality RT data. Furthermore, the system has become a valued tool for transferring data, generating datasets for audit trials, and performing QA of data in ongoing trials.