NSW Cancer Registry: What we did
In collaboration with the NSW cancer system, the Registry has developed and utilised new technologies to implement novel data engineering tools. These tools have allowed for a widely expanded population-based clinical data collection for all people diagnosed with cancer across the state, including rural and remote cancer notification centres, by the following means:
- Electronic reporting (ePath) of pathology reports has been implemented to supersede manual processing. ePath facilitates, streamlines, and automates the capture and transfer of notifications between laboratories and the Registry.
- Machine learning, a subfield of artificial intelligence (AI), is a system that can perform complex tasks in a way that is similar to how humans solve problem. AI cluster algorithms have been developed to group and standardise free text fields such as Radiotherapy Site Name and Course Name so that local health district resources are not required to maintain a manual mapping document at each site to enable these fields to be collected. Natural language processing software is being used to extract data automatically from electronic pathology reports (80 data items have been evaluated and 49 deemed suitable for semi-automatic collection).
- Dashboard templates (n = 16) have been developed in house and then configured per facility and handed over for customisation and management at the site. This provides radiotherapy facilities with a rapid view of the utilisation and treatment data at their centres, and allows them to monitor quality, identify erroneous data, and visualise issues as they happen.
Overall results
New technologies have not only increased the speed, efficiency and accuracy of data operations but also the completeness and reach of high quality cancer data, making it available state-wide for both public and private sector quality improvement, statistics and research.
