The PSI has medical accelerators to treate cancer patients by targeting the specific cancer cells. These machines produces an enourmous amount of log-files, when special occurences such as crashes happen. The dashboard visualization tool for medical accelerators was develop to help people at PSI understand the log-files generated by their medical accelerators.
The goal of this project was to develop a dashboard visualization tool for medical accelerators at PSI. The tool should help people at PSI understand the log-files generated by their medical accelerators. Important for the client is to see, how these signals in the logfiles change over time, so they can take conclusion and see trends. In the end the tool can help find faults in the machines, so they can be fixed for future uses.
The current solution could not be used to look at the log-files in detail and see changes over time. It implemented log file analysation more as an after thought and would also frequently crash, which would render it not very usable to the client. This means that these generated log-files were not at all interpretable and more or less useless to the client.
A dashboard application was implemented using dash-app, a python framework for creating dashboard visualization tools. The dashboard could show when an incident happened and it let the user look at a specific incident to check what actually happened in this log. The dashboard also featured various interactive featres, such as date, machine and mastership-filtering. All these interactive features together with the implement visualizations helped the client to understand the log-files generated by their medical accelerators. Test-results showed that the application succeeded on all goals it tried to reach. Feedback from the client also confirmed they are very happy and think it will be a very useful tool in their daily work.
Bachelor Thesis, Project length: 1 Semester, Hours of work: 12 ETCS (360 STD) per person, Team size: 2
Paul Scherrer Institut (PSI)
Leticia Fernández Moguel (leticia.fernandezmoguel@fhnw.ch), Arzu Cöltekin (arzu.coltekin@fhnw.ch)
