This project involved a multi-disciplinary approach implementing new radiation oncology through Artificial Intelligence (AI) treatment software.
Radiation Oncology departments rely on a variety of specialties working together to provide high quality patient care. Embedded in this multidisciplinary team (MDT) environment are the complexities of radiation treatment planning and delivery. Once a radiation oncologist prescribes radiotherapy to a patient, the patient will undergo a planning Computed Tomography (CT) scan, which is then imported into treatment planning software to create a unique radiation treatment plan to be delivered to the patient. One key step in this process is the act of contouring, which is the process of anatomical delineation, undertaken by highly trained radiation therapists and radiation oncologists. This contouring step will ultimately contribute to the accuracy of dose delivered, as well as the way in which the plan is assessed for quality.
In early 2024 Radiation Oncology Princess Alexandra Raymond Terrace (ROPART) accepted, commissioned and implemented a state-of-the-art contouring software, MIM Maestro, which filled a technical gap in the existing treatment workflow. Such implementation required a robust commissioning approach from all teams involved, which included input from information technologists, physicists, oncologists and radiation therapists.
The commissioning process was split into a staged approach, with view to prioritise clinical implementation and improved processes. Through collaboration, MIM Maestro was successfully implemented in February 2024 after one month of planning, testing, system set up and infrastructure design, documentation and workflow discussions. The commissioning itself included regular MDT meetings and discussions to ensure each step was approached with consideration, and the entire team are very proud to now be working with MIM Maestro clinically.
A new workflow for the department would include exporting the patient CT scans directly from the CT scanners into MIM Maestro, where diagnostic MRI or CT/PET images are fused to the planning CT in a fraction of the time it would previously take to further assist in tissue delineation. Simultaneously, automatically generated contours would be added to the patient CT scan, through MIM Maestro’s Contour Protege AI contouring solution. The act of automatically generating patient contours for the internal anatomy of a patient is a significant time save for the planning team, but also removes inter-user variability. In anticipation of MIM Maestro entering in the department, a team of IT specialists, radiotherapists, physicists and oncologists was established to ensure a streamlined, collaborative approach to MIM implementation. Timelines and tasks were divided appropriately and adjusted when required, through regular MDT meetings. Further to this, all specialties were invited to the vendor provided training and a master project management excel sheet was used to keep track of progress throughout the project. Commissioning tests included infrastructure set up, server creation and system install and configuration, data management and application deployment, data integrity testing, workflow commissioning, training and educational resource development and documentation across all levels, which ultimately resulted in a robust, expedient process.
On all accounts, and from each specialty group, the MIM commissioning process passed all required tests. Any hurdles were dealt with on a case-by-case basis and implementation techniques were adapted as required. The first two weeks of patients were completed in February 2024 for the first oncology patient group, providing an opportunity for group reflection and adaption.
Now that the initial implementation of MIM has successfully begun and brought together a MDT with a shared vision for its clinical application, the opportunity exists to pursue further avenues of automation and workflow improvement using this technology. An early benefit of this technology has been the reduction of manual processes within typical clinical workflows that would previously have taken up valuable clinical time. However, this powerful program provides the scope to evaluate multiple parts of the clinical workflow, and with an established team now formed and collaborating effectively, these areas of future development can be discussed and developed with confidence in the team's ability to execute these plans.
