Scaling AI Contouring Across an International Cancer Network

"AutoContour easily adapted into our current workflow, produces high-quality contours, and has great functionality in image registration."

This month, we’re spotlighting a collaboration between two important roles within Icon Group (Icon), Huong Nguyen, Group Manager - Icon Plan, and Aidan Leong, Group Manager - Radiation Therapy Education and Training. Icon and Radformation recently announced a strategic partnership focusing on collaboration and advancing radiation therapy software solutions. Huong and Aidan share the adoption and integration of AutoContour across their extensive network, showcasing a strategic approach to technology selection, deployment, and ongoing education.
 
Icon’s extensive network, encompassing 50 plus cancer centers across Australia, New Zealand, Singapore, Mainland China, Hong Kong, and Malaysia with over 50 Linacs and more than 300 radiation therapy staff members, presents a formidable challenge when selecting and implementing a software solution. That’s where Huong and Aidan shine in their critical roles and were pivotal in navigating these complexities.  
 
Huong spearheaded Icon’s assessment of multiple AI contouring solutions with criteria focusing on contour quality, adherence to contouring protocols, range of structures identified, diagnostic imaging capability, efficiency gains, cost implications, and workflow integration. Based on these criteria, Icon chose AutoContour as their automated contouring solution. Underscoring the strategic advantages of the solution, Huong details, “AutoContour has high-quality contours, great flexibility to create structure templates (including high resolution structures), great functionality in regard to image registration, and a streamlined process for contour checking prior to import into TPS.” 
 
AutoContour was quickly adapted to fit Icon’s current workflow, making implementation and network-wide adoption a massive but manageable operation. As Head of Radiation Therapy Education and Training, Aidan developed a comprehensive education program tailored to the unique requirements of Icon’s international network, empowering Icon to leverage AutoContour’s full potential. His focus on ensuring a deep understanding of grasping both the principles and technical intricacies of AI technologies, among the radiation therapy staff is essential for leveraging AutoContour’s full potential. He shares, “Our network’s rollout of AutoContour represents a significant milestone in the incorporation of AI into standard clinical workflows. Central to this rollout has been consideration of how we can most effectively support our RT workforce to understand the underlying principles as well as the technical operation of a deep learning platform like AutoContour. I see this as being key to empowering RTs to leverage the maximum benefit of AI tools while ensuring safety and efficiency.”

Since the integration of AutoContour, Icon has quickly noticed a positive impact in their planning workflow. Huong explains her team’s new process, “AutoContour is led by the RTs and completed at the time of CT import, which allows all OARs to be present for the Radiation Oncologist when they draw their targets.”  When referring to the noteworthy improvements, Huong shares, “Overall, we have found a reduction in planning time, improved contour consistency, and an increase in the number of OARs for dose reporting.” 
 
Feedback from Aidan's team echoes resoundingly positive sentiments. He shares, “We’ve received incredibly positive feedback throughout the implementation process of AutoContour. This goes not only for the RTs themselves who are responsible for the majority of OAR delineation in Australia/New Zealand, but also for clinicians who have reported high value from the availability and accuracy of AI-generated structures.”  Leveraging lessons learned from the New Zealand rollout, Aidan navigated the complexities of scaling implementation across the remaining centers over an impressive four week timeframe, all while maintaining optimal user experience. Aidan explains, “A key factor to this success has been the framework of partnerships facilitated between centres in each region, initially to support the cascade of training throughout teams but then to consolidate user experience and guide workflow refinements.”
 
Through effective communication and intentional training initiatives, Huong and Aidan have ensured widespread AutoContour acceptance and proficiency throughout Icon’s Radiation Oncology departments. Their collective efforts and strategic approach to integrating AutoContour into their network exemplifies Icon's commitment to leveraging advanced technologies to enhance clinical outcomes and operational efficiencies.

In her free time, Huong spends her days with her kids at the park and the beach, enjoying the natural beauty outside. She’s also a Eurovision fanatic and loves supporting Australia in the singing contest!

Outside of work, "like many Kiwis," Aidan tries to get out to enjoy the beautiful nature in Aotearoa, New Zealand. These days, he often coaxes his young kids along trails (or chases after them!).