Overview
For many radiotherapy departments, the integration of new technologies must complement rather than replace established planning processes. The Northern Centre for Cancer Care (NCCC) in Newcastle, United Kingdom, offers a valuable example of how MVision AI’s Dose+ can provide clinical value even within mature automated workflows.
Led by Dr. Samantha Warren and Lois Daniel-Black, the team presented a validation study comparing Dose+ predictions against both automated and manual planning in prostate cancer cases of varying complexity, using the RayStation® treatment planning system.
Study
The evaluation included 60 prostate cancer patients, divided into two cohorts:
- 30 standard prostate cases treated with 60 Gy in 20 fractions
- 30 complex prostate cases involving pelvic lymph nodes, treated with 60 Gy to the prostate and 47 Gy to the nodes in 20 fractions
All planning was performed using RayStation® TPS, providing a consistent environment to compare Dose+ predictions across clinical planning methods.
Evaluation Focus
Several key aspects of Dose+ performance were evaluated:
- Agreement between Dose+ predictions and an existing automated planning solution in standard prostate cases
- The model’s ability to support treatment planning efficiency in more complex node-involved cases
- The potential to reduce optimization workload by identifying cases requiring dose trade-offs early
- Dose+ applicability across both automated and manual workflows
This evaluation provided insight into the model’s clinical utility across a range of case types and planning environments.
Key Findings
The study identified several ways in which Dose+ added clinical value within both automated and manual planning workflows:
- Strong agreement with established automated plans in standard prostate cases
In the cohort of 30 standard prostate cancer patients (60 Gy in 20 fractions), Dose+ predictions closely aligned with NCCC’s automated planning solution:
◦ Rectum V48Gy: Clinical plans averaged 8.6% (range 1.1–16.8%) vs. Dose+ 9.6% (range 1.2–18.5%)
◦ Bladder V48Gy: Clinical plans averaged 13.0% (range 4.0–31.0%) vs. Dose+ 12.1% (range 3.7–28.0%) - Enhanced treatment planning efficiency in complex nodal cases
In the 30 complex prostate cases involving pelvic lymph nodes (60 Gy to prostate, 47 Gy to nodes in 20 fractions), Dose+ supported more efficient planning in scenarios that typically require experienced judgment to balance nodal coverage against bowel dose constraints (e.g., V48Gy < 28cc). - Accurate prediction of cases requiring coverage-sparing trade-offs
Dose+ identified approximately one-third of complex cases where compromises between target coverage and OAR sparing would be necessary — helping reduce trial-and-error and planning iterations. - Support for planning standardization amid rising clinical demand
With a recent doubling in prostate and node-involved cases at NCCC, the ability of Dose+ to integrate into existing workflows helped the department manage increased workload while maintaining consistent planning quality. - Validated utility across planner experience levels and workflow types
The study demonstrated Dose+ effectiveness across a range of planning complexities and user experience, enhancing both automated processes and supporting less experienced staff in manual planning contexts.
Conclusion
Dose+ integrates seamlessly into established clinical environments, adding value across both standard and complex prostate treatment workflows. This impact supports physicists, clinicians, and planners in different but meaningful ways.
For physicists, the study provides quantitative dose comparisons and agreement metrics. For clinicians, the tool’s predictive ability supports informed trade-offs in complex cases. For planners, Dose+ enhances manual and automated planning alike, helping manage increased volumes and clinical variation without increasing planning burden.
Overall, the study reinforces Dose+ as a clinically valuable tool for supporting efficient, consistent, and high-quality radiotherapy planning.