Radiation oncology
Credential the judgement around the output, not the software
AI-assisted segmentation, registration, synthetic imaging, automated planning and adaptive decision gates all leave a named person accountable for the output. These starter packages build the judgement around that output, for each role.
Accept, edit, reject or escalate
Each package is built around that decision, with benchmark sets that contain deliberate failures, a delegation directive template, checklists, a register and maintenance rules. Acceptance criteria, autonomy levels, out-of-distribution triggers and monitoring intervals are yours to set.
A package for each role
Each package holds the T0 to T5 learning journey, the competency standard for that profession, a credentialling checklist and assessment forms. Everything is a starting point for your credentialling committee to set and ratify before use.
Radiation therapist
Accept, edit, reject and escalate.
Covers reviewing AI output at the standard you would apply to a colleague’s work, recognising errors that look anatomically convincing, spotting patients the model was not built for, and using the fallback pathway without hesitation.
See the packageRadiation oncologist
What may be accepted, by whom, and when to stop.
Covers setting what may be accepted without further review, owning the out-of-distribution list, recognising output that is plausible but wrong, and judging when a patient falls outside what the model can be trusted on.
See the packageMedical physicist
The judgement around the output, not the software.
Covers AI-assisted segmentation, registration and propagation, synthetic imaging, automated planning and adaptive decision gates, with the competency standard built around the accept, edit, reject and escalate decision.
See the packageThe tiers behind every package come from the AyniHealth credentialling methodology.
Talk to us about AI-enabled workflow
We can show how the packages fit your service and what it takes to set them up.