EXPOSED
The core of the job — building and optimizing radiation treatment plans on a TPS workstation — is exactly the target of knowledge-based planning, auto-contouring, and vendor auto-planning tools that already produce clinically usable plans in minutes for prostate, breast, and standard head-and-neck cases. What persists is plan evaluation under competing constraints (organ-at-risk tradeoffs in re-irradiation, unusual anatomy, brachytherapy and SBRT cases), second-check QA, and being the person in the room with the physicist and radiation oncologist when a plan is deviated from. The occupation isn't licensed at the state level in most places — CMD certification plus physician sign-off is the shield — so protection comes from institutional QA culture more than statute.
Mixed — a routine tier and a judgment tier. Beam angle selection, MLC segmentation, dose-volume optimization loops, and DVH review are all keystroke work inside Eclipse/RayStation/Monaco that vendor auto-planning and knowledge-based planning already reproduce for the standard prostate, whole-breast, and 3D palliative cases that fill most queues — the 8 reflects that the surviving irreducible work is narrower: fusing a re-irradiation case against prior dose, HDR applicator reconstruction, and catching a contour error the model propagated silently.
Some physical or field component. You are physically present for CT/MR sim on complex cases, in the vault for HDR applicator geometry and catheter reconstruction, and doing hands-on ion chamber/film measurements for patient-specific QA on some setups — but that is a fraction of a week otherwise spent seated at dual monitors, which is why this sits at 7 rather than up with therapists who position patients daily.
Certification preferred, not legally required. MDCB certification is the credential most employers require and Texas is nearly alone in licensing dosimetrists, so when a plan error reaches a patient the medical event report under 10 CFR 35.3045 lands on the authorized user physician and the medical physicist who signed the plan check — a 9 credits the real barrier that no hospital lets an uncertified person plan, without pretending you hold personal statutory liability.
Some relationship component. Patients almost never learn your name; the relationships that matter are internal ones — the oncologist who knows your judgment on cord tolerance, the physicist who trusts your second check — and those carry weight at a specific institution but are not the product being purchased, which caps this at 6.
Meaningful discretion. Choosing which OAR constraint to break when the PTV abuts brainstem, deciding whether a 2 mm shift on a fused prior plan is acceptable cumulative dose, and flagging a plan you believe is unsafe are genuinely ambiguous calls you make first — the 12 rather than higher is because the physician's signature and the physics plan check both sit downstream of every decision you make.
Orthodontists SAFE
Has AI actually changed your work?