{"stable_id":"b8243d473b7bd45c","slug":"prostate-ml-radiotherapy-planning","company":"Princess Margaret Cancer Centre / University Health Network","workflow_name":"ML-assisted treatment-plan generation and blinded physician selection","function_code":"clinical_ops","pattern_codes":["continuous_decisioning","creator_to_judge"],"changed_assumption":"A machine can generate a candidate plan, but physicians and peer review select the plan delivered.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"2 peer reviewed; publication outcomes are verified.","caveat_summary":"Single-center prospective deployment; 61% of ML plans were selected prospectively versus 83% in simulation.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Medical dosimetrist / physicist","action":"Manually specifies objectives and iteratively tunes a prostate radiotherapy plan.","actor_type":"control"},{"actor":"Treating physician","action":"Reviews the human-generated plan against standardized criteria.","actor_type":"human"}],"hinge":"A machine can generate a candidate plan, but physicians and peer review select the plan delivered.","after":[{"actor":"Random-forest planning system and human planner","action":"Generates an ML plan in parallel with the conventional human plan.","actor_type":"human"},{"actor":"Treating physician","action":"Blindly compares candidates and selects the clinically acceptable plan for delivery.","actor_type":"human"}],"decision_mode":"moved","decision_marker":"Selection moves from a fixed rule to the model"},"before":[{"order":1,"actor":"Medical dosimetrist / physicist","action":"Manually specifies objectives and iteratively tunes a prostate radiotherapy plan.","handoff_to":"Treating radiation oncologist","control":"Standard planning system, dose constraints, and peer review."},{"order":2,"actor":"Treating physician","action":"Reviews the human-generated plan against standardized criteria.","handoff_to":"Treatment delivery","control":"Physician and peer-review approval; median end-to-end baseline 118 hours."}],"after":[{"order":1,"actor":"Random-forest planning system and human planner","action":"Generates an ML plan in parallel with the conventional human plan.","handoff_to":"Blinded physician comparison","control":"Same clinical criteria and standardized peer review; two independent candidate paths retained."},{"order":2,"actor":"Treating physician","action":"Blindly compares candidates and selects the clinically acceptable plan for delivery.","handoff_to":"Treatment and peer-review workflow","control":"Physician owns acceptance; unacceptable ML plans are discarded for human plans."}],"decision_rights":"Treating physicians and peer review retain final plan selection and delivery authority.","exception_path":"An unacceptable ML plan is rejected in favor of the human-generated plan.","removed_work":["Some objective specification and iterative tuning"],"outcomes":[{"metric":"End-to-end planning time","baseline":"Median 118 hours","result":"Median 47 hours; 60.1% reduction","period":"Prospective 50-patient deployment","scale":"50 retrospective simulations plus 50 prospective patients","attribution_caveat":"Single-center prospective deployment; 61% of ML plans were selected prospectively versus 83% in simulation.","evidence_label":"verified"}],"executive_lesson":"Parallel human and machine plans can expose real adoption gaps that retrospective accuracy misses.","anti_pattern":"Do not publish simulated acceptance as the prospective adoption rate.","questions_for_leaders":["Where is the operating threshold set and who can override it?","What measured result would trigger rollback or retraining?","Which residual decisions must remain human-owned?"],"collections":[],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/prostate-ml-radiotherapy-planning"}