{"stable_id":"fb7af1efc542b6d1","slug":"gi-genius-colonoscopy-detection","company":"Three Italian endoscopy centers using GI Genius","workflow_name":"Real-time AI computer-aided adenoma detection during colonoscopy","function_code":"clinical_ops","pattern_codes":["creator_to_judge"],"changed_assumption":"That the endoscopist must be the only visual observer capable of flagging a lesion during withdrawal.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"1 peer reviewed; publication outcomes are verified.","caveat_summary":"Randomized detection endpoint; the trial did not establish reduced colorectal-cancer mortality, and much of the uplift was in small lesions.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Endoscopist","action":"Inspect the live colonoscopy feed and visually identify suspected lesions.","actor_type":"control"},{"actor":"Pathologist","action":"Determine whether removed tissue is neoplastic.","actor_type":"control"}],"hinge":"That the endoscopist must be the only visual observer capable of flagging a lesion during withdrawal.","after":[{"actor":"GI Genius CADe","action":"Processes the live video and superimposes a green box over a suspected lesion.","actor_type":"system"},{"actor":"Endoscopist","action":"Inspects the highlighted area and decides whether to remove tissue.","actor_type":"system"}],"decision_mode":"shared","decision_marker":"Human moves from creator to judge"},"before":[{"order":1,"actor":"Endoscopist","action":"Inspect the live colonoscopy feed and visually identify suspected lesions.","handoff_to":"Resection and pathology","control":"Clinical judgment and minimum withdrawal time"},{"order":2,"actor":"Pathologist","action":"Determine whether removed tissue is neoplastic.","handoff_to":"Clinical record","control":"Histopathology reference standard"}],"after":[{"order":1,"actor":"GI Genius CADe","action":"Processes the live video and superimposes a green box over a suspected lesion.","handoff_to":"Endoscopist","control":"Real-time computer-aided detection"},{"order":2,"actor":"Endoscopist","action":"Inspects the highlighted area and decides whether to remove tissue.","handoff_to":"Pathology","control":"Physician retains procedural decision rights"}],"decision_rights":"The AI draws attention; the endoscopist decides whether a finding is real and whether to resect.","exception_path":"The physician ignores false alerts, inspects missed or ambiguous tissue directly, and relies on histopathology for final classification.","removed_work":["Reliance on one unaided visual observer for every frame"],"outcomes":[{"metric":"Adenomas detected per colonoscopy","baseline":"0.71 mean adenomas per colonoscopy in the standard-colonoscopy control group","result":"1.07 in the CADe group; incidence-rate ratio 1.46, with no increase in withdrawal time","period":"September-November 2019 randomized trial","scale":"685 subjects at three Italian centers","attribution_caveat":"Randomized detection endpoint; the trial did not establish reduced colorectal-cancer mortality, and much of the uplift was in small lesions.","evidence_label":"verified"}],"executive_lesson":"The model changed the attention workflow rather than the clinical decision right; value came from watching every frame without slowing the procedure.","anti_pattern":"Calling every AI box a diagnosis or extrapolating detection uplift to mortality.","questions_for_leaders":["Does the AI change attention or authority?","What is the false-alert burden at operating speed?"],"collections":["human-still-decides","regulated-autonomy"],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/gi-genius-colonoscopy-detection"}