Optical Diagnosis: See It, Name It, Leave It, Prove It
Real-time histology is arguably the largest single value opportunity in everyday colonoscopy — and this session earns it the hard way: not advocacy, but audit. GIEQs shows its own evidence, then holds the field's biggest rollout to the meeting's standard.
Running order (indicative)
- Head
- Dr John Andersonanchors the studio throughout
- Talk 1
- Dr Maria Eva ArgenzianoThe GIEQs Algorithm: Blink, Demarcation, Vessels — Under Audited Competence
- Talk 2
- Prof Matthew RutterResect and Discard Under Scrutiny: By What Criteria Should It Be Judged?
- Talk 3
- Dr Pieter Jan PoortmansCan AI Rescue Resect and Discard?
- Live case
- Dr Lobke DesomerDr Mariana Figueiredo FerreiraColon with several polyps for live NICE/JNET optical characterisation
- Discussion
- Allremaining session time
The session, as written
Lecture 1 — The GIEQs algorithm: blink, demarcation, vessels — under audited competence. (Poortmans). A structured optical-diagnosis algorithm approaching a 400–500-patient publication, announced from its own stage: what audited competence looks like, and what a confident diagnosis is actually worth.
Lecture 2 — Resect and discard under scrutiny: by what criteria should it be judged? (Rutter). The English rollout, honestly examined: three studies that never reached their primary endpoints, a training programme steamrolled through anyway, and the surveillance time-bomb of hyperplastic polyps lumped with sessile serrated lesions. Judged by post-colonoscopy cancers, missed cancers and opportunity cost — not by enthusiasm.
Lecture 3 — Can AI rescue resect and discard? (Poortmans). AI supports optical diagnosis well distally, poorly for larger right-colon lesions — where machines genuinely help the strategy, and where they become a licence to over-surveil. Competence support, not a bypass.
Live case — predict before the reveal. Polyps characterised live; audience and operator commit to histology and surveillance interval before pathology settles it.
