Robert O'Neill

Packed auditorium as Chris Jones gets the afternoon underway #CamSOC2024

As always a really thoughful presentation of fundamental questions in organoid resarch by John Zhuang #CamSOC2024

The Cambridge Symposium on Oesophageal Cancer #camsoc2024 has kicked off! Thanks to Rob O’Neill for the welcome, we start with a talk from @RC_Fitzgerald ‘The past, present and future of early detection for oesophageal cancer’

How can you make probability distributions and cancer evolution accessible? ask Trevor Graham! #CamSOC2024

Delighted to open the inaugural Cambridge symposium on oesophageal cancer #CamSOC2024. A fantastic line up of speakers. Oustanding first talk @RFitzgerald_lab

@ProfW_edinsurg @ewenharrison @EdinSurg @edinsurg_online @SurgicalSaber @SurgData Wow Ewen, masssive congratulations!

Many congratulations to our colleague @ewenharrison who was awarded an OBE in the New Year's Honours list for his data science work supporting government committees during the COVID 19 pandemic. What a star🌟 @EdinSurg @edinsurg_online @SurgicalSaber @SurgData @UsherInstitute

A fantastic talk by @scarlettmcnally. Thanks!

@MrStuartAndrews Hi stuart, I am basically never on here now. get in touch on my email robertoneill@nhs.net

@Rob_UGIsurgeon Hi Rob, would like to get in touch about the National Hiatal Surgical Registry

Questions arising - Is this the case with a majority FLOT treated cohort (no reason not to think so), Is the benefit across all levels of Mandard TRG?

Two important papers in @BJSurgery using @Augishealth #NOGCA data. A prediction model for prognosis after gastric cancer surgery in a UK population and confirmation of the value of adjuvant chemo. Important data to inform UGI MDTs. Fantastic work @SaqRahman

@EwenGriffiths @salo75 @Cirbosque @LiangRhea @pferrada1 @SWexner @juliomayol @MISIRG1 @NeilFlochMD @DissanaikeMD @PipeCabreraV @CUH_NHS @AmCollSurgeons @ISDE_net @Augishealth @NIHRresearch But on occasion a thoracotomy is still needed for source control. This can usually be decided when scoping on table. Big question is what to do with the hole in the oesophagus. We have abandoned t-tubes/repairs/flaps in favour of endovac.

@EwenGriffiths @salo75 @Cirbosque @LiangRhea @pferrada1 @SWexner @juliomayol @MISIRG1 @NeilFlochMD @DissanaikeMD @PipeCabreraV @CUH_NHS @AmCollSurgeons @ISDE_net @Augishealth @NIHRresearch Its is fairly rare to have large amounts of solid contamination, most of the time it is liquid and can be washed away endoscopically or using a roth net. Early source control is critical and is achieved with the endoscope not the sponge. The sponge helps establish granulation.

Management algorithm for spontaneous esophageal perforation👇⤵️💥🔪🌎🌍! #SoMe4Surgery #MedTwitter @LiangRhea @pferrada1 @SWexner @juliomayol @misirg1 @NeilFlochMD @DissanaikeMD @PipeCabreraV @salo75

Presenting @Augishealth today on behalf of @ArrowStudy. A large, @RCSnews supported trainee-led collaborative audit. First six months of data with more to come!!! Keep it coming guys....@misshollyman @NatalieBlencowe @MarkarSheraz @TomWiggins23 @ShameenJaunoo

I am sure the hospitality will more than make up for the early start today on the way to @Augishealth #UGI2021. Looking forward to a fantastic meeting in Belfast #Nozoomtoday

Fantastic news. Well deserved

@TimTheSurgeon @MarkarSheraz @SurgJournal @Sivesh93 @SWexner This too. Make robust data collection part of routine cancer care and contribute to research projects.

@MarkarSheraz @SurgJournal @Sivesh93 @SWexner But it’s not quite as simple as that, because the best ML/AI models will learn and refine as cases are added, so not fixed or finite.

Conclusion after analysis >25,000 patients “Machine learning models and the integer-based risk score performed well in predicting risk of postoperative mortality in gastroduodenal ulcer perforation. These models will help in decision making.” @SurgJournal surgjournal.com/article/S0039-…

@MarkarSheraz @TimTheSurgeon @SurgJournal @Sivesh93 @SWexner External validation is not necessarily the answer. Population specific models may be more reliable than an over simplified universal model. In the case of AUGIS-surv this is more powerful because a large part of the total population data are used for the model.

@TimTheSurgeon @SurgJournal @Sivesh93 @SWexner Key will be testing external validity of this and other ML algorithms in a completely different dataset. Work in this area ongoing. External validity of ML algorithms is key step for clinical confidence / integration.