Thomas Day

Our cardiology consultants @ajayding @thomasgday and @jmsimpsonmd were on the editorial team of the first ever Oxford Handbook of Paediatric Cardiology! They worked with a team of 31 expert authors (21 of them from across @GSTTnhs) to write this important handbook ⬇

Great chat with @ruskin147 about our recent work on AI to help fetal ultrasound screening at @GSTTnhs @EvelinaLondon @KingsImaging rorycellanjones.substack.com/p/can-ai-cut-b…

Delighted to be awarded eur500 prize for best fetal oral abstract at @AEPCcongenital conference in Porto, for work with @jmsimpsonmd @BernhardKainz1 and Reza Razavi. Full paper now published: doi.org/10.1002/uog.27…

Available for pre-order now! Our new handbook of paediatric cardiology. Really proud of this after a few years of work, with an amazing group of colleagues global.oup.com/academic/produ… @ajayding @jmsimpsonmd @GSTTnhs @OUPMedicine

Lots of exciting things happening in the first couple of weeks in July…. @thomasgday @jmsimpsonmd @EvelinaLondon

Strong representation of young researchers in in fetal cardiology from @EvelinaLondon in abstract session @AEPCcongenital. Great talks on AI from @thomasgday, fetal ECG from Sian Chivers and long term TGA outcomes from Peter Lillitos @jmsimpsonmd @trisha_vig

Delighted to be awarded this fellowship from @Kings_CATO, allowing me to continue researching the use of AI to improve detection of fetal heart disease

How can we help clinicians trust AI when it is correct, and ignore it when it is wrong? Our new paper in #uogjournal explores this in the context of fetal CHD and ultrasound: obgyn.onlinelibrary.wiley.com/doi/10.1002/uo… @jmsimpsonmd @BernhardKainz1 @trisha_vig

It was a pleasure to present the preliminary results of the PROMETHEUS trial at @BMUS_Ultrasound this week, showing how AI can be used in fetal ultrasound. Full paper coming soon!

Here’s another of the many 𝘀𝗽𝗲𝗮𝗸𝗲𝗿 𝗵𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀 of 𝗨𝗹𝘁𝗿𝗮𝘀𝗼𝘂𝗻𝗱 𝟮𝟬𝟮𝟯. For more details, see our 𝗖𝗼𝗻𝗳𝗲𝗿𝗲𝗻𝗰𝗲 𝗣𝗿𝗼𝗴𝗿𝗮𝗺𝗺𝗲 at bmus.org/ultrasound-202…

@DrXiaoLiu We've just finished PROMETHEUS, our randomised controlled trial of AI to assist in fetal ultrasound screening. Writing up the paper now. ifindproject.com/news-and-media/

UK #AI #ML in #healthcare folks: Could you tell me who is doing prospective clinical #evaluations of AI products in the UK right now (observational trials, RCTs, silent trials, etc)?

One of our latest papers develops diffusion models to generate video sequences from single ultrasound images, combined with clinical parameters. 📈 This work holds potential for robust operator training, counterfactual generation & clinical insights; next week at #MICCAI2023 🚀

@jmsimpsonmd @EvelinaLondon @KingsImaging @GSTTresearch Nice work Tom, John & team. Neatly comparing to current standard, you elegantly demonstrate the importance of screen positive rates. For most fetal conditions (rare) it remains high-a BIG challenge. We must find ways to reduce this. Not easy given characteristics of ultrasound.

How good do AI models need to be for fetal screening? We show that specificity is key!

Who wins the prize for the cheesiest smile 🧀🤓? Great discussions in @thomasgday and my 'fetal face and heart imaging' PPI group. Facilitated by the wonderful @MaryNewburn1 and Vita! Always humbled to hear our #PPI members' care journeys and perspectives - thank you ❤️

Ta da! Fetal cardiologist Tom Day presenting some work on AI in fetal cardiology from ⁦@EvelinaLondon⁩ at the world congress ⁦@8thWCPCCS2023⁩ ⁦@thomasgday⁩

More of the @EvelinaLondon cardiac team on the international stage @8thWCPCCS2023 Trisha Vigneswaran is our Fetal Cardiac lead and was talking on incorporating fetal CMR into the clinical service @trisha_vig

Strong representation from the @evelinalondon cardiac team at World Congress this year! @8thWCPCCS2023

If AI is to be used successfully in fetal ultrasound, collaboration will be required between the AI and human operators. We explore the implications, benefits and pitfalls of this in our new opinion piece, with @jmsimpsonmd @BernhardKainz1 @JackieMatthewJM obgyn.onlinelibrary.wiley.com/doi/10.1002/uo…