The workshop should be hands on and do lots of scans. Remember 20-30 is for quantitative (measuring gastric volume by tracing) which is tedious. I’d concentrate on the qualitative (image recognition) which is much easier & faster. End of day all you want is “elective = defer, emergency = RSI”.

06.04.2025 05:39 — 👍 0    🔁 0    💬 1    📌 0

Good for you Anthony. About 20-30 supervised scans. The quantitative image from a sagittal probe below the costal margin takes 1 sec and is almost a spot diagnosis.
See Perlas et al. Validation of a mathematical model for ultrasound assessment of gastric volume A & A 2013

06.04.2025 05:30 — 👍 0    🔁 0    💬 1    📌 0

At the end of the day I treat my patient in front of me, balancing risks and benefits of all decisions.
I thank the committee that wrote the guidance and will adapt it to local policy and to the individual patient

05.04.2025 21:14 — 👍 2    🔁 0    💬 0    📌 0

We should be setting better standards, not reducing them. We insist on training in FOB and FONA even though it’s thankfully rare clinically. Training to use gastric US is far simpler, and then have actionable scan results. Why are we dumbing ourselves down? Aim high and set the competency

05.04.2025 21:07 — 👍 2    🔁 0    💬 3    📌 0

Flat disagree.
Why was gastric US not promoted more? A 20 second qualitative screen, or 1 minute quantitative diagnostic scan, to assess gastric volume.
Why promote prokinetics when efficacy cannot be determined?
Aspiration is a *preventable* outcome.
@drnavsidhu.bsky.social

05.04.2025 13:07 — 👍 1    🔁 0    💬 1    📌 0

2nd-gen AI could use different methods; eg. dynamic segmentation models. Extreme frequency (20-45MHz) US is just around the corner - we can start including these high res images into our dataset too. Perhaps this will deliver real clinical value. It’s an exciting time to be working in this space

04.04.2025 01:29 — 👍 1    🔁 0    💬 0    📌 0

Nerve US relies on dynamic scanning. Anatomical variability and low resolution also leads to poor ground truth even amongst expert anaesthesiologists. Finally, our training dataset is limited to 100s of images, compared to 100,000s used in other AI.
2/3

04.04.2025 01:29 — 👍 0    🔁 0    💬 1    📌 0

At the ASURA conf, I spoke that AI allows novices to confidently diagnose obstetrics, lung, gastric & cardiac ultrasound (US). But for nerve blocks: AI performs well on easy but not in hard anatomy

Why? 1st-gen AI used convolutional networks, U-Net & model tuning, but perhaps we shouldn’t
1/3

04.04.2025 01:29 — 👍 2    🔁 0    💬 1    📌 0