Do you routinely measure #MAPSE?
Did you know that MAPSE beats #GLS and #LVEF regarding prognosis?
A new multicenter collab study out, led by @KellmanPeter
Ref: pubmed.gov/35132872#WhyCMR 🫀🧲 #EchoFirst 🫀🔊
A thread 🧵. 1/n
This is worth a trip to blue sky for. Martin Ugander is brilliant and his calculations are highly relevant and very interesting!
They strengthen the case for our study conclusions and I’m glad to get to see the work from a different, clever angle.
nature.com/articles/s41591-0…@PHRIresearch@lunduniversity@NatureMedicine
Finland's Publication Forum (or JUFO @tsv_media) announced that from January 2025 it was downgrading its classification of 271 open access journals to the lowest level, zero (this includes 187 journals from @MDPIOpenAccess and 82 from @FrontiersIn).
julkaisufoorumi.fi/en/news/c…
Great work on #ECV#WhyCMR in mod or severe asymp #AS
led by HJ Lee & SP Lee w @bernhard_gerber, Gerry McCann, Marc Dweck, @PPibarot
✅ ECV did not assoc with AS severity
✅ ECV independently assoc w outcomes
➡️ AS is about myocardium, not just the valve!
Our recent paper:
Iyer, et al. 🫁Lung water density is⬆️in pts at risk of 🫀 failure and is largely independent of conventional #WhyCMR measures. EHJ-IMP, 2024
Is now 🔥 discussed (15 mins) by @Nithiniyer on the EHJ-IMP podcast
academic.oup.com/ehjimp/page…
💦🫁Lung water density is ⬆️ in patients at risk of heart 🫀failure and is largely independent of conventional #WhyCMR measures
Our latest 🇦🇺paper out in #EHJIMP
Led by @Nithiniyer in collab w Calvin Chin in Singapore 🇸🇬and Rich Thompson in Edmonton 🇨🇦
REGISTER TODAY!! STRESS CMR: Hands-On Workshop
Join SCMR for a two-day scientific meeting on visual and quantitative assessment of myocardial perfusion will include a mix of state-of-the art & more advanced presentations on a variety of topics!
scmr.org/event-calendar/scmr…
Diagnostic performance measures in clinical medicine (R2, AUC, OR) are all mathematically related, but far from linearly.
I made these figs to discuss w my PhD students.
@venkmurthy@atwater_brett@hauselin source data: escal.site
Actually it depends on whether the predictor is continuous or binary and, if the latter, the prevalence of the predictor. If a marker has an OR of 20, but is found in only 0.1% of population, AUC will be barely above 0.5.