At test-time, Atlas-ISTN not only produces accurate segmentations of anatomical structures but also establishes correspondences to a canonical atlas space, providing inter-subject registration in the process and enabling additional analyses
Our research paper on Atlas-ISTN for highly accurate and anatomically plausible image segmentation is now published in Medical Image Analysis
Latest joint work from the @HeartFlow-@imperialcollege research lab @BioMedIAICL@ICComputing
Open Access
https://t.co/4zNnXnTWk1
Atlas-ISTN helps overcome noisy pixel-wise predictions and allows for configurable stiffness (λ) of the atlas registration at test-time, as shown below with the letter B
Want to make impactful research and make a difference in LMIC? Join the #VITAL project to work on #AI-enabled #ultrasound to assist cardiac imaging for #ICU patient management. #Postdoc position at @KingsImaging w @OUCRU_Vietnam @CLThwaites @sophieyacoub https://t.co/r47na2Klhq
Our paper on population-level cardiac imaging phenotyping and analysis is finally published on Nature Medicine at https://t.co/eLLLK0TSvh. Sincerely thank all collaborators for your support and @uk_biobank and all volunteers. @ImperialDSI@ImperialBrains@BioMedIAICL@ICComputing
The World Health Organization warns young people: “You are not invincible.”
COVID-19 could "put you in hospital for weeks, or even kill you. Even if you don’t get sick, the choices you make about where you go could be the difference between life and death for someone else.”
I think most people aren’t aware of the risk of systemic healthcare failure due to #COVID19 because they simply haven’t run the numbers yet. Let’s talk math. 1/n