It was an honor to play a small part on this "foundational" work! #datascience #Ophthalmology #ophtwitter

Very interesting paper: using generative AI to produce text or images emits 3 to 4 orders of magnitude *less* CO2 than doing it manually or with the help of a computer. arxiv.org/abs/2303.06219

Everyone! If you know anyone who is interested in taking a supported year off to learn more about AI/ML, please ask them to apply to our internship program!

Bridge2AI F2F in the city of the stars ✨✨

Lack of adoption of ocular imaging standards (DICOM) is bad for research & clinical care. NEI U.S. FDA @ONC_HealthIT have developed a roadmap for interoperability together. We will consider this strongly in future research grants Ophthalmology: sciencedirect.com/science/articl…

Excited that our Diabetes Care, a research journal of the ADA 🌟 review paper and illustration on #AI was referenced by Forbes News! Congrats to Anand Rajesh diabetesjournals.org/care/article/4…

Please come join us as we have our first open house for the NIH #Bridge2AI Program!!

April 18 and 19: NIH Common Fund Bridge2AI open house. Great opportunity for scientists, ethicists, coders, and clinicians, and others to learn about these flagship biomedical datasets & using them to answer big questions. Register: bit.ly/3xavrkA

.NIH Common Fund #Bridge2AI F2F Aaron Lee presenting on AI-READI (aireadi.org) - I’m proud of our collaborative work on new health data license- we consider downstream implications of AI - w/ Barbara Evans, Jorge Contreras, Samm Hurst PhD, MA Fair Data Innovations Hub & Cecilia Lee

Thank you to everyone who came to #ARVO2024 in Seattle this year! It was an amazing conference and wonderful to see everyone again. So much brilliant work in the field for data science and AI in eyecare from around the world!

Excited to share OCTCube (arxiv.org/abs/2408.11227), the first *3D* foundation model for retinal OCT images. OCTCube is trained on 26k 3D OCT volumes encompassing 1.62 million 2D images. OCTCube achieved significantly best performance on 27 out of 29 tasks.

If you use #datasets for #AI in #MedicalResearch, please help us understand whether Datasheets, Healthsheets, data card, data nutrition label, and accountability are best for #Transparency with this survey: ucsd.co1.qualtrics.com/jfe/form/SV_e8…

Retinal vessel plexus differentiation based on OCTA using deep learning ow.ly/S7GO50T9Kn8 Aaron Lee #ophthalmology

#AIREADI Y2 data for #Bridge2AI is now live for download with >165,000 files, 2TB of data on #diabetes on >1000 people. Our marker paper was also published today at Nature Metabolism! rdcu.be/dZyyH Request access: aireadi.org #medtwitter #DataScience #AI

And read about our Marker Paper here! nature.com/articles/s4225…

The inhome environmental sensor has some really interesting data and could give a lot more information about each individual’s exposome!

V2 has more than a 1000 participants data and is available at aireadi.org

Excited to announce the publication of our paper in Nature Communications showing a solution for arriving at the Retinal Pigment Score and finding biomarkers to use instead of race: ethnicity is not biology. Congrats to Anand Rajesh! nature.com/articles/s4146…

Thrilled to announce Cecilia & I will join WashU Medicine where I'll serve as Chair of the John F. Hardesty, MD, Dept of Ophthalmology & Visual Sciences. Grateful to University of Washington & Russ Van Gelder for incredible mentorship, and Todd Margolis for his leadership! medicine.washu.edu/news/lee-named…

If you are attending MICCAI this year and are interested in ophthalmic imaging, please participate in the challenge being hosted by Mediwhale where our NIH Bridge2AI dataset (AI-READI) is being used! mucard.mediwhale.net